Menu
CS

Search

Persons

prof. Ing. Roman Čmejla, CSc.

Research topics

Signal processing

All publications

Correlation of spectral moments and age in children's diadochokinetic tasks

  • Authors: Maduda, S., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 100. akustický seminář. Praha: Czech Technical University in Prague, 2021. p. 61-64. ISBN 978-80-01-06888-5.
  • Year: 2021
  • Department: Department of Circuit Theory
  • Annotation:
    Oral diadochokinetic tasks (DDK) are standard motor speech assessment tool often used in children speech-impairment assessment. While main interest of DDK tasks research is temporal domain, mostly maximal speech rate, spectral domain could also bring new insights in the speech impairment research. In this study spectra of plosive consonants obtained through the DDK tasks were compared between children from 3 to 16 years of age. Spectrum was represented by four spectral moments – single numbers describing the shape of spectrum. Spectral moments were tested for correlation with age of speakers to confirm changes in spectrum over the period of growing up. Study showed significant correlation between spectral moments and age for all plosive consonants (/p/, /t/, /k/). These results suggest differences in pronunciation of syllables /pa/-/ta/-/ka/ over time that manifest in changes of spectrum.

Effect of Ageing on Acoustic Characteristics of Voice Pitch and Formants in Czech Vowels

  • DOI: 10.1016/j.jvoice.2020.02.022
  • Link: https://doi.org/10.1016/j.jvoice.2020.02.022
  • Department: Department of Circuit Theory
  • Annotation:
    Background: The relevance of formant-based measures has been noted across a spectrum of medical, technical, and linguistic applications. Therefore, the primary aim of the study was to evaluate the effect of ageing on vowel articulation, as the previous research revealed contradictory findings. The secondary aim was to provide normative acoustic data for all Czech monophthongs. Methods: The database consisted of 100 healthy speakers (50 men and 50 women) aged between 20 and 90. Acoustic characteristics, including vowel duration, vowel space area (VSA), fundamental frequency (fo), and the first to fourth formant frequencies (F1−F4) of 10 Czech vowels were extracted from a reading passage. In addition, the articulation rate was calculated from the entire duration of the reading passage. Results: Age-related changes in pitch were sex-dependent, while age-related alterations in F2/a/, F2/u/, VSA, and vowel duration seemed to be sex-independent. In particular, we observed a clear lowering of fo with age for women, but no change for men. With regard to formants, we found lowering of F2/a/ and F2/u/ with increased age, but no statistically significant changes in F1, F3, or F4 frequencies with advanced age. Although the alterations in F1 and F2 frequencies were rather small, they appeared to be in a direction against vowel centralization, resulting in a significantly greater VSA in the older population. The greater VSA was found to be related partly to longer vowel duration. Conclusions: Alterations in vowel formant frequencies across several decades of adult life appear to be small or in a direction against vowel centralization, thus indicating the good preservation of articulatory precision in older speakers.

Modeling Dysphonia Severity as a Function of Roughness and Breathiness Ratings in the GRBAS Scale

  • Authors: Ferrer, C.A., Aragón, E., Hdez-Díaz, M.E., de Bodt, M.S., prof. Ing. Roman Čmejla, CSc., Englert, M., Behlau, M., Nöth, E.
  • Publication: Proceedings of Interspeech 2021. Brno: Brno University of Technology, 2021. p. 4835-4839. ISSN 1990-9772. ISBN 9781713836902.
  • Year: 2021
  • DOI: 10.21437/Interspeech.2021-1540
  • Link: https://doi.org/10.21437/Interspeech.2021-1540
  • Department: Department of Circuit Theory
  • Annotation:
    Dysphonia comprises many perceptually deviating aspects of voice, and its overall severity perception is made by the listener according to methods of aggregating the single dimensions which are personally conceived and not well studied. Roughness and breathiness are constituent dimensions in most devised rating scales in clinical use. In this paper, we evaluate several ways to model the mapping of the overall severity as a function of the particular ratings of roughness and breathiness. The models include the simple linear averaging as well as several non-linear variants suggested elsewhere, and some minor adjustments. The models are evaluated on four datasets from different countries, allowing a more global evaluation of how the mapping is conceived. Results show the limitations of the most widely assumed linear approach, while also hinting at a need for a more uniform coverage of the sample space in voice pathology datasets. The models explored in this paper can be expanded to higher-dimensional scales.

Speaking rate estimation in Czech children using a phoneme recognizer

  • Authors: Vimr, J., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 100. akustický seminář. Praha: Czech Technical University in Prague, 2021. p. 83-86. ISBN 978-80-01-06888-5.
  • Year: 2021
  • Department: Department of Circuit Theory
  • Annotation:
    We analysed speech rate as well as articulation rate in utterances from 243 healthy Czech children between the age of 5 and 16 years by utilizing a phoneme recognizer to detect syllables and speech pauses. We observed that although this approach is believed to be relatively precise in comparison to other freely available methods there is a tendency to underestimate both speech rate and articulation rate in faster speech and overestimate in slower speech.

Acoustic Tracking of Pitch, Modal, and Subharmonic Vibrations of Vocal Folds in Parkinson's Disease and Parkinsonism

  • DOI: 10.1109/ACCESS.2019.2945874
  • Link: https://doi.org/10.1109/ACCESS.2019.2945874
  • Department: Department of Circuit Theory
  • Annotation:
    The prominent and early presence of dysphonia is considered a valuable marker for differentiation of idiopathic Parkinsons disease and parkinsonian syndromes. Objective quantification of vibrational regimes represented by modal and subharmonic vibrations may thus be vital for improving accuracy of diagnostic decision. The rationale for analyzing vibrational regimes is that abnormal subharmonic vibrations might be the key factor causing dysphonia in parkinsonian syndromes. This study introduces a new fully automated methodology based on robust pitch tracker for decoupling vibrations controlled by laryngeal muscles from the effect of subharmonics that provides distinguishing features of Parkinsons disease and atypical parkinsonian syndromes. We tested the method on resynthesized signals with known parameters and demonstrated that vibrations controlled by laryngeal muscles as well as subharmonics can be detected reliably with a precision that outperforms available technologies. We analysed 337 sustained vowels of 22 patients with PD, 21 patients with multiple system atrophy, 18 patients with progressive supranuclear palsy and 22 healthy controls. Our results showed that subharmonics are more prominent in atypical parkinsonian syndromes compared to Parkinsons disease. Also, increased modulation by laryngeal muscles appears to be a distinctive symptom of multiple system atrophy. Developed algorithm and proposed resynthesized voice signals provide further critical step to understanding and evaluation of dysphonia in Parkinsonism.

The sub-regional functional organization of neocortical irritative epileptic networks in pediatric epilepsy

  • DOI: 10.3389/fneur.2018.00184
  • Link: https://doi.org/10.3389/fneur.2018.00184
  • Department: Department of Circuit Theory
  • Annotation:
    Between seizures irritative network generates frequent brief synchronous activity, which manifests on the EEG as interictal epileptiform discharges (IEDs). Recent insights into the mechanism of IEDs at the microscopic level have demonstrated a high variance in the recruitment of neuronal populations generating IEDs and a high variability in the trajectories through which IEDs propagate across the brain. These phenomena represent one of the major constraints for precise characterization of network organization and for the utilization of IEDs during presurgical evaluations. We have developed a new approach to dissect human neocortical irritative networks and quantify their properties. We have demonstrated that irritative network has modular nature and it is composed of multiple independent sub-regions, each with specific IED propagation trajectories and differing in the extent of IED activity generated. The global activity of the irritative network is determined by long-term and circadian fluctuations in sub-region spatiotemporal properties. Also, the most active sub-region co-localizes with the seizure onset zone in 12/14 cases. This study demonstrates that principles of recruitment variability and propagation are conserved at the macroscopic level and that they determine irritative network properties in humans. Functional stratification of the irritative network increases the diagnostic yield of intracranial investigations with the potential to improve the outcomes of surgical treatment of neocortical epilepsy.

Automated analysis of connected speech reveals early biomarkers of Parkinson’s disease in patients with rapid eye movement sleep behaviour disorder

  • DOI: 10.1038/s41598-017-00047-5
  • Link: https://doi.org/10.1038/s41598-017-00047-5
  • Department: Department of Circuit Theory
  • Annotation:
    For generations, the evaluation of speech abnormalities in neurodegenerative disorders such as Parkinson’s disease (PD) has been limited to perceptual tests or user-controlled laboratory analysis based upon rather small samples of human vocalizations. Our study introduces a fully automated method that yields significant features related to respiratory deficits, dysphonia, imprecise articulation and dysrhythmia from acoustic microphone data of natural connected speech for predicting early and distinctive patterns of neurodegeneration. We compared speech recordings of 50 subjects with rapid eye movement sleep behaviour disorder (RBD), 30 newly diagnosed, untreated PD patients and 50 healthy controls, and showed that subliminal parkinsonian speech deficits can be reliably captured even in RBD patients, which are at high risk of developing PD or other synucleinopathies. Thus, automated vocal analysis should soon be able to contribute to screening and diagnostic procedures for prodromal parkinsonian neurodegeneration in natural environments.

Distinct patterns of imprecise consonant articulation among Parkinson's disease, progressive supranuclear palsy and multiple system atrophy

  • DOI: 10.1016/j.bandl.2016.11.005
  • Link: https://doi.org/10.1016/j.bandl.2016.11.005
  • Department: Department of Circuit Theory
  • Annotation:
    Distinct speech characteristics that may aid in differentiation between Parkinson's disease (PD), progressive supranuclear palsy (PSP) and multiple system atrophy (MSA) remain tremendously under-explored. Here, the patterns and degree of consonant articulation deficits across voiced and voiceless stop plosives in 16 PD, 16 PSP, 16 MSA and 16 healthy control speakers were evaluated using acoustic and perceptual methods. Imprecise consonant articulation was observed across all Parkinsonian groups. Voice onset time of voiceless plosives was more prolonged in both PSP and MSA compared to PD, presumably due to greater severity of dysarthria and slower articulation rate. Voice onset time of voiced plosives was significantly shorter only in MSA, likely as a consequence of damage to cerebellar structures. In agreement with the reduction of pre-voicing, MSA manifested increased number of voiced plosives misclassified as voiceless at perceptual evaluation. Timing of articulatory movements may provide important clues about the pathophysiology of underlying disease. (C) 2016 Elsevier Inc. All rights reserved.

Dysprosody Differentiate Between Parkinson’s Disease, Progressive Supranuclear Palsy, and Multiple System Atrophy

  • DOI: 10.21437/Interspeech.2017-762
  • Link: https://doi.org/10.21437/Interspeech.2017-762
  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson’s disease (PD), progressive supranuclear palsy (PSP), and multiple system atrophy (MSA) are distinctive neurodegenerative disorders, which manifest similar motor features. Their differentiation is crucial but difficult. Dysfunctional speech, especially dysprosody, is a common symptom accompanying PD, PSP, and MSA from early stages. We hypothesized that automated analysis of monologue could provide speech patterns distinguishing PD, PSP, and MSA. We analyzed speech recordings of 16 patients with PSP, 20 patients with MSA, and 23 patients with PD. Our findings revealed that deviant pause production differentiated between PSP, MSA, and PD. In addition, PSP showed greater deficits in speech respiration when compared to MSA and PD. Automated analysis of connected speech is easy to administer and could provide valuable information about underlying pathology for differentiation between PSP, MSA, and PD.

ACTIVITY OF UPPER BODY MUSCLES IN DOUBLE POLING AND SKIERG WORKOUT

  • Authors: Horyna, R., Bačáková, R., Chrástková, R., Sedlák, J., prof. Ing. Roman Čmejla, CSc., Kračmar, B.
  • Publication: Baltic Journal of Sport & Health Sciences. 2016, 101(2), 31-37. ISSN 2351-6496.
  • Year: 2016
  • Department: Department of Circuit Theory
  • Annotation:
    Background. The aim of the study was to compare the involvement of upper body muscles during double poling and SkiErg Concept 2 workout and verify its specificity for cross-country skiing. Methods. Ten elite Czech cross-country skiers performed double poling and SkiErg workout. Electromyography of selected upper body muscles and cycle characteristics were analysed. To monitor the electrical activity of muscles, we used the device ME6000. Data were analysed using Mega Win and MATLAB software version R2012b. Results. Relative poling phase during double poling was 30.30 ± 2.02% and during SkiErg workout 54 ± 3.36%. Pre-activation of trunk flexors was significantly higher during double poling due to high and forward body position before pole plant. Pre-activation of trunk flexors was not significantly different as pre-activation of shoulder and elbow extensors during SkiErg workout. Deactivation of these muscles came significantly later during SkiErg workout. Conclusion. SkiErg cannot be considered a specific training method for cross-country skiing. It can be recommended to obtain specific power, but long-term application may cause disruption of double poling technique, especially timing of trunk flexors, shoulder and elbow extensors.

Circadian Dynamics of High Frequency Oscillations in Patients with Epilepsy

  • DOI: 10.5220/0005827602840289
  • Link: https://doi.org/10.5220/0005827602840289
  • Department: Department of Circuit Theory
  • Annotation:
    High frequency oscillations (HFOs) are novel biomarker of epileptogenic tissue. HFOs are currently used to localize the seizure generating areas of the brain, delineate the resection and to monitor the disease activity. It is well established that spatiotemporal dynamics of HFOs can be modified by sleep-wake cycle. In this study we aimed to evaluate in detail circadian and ultradian changes in HFO dynamics using techniques of automatic HFO detection. For this purpose we have developed and implemented novel algorithm to automatic detection and analysis of HFOs in long-term intracranial recordings of six patients. In 5/6 patients HFO rates significantly increased during NREM sleep. The largest NREM related increase in HFO rates were observed in brain areas which spatially overlapped with seizure onset zone. Analysis of long-term recording revealed existence of ultradian changes in HFO dynamics. This study demonstrated reliability of automatic HFO detection in the analysis of long-term intracranial recordings in humans. Obtained results can foster practical implementation of automatic HFO detecting algorithms into presurgical examination, dramatically decrease human labour and increase the information yield of HFOs

Effects of dopaminergic replacement therapy on motor speech disorders in Parkinson's disease: longitudinal follow-up study on previously untreated patients

  • DOI: 10.1007/s00702-016-1515-8
  • Link: https://doi.org/10.1007/s00702-016-1515-8
  • Department: Department of Circuit Theory
  • Annotation:
    Although speech disorders represent an early and common manifestation of Parkinson’s disease (PD), little is known about their progression and relationship to dopaminergic replacement therapy. The aim of the current study was to examine longitudinal motor speech changes after the initiation of pharmacotherapy in PD. Fifteen newly-diagnosed, untreated PD patients and ten healthy controls of comparable age were investigated. PD patients were tested before the introduction of antiparkinsonian therapy and then twice within the following 6 years. Quantitative acoustic analyses of seven key speech dimensions of hypokinetic dysarthria were performed. At baseline, PD patients showed significantly altered speech including imprecise consonants, monopitch, inappropriate silences, decreased quality of voice, slow alternating motion rates, imprecise vowels and monoloudness. At follow-up assessment, preservation or slight improvement of speech performance was objectively observed in two-thirds of PD patients within the first 3–6 years of dopaminergic treatment, primarily associated with the improvement of stop consonant articulation. The extent of speech improvement correlated with l-dopa equivalent dose (r = 0.66, p = 0.008) as well as with reduction in principal motor manifestations based on the Unified Parkinson’s Disease Rating Scale (r = −0.61, p = 0.02), particularly reflecting treatment-related changes in bradykinesia but not in rigidity, tremor, or axial motor manifestations. While speech disorders are frequently present in drug-naive PD patients, they tend to improve or remain relatively stable after the initiation of dopaminergic treatment and appear to be related to the dopaminergic responsiveness of bradykinesia.

Hypernasality associated with basal ganglia dysfunction: evidence from Parkinson’s disease and Huntington’s disease

  • DOI: 10.7717/peerj.2530
  • Link: https://doi.org/10.7717/peerj.2530
  • Department: Department of Circuit Theory
  • Annotation:
    Background. Although increased nasality can originate from basal ganglia dysfunction, data regarding hypernasality in Parkinson's disease (PD) and Huntington's disease (HD) are very sparse. The aim of the current study was to analyze acoustic and perceptual correlates of velopharyngeal seal closure in 37 PD and 37 HD participants in comparison to 37 healthy control speakers. Methods. Acoustical analysis was based on sustained phonation of the vowel /i/ and perceptual analysis was based on monologue. Perceptual analysis was performed by 10 raters using The Great Ormond Street Speech Assessment '98. Acoustic parameters related to changes in a 1/3-octave band centered on 1 kHz were proposed to reflect nasality level and behavior through utterance. Results. Perceptual analysis showed the occurrence of mild to moderate hypernasality in 65% of PD, 89% of HD and 22% of control speakers. Based on acoustic analyses, 27% of PD, 54% of HD and 19% of control speakers showed an increased occurrence of hypernasality. In addition, 78% of HD patients demonstrated a high occurrence of intermittent hypernasality. Further results indicated relationships between the acoustic parameter representing fluctuation of nasality and perceptual assessment (r D 0:51, p<0:001) as well as the Unified Huntington Disease Rating Scale chorea composite subscore (r D0:42, pD0:01). Conclusions. In conclusion the acoustic assessment showed that abnormal nasality was not a common feature of PD, whereas patients with HD manifested intermittent hypernasality associated with chorea.

Longitudinal Study of Age-related Changes in Acoustic Characteristics of Speech

  • Department: Department of Circuit Theory
  • Annotation:
    In this paper a longitudinal study is presented of age-dependent speech acoustic characteristics from the utterances of Czech actresses, acquired from movie databases recorded in the years from 1985 to 2014. The age dependence of phonation is analyzed using the parameters fundamental frequency, jitter, and shimmer. For assessment of articulation, the formant-based voice area index of vocals and the length of the burst from a palatal stop consonant is used. For prosody, no characteristics were found that were applicable to the analysis of film archives. The results of age dependencies described by the phonetic categories are consistent with the assumptions and conclusions presented in the literature. This pilot project also confirms the possibility of creating valuable studies from publicly available sources.

Robust detection of fundamental frequency in dysarthric speech

  • Department: Department of Circuit Theory
  • Annotation:
    This study present robust method for detection of fundamental frequency (F0) in dysarthric speech. Method was validated on large sample of 345 artificial phonations based on real data. Artificial signals were synthesized from parameters of F0 contour, jitter, shimmer, and harmonic-to-noise ratio measured on real signals using supervised PRAAT software. The method compares F0 probability model of speaker with F0 candidates obtained as local maxima of frequency spectrum. The F0 probability model is predicted by Kalman filter. Method detect pitch breaks using cepstral analysis. Presented method showed sufficiently high sensitivity compared to conventional methods of F0 detection (PRAAT,DYPSA, RAPT, PEFAC, SHRP).

Speech changes after coordinative training in patients with cerebellar ataxia: A pilot study

  • DOI: 10.1007/s10072-015-2379-7
  • Link: https://doi.org/10.1007/s10072-015-2379-7
  • Department: Department of Circuit Theory
  • Annotation:
    Objectives: Although rehabilitative training is a necessary adjunct in the management of gait ataxia, it remains unknown whether the possible beneficial effect of intensive coordinative training may translate to activities of daily living, which are closely connected with postural alignment. The aim of the present study was to examine the effectiveness of a 2-week intensive coordinative motor training on speech production. Methods: Speech and motor performances in a cohort of 10 individuals with cerebellar degeneration were examined three times; before the introduction of training; directly and four weeks after the last training session. Each patient was instructed to perform a speaking task of fast syllable repetition and monologue. Objective acoustic analyses were used to investigate six key aspects of speech production disturbed in ataxic dysarthria including accuracy of consonant articulation, accuracy of vowel articulation, irregular alternating motion rates, prolonged phonemes, slow alternating motion rates and inappropriate segmentation. Results: We found that coordinative training had a mild beneficial effect on speech in cerebellar patients. Immediately after the last training session, slight speech improvements were evident in all 10 patient. Furthermore, follow-up assessment performed 4 weeks later revealed that 90% of the patients showed better speech performance than before initiation of the therapy.

ANALYSIS OF HIGH-FREQUENCY ACTIVITY IN ELECTROENCEPHALOGRAM OF ANIMAL EPILEPSY MODEL

  • Authors: Havel, T., Kudláček, J., Jiruška, P., Ing. Petr Ježdík, Ph.D., prof. Ing. Roman Čmejla, CSc.,
  • Publication: INTERNATIONAL WORKSHOP ON COMPUTATIONAL INTELLIGENCE FOR MULTIMEDIA UNDERSTANDING. Praha: Czech Technical University in Prague, 2015. ISBN 978-1-4673-8457-5.
  • Year: 2015
  • DOI: 10.1109/IWCIM.2015.7347072
  • Link: https://doi.org/10.1109/IWCIM.2015.7347072
  • Department: Department of Circuit Theory
  • Annotation:
    High frequency oscillations (HFO) are believed to be a new specific biomarker of epileptogenic tissue. According to the reported findings, we hypothesized that HFOs occur more specific to the epileptogenic tissue. To test this hypothesis, we analyzed intracranial electroencephalograms (iEEG) of three subjects of animal epilepsy model. Overall 180 minutes of iEEG records were processed by the automatic high frequency activity detector and further analyzed. The dominant frequency of each segment was determined and categorized as ripples (80-200 Hz) or fast ripples (200-1000 Hz). Even though overall number of HFO detections predominate in the hemisphere where epileptic focus is located, in comparison of each brain structure separately our hypothesis cannot be fully confirmed.

Automatic detection of voice onset time in dysarthric speech

  • DOI: 10.1109/ICASSP.2015.7178790
  • Link: https://doi.org/10.1109/ICASSP.2015.7178790
  • Department: Department of Circuit Theory
  • Annotation:
    Although a number of speech disorders reflect varying involvement of brain areas, recently published automatic speech analyses have primarily been limited to hypokinetic dysarthria in Parkinson's disease (PD). Therefore, the aim of the present study was to provide an automatic algorithm suitable for the assessment of voice onset time (VOT) in various dysarthria types.

Automatic evaluation of speech rhythm instability and acceleration in dysarthrias associated with basal ganglia dysfunction

  • DOI: 10.3389/fbioe.2015.00104
  • Link: https://doi.org/10.3389/fbioe.2015.00104
  • Department: Department of Circuit Theory
  • Annotation:
    Speech rhythm abnormalities are commonly present in patients with different neurodegenerative disorders. These alterations are hypothesized to be a consequence of disruption to the basal ganglia circuitry involving dysfunction of motor planning, programming and execution, which can be detected by a syllable repetition paradigm. Therefore, the aim of the present study was to design a robust signal processing technique that allows the automatic detection of spectrally-distinctive nuclei of syllable vocalizations and to determine speech features that represent rhythm instability and acceleration. A further aim was to elucidate specific patterns of dysrhythmia across various neurodegenerative disorders that share disruption of basal ganglia function. Speech samples based on repetition of the syllable /pa/ at a self-determined steady pace were acquired from 109 subjects, including 22 with Parkinson's disease (PD), 11 progressive supranuclear palsy (PSP), 9 multiple system atrophy (MSA), 24 ephedrone-induced parkinsonism (EP), 20 Huntington's disease (HD), and 23 healthy controls. Subsequently, an algorithm for the automatic detection of syllables as well as features representing rhythm instability and rhythm acceleration were designed. The proposed detection algorithm was able to correctly identify syllables and remove erroneous detections due to excessive inspiration and nonspeech sounds with a very high accuracy of 99.6%. Instability of vocal pace performance was observed in PSP, MSA, EP and HD groups. Significantly increased pace acceleration was observed only in the PD group. Although not significant, a tendency for pace acceleration was observed also in the PSP and MSA groups. Our findings underline the crucial role of the basal ganglia in the execution and maintenance of automatic speech motor sequences. We envisage the current approach to become the first step towards the development of acoustic technologies allowing automated assessment of rhythm in dysarthrias.

CASE STUDY OF INTRACRANIAL EEG RECORDS OF PATIENTS WITH FOCAL CORTICAL DYSPLASIA TYPE I AND II

  • Authors: Balach, J., prof. Ing. Roman Čmejla, CSc., Ing. Petr Ježdík, Ph.D., Kršek, P., Jiruška, P.
  • Publication: INTERNATIONAL WORKSHOP ON COMPUTATIONAL INTELLIGENCE FOR MULTIMEDIA UNDERSTANDING. Praha: Czech Technical University in Prague, 2015. p. 1-4. ISBN 978-1-4673-8457-5.
  • Year: 2015
  • DOI: 10.1109/IWCIM.2015.7347080
  • Link: https://doi.org/10.1109/IWCIM.2015.7347080
  • Department: Department of Circuit Theory
  • Annotation:
    In this study we try to find out if it is possible to differentiate type of focal cortical dysplasia by features obtained from intracranial EEG. We compare occurrence and rates of three biomarkers present in epilepsy in patients with focal cortical dysplasia type I and II. Case study is made on long term night records of 6 pediatric patients. Detection of interictal epileptiform discharges and high-frequency oscillations is made by automated algorithms, delta brush are marked visually. Position of lesion and electrodes inside were obtained from MRI. In individual rates were not found difference on significant level. No major significance were found, but as promising seem to be ratio inside to outside rates of high-frequency oscillations and presence of delta brush, which were found

Comparison of developmental and neurogenic stuttering

  • Department: Department of Circuit Theory
  • Annotation:
    Although the full etiological nature of developmental stuttering is still unknown, the key role of disturbed basal ganglia function along with the role of dopamine system have been thoroughly discussed. Therefore, the aim of the current study was to survey the characteristics of neurogenic stuttering based on patients with Parkinson's disease (PD) and compared them to the characteristics of developmental stuttering. The database consists of 14 persons with developmental stuttering (pDS) and 14 patients with idiopathic PD. In addition, 14 sex-matched healthy controls (HC) were recruited. Each participant was instructed to perform two-minutes long monolog on given topic. Analysis of dysfluency was conducted according to the Lidcombe behavioral taxonomy of stuttering. Our results showed that pDS subjects manifested significantly more vocal blocs (p 0.01) and filling words (p 0.05) compared to PD patients. On the other hand, the neurogenic stuttering of PD was mainly characterized by incomplete syllable repetitions (p 0.05) and prolongations (p 0.01) as compared to HC. In conclusion, our study demonstrates that there seems to be more differences rather than similarities between neurogenic and developmental stuttering.

Detection of Interictal Epileptiform Discharges Using Signal Envelope Distribution Modelling: Application to Epileptic and Non-Epileptic Intracranial Recordings

  • DOI: 10.1007/s10548-014-0379-1
  • Link: https://doi.org/10.1007/s10548-014-0379-1
  • Department: Department of Circuit Theory
  • Annotation:
    Interictal epileptiform discharges (spikes, IEDs) are electrographic markers of epileptic tissue and their quantification is utilized in planning of surgical resection. Visual analysis of long-term multi-channel intracranial recordings is extremely laborious and prone to bias. Development of new and reliable techniques of automatic spike detection represents a crucial step towards increasing the information yield of intracranial recordings and to improve surgical outcome. In this study, we designed a novel and robust detection algorithm that adaptively models statistical distributions of signal envelopes and enables discrimination of signals containing IEDs from signals with background activity. This detector demonstrates performance superior both to human readers and to an established detector. It is even capable of identifying low-amplitude IEDs which are often missed by experts and which may represent an important source of clinical information. Application of the detector to non-epileptic intracranial data from patients with intractable facial pain revealed the existence of sharp transients with waveforms reminiscent of interictal discharges that can represent biological sources of false positive detections. Identification of these transients enabled us to develop and propose secondary processing steps, which may exclude these transients, improving the detector’s specificity and having important implications for future development of spike detectors in general.

Effect of dopaminergic medication on speech dysfluency in Parkinson's disease: a longitudinal study

  • DOI: 10.1007/s00702-015-1363-y
  • Link: https://doi.org/10.1007/s00702-015-1363-y
  • Department: Department of Circuit Theory
  • Annotation:
    Although speech dysfluencies have been hypothesized to be associated with abnormal function of dopaminergic system, the effects of dopaminergic medication on speech fluency in Parkinson's disease (PD) have not been systematically studied. The aim of the present study was, therefore, to investigate the long-term effect of dopaminergic medication on speech fluency in PD. Fourteen de novo PD patients with no history of developmental stuttering and 14 age- and sex-matched healthy controls (HC) were recruited. PD subjects were examined three times; before the initiation of dopaminergic treatment and twice in following 6 years. The percentage of dysfluent words was calculated from reading passage and monolog. The amount of medication was expressed by cumulative doses of L-dopa equivalent. After 3-6 years of dopaminergic therapy, PD patients exhibited significantly more dysfluent events compared to healthy subjects as well as to their own speech performance before the introduction of dopaminergic therapy. In addition, we found a strong positive correlation between the increased occurrence of dysfluent words and the total cumulative dose of L-dopa equivalent. Our findings indicate an adverse effect of prolonged dopaminergic therapy contributing to the development of stuttering-like dysfluencies in PD. These findings may have important implication in clinical practice, where speech fluency should be taken into account to optimize dopaminergic therapy.

Grunting in a Genetically Modified Minipig Animal Model for Huntington’s Disease – a Pilot Experiments

  • DOI: 10.14735/amcsnn20152S61
  • Link: https://doi.org/10.14735/amcsnn20152S61
  • Department: Department of Circuit Theory
  • Annotation:
    Huntington’s disease (HD) is an autosomal-dominant neurodegenerative disorder characterized by the impairment of voluntary and involuntary movements, behavioral disorders and cognitive decline. Besides the main motor symptoms, voice and speech disorders have been documented in a large majority of patients with HD. The animal model of pigs is often used in preclinical studies. Although there are obvious differences in the anatomy of the articulation organs between pigs and humans, the same trends in pathophysiological mechanisms can be expected in both grunting and human phonation. The main aim of the study was therefore to design a suitable experiment that would allow for acquisition of a sufficiently long recording of grunting from as many pigs as possible. The second goal was to perform the final version of the experiment in all available pigs and to evaluate the amount and quality of the acquired recordings. The database consists of 17 HD transgenic minipigs and 16 healthy siblings. Tested variants of the experiment, performed on subgroup of four sows, were divided into four subgroups: (a) positive – feeding, (b) positive – sound stimulation, (c) negative – hindering in movement, (d) negative – unpleasant touch. The evaluation of the quality of the elicited recording was performed using audio software where pure pig grunting was selected and all acoustic artefacts deleted. The best results were reached using the experiment in which: (i) a recording device is put on the pig’s body, (ii) the pig is left alone for few minutes in the pen in order to calm down, and (iii) a person enters the room and tries to offer the pig food while walking backwards. As a result, the pig follows the person and grunts. Sufficiently long (20 single grunts or more) and clear recordings were received from 24 out of 33 pigs (73%). The realisation of the experiment is therefore possible.

Hypernasality in dysarthric speech

  • Department: Department of Circuit Theory
  • Annotation:
    Study is aimed on presence of hypernasality in dysarthric utterances of Parkinson' s disease and Huntington's disease sufferers.

Language-independent method for analysis of German stuttering recordings

  • Authors: Lustyk, T., Bergl, P., Haderlein, T., Nöth, E., prof. Ing. Roman Čmejla, CSc.,
  • Publication: INTERSPEECH 2015. Bochum: ISCA - International Speech Communication Association, 2015. ISSN 2308-457X.
  • Year: 2015
  • Department: Department of Circuit Theory
  • Annotation:
    The paper describes experiments where automatic acoustic algorithms initially intended to be used on Czech stuttering speakers were applied on recordings of German stuttering speakers. Four algorithms based on voice activity and abrupt spectral changes detection are introduced. The database consists of 34 speakers. The measure, the number of abrupt spectral changes in speech segments, reached a correlation with fluency rating of 0.85. The other measures have also good agreement with subjective evaluation. Results indicate that it could be basically possible to do language–independent analysis of stuttering, here demonstrated on read recordings of German speakers.

Rate and rhythm of speech in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    The aim of this paper was to evaluate rate and rhythm of spoken utterances in Parkinson's disease.

Using of transgenic minipigs as a model for research of Huntington's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Huntington's disease is a rare neurodegenerative disorder characterized by uncoordanitave body movements and decine in cognitive function.

Acoustic Investigation of Stress Patterns in Parkinson's Disease

  • DOI: 10.1016/j.jvoice.2013.07.001
  • Link: https://doi.org/10.1016/j.jvoice.2013.07.001
  • Department: Department of Circuit Theory
  • Annotation:
    Objectives. Although reduced stress is thought to be one of the most deviant speech dimensions in hypokinetic dysarthria associated with Parkinson’s disease (PD), the mechanisms of stress production in PD have not been thoroughly explored by objective methods. The aim of the present study was to quantify the effect of PD on prosodic characteristics and to describe contrastive stress patterns in parkinsonian speech. Methods. The ability of 20 male speakers with early PD and 16 age- and gender-matched healthy controls (HCs) to signal contrastive stress was investigated. Each participant was instructed to unnaturally emphasize five key words while reading a short block of text. Acoustic analyses were based on the measurement of pitch, intensity, and duration. In addition, an innovative measurement termed the stress pattern index (SPI) was designed to mirror the effect of all distinct acoustic cues exploited during stress production. Results. Although PD patients demonstrated a reduced ability to convey contrastive stress, they could still notably increase pitch, intensity, and duration to emphasize a word within a sentence. No differences were revealed between PD and HC stress productions using the measurements of pitch, intensity, duration, and intensity range. However, restricted SPI and pitch range were evident in the PD group. Conclusions. A reduced ability to express stress seems to be the distinctive pattern of hypokinetic dysarthria, even in the early stages of PD. Because PD patients were able to consciously improve their speech performance using multiple acoustic cues, the introduction of speech therapy may be rewarding.

Acoustic Parameters for Classification of Breathiness in Continuous Speech According to the GRBAS Scale

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Journal of Voice. 2014, 28(5), 653.e9-653.e17. ISSN 0892-1997.
  • Year: 2014
  • DOI: 10.1016/j.jvoice.2013.07.016
  • Link: https://doi.org/10.1016/j.jvoice.2013.07.016
  • Department: Department of Circuit Theory
  • Annotation:
    The purpose of this study was to classify objectively breathiness in continuous speech according to a subjective evaluation of voice based on the GRBAS scale. The level of breathiness in continuous speech can be effectively described by automatic system–based analysis of acoustic measures. The proposed automatic system is able to determine the level of breathiness in continuous speech with sufficient precision.

Automatic assessment of speech pauses in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    The study aims on the method of speech / pause classification and feature assessment for healthy speech and speech impaired due to Parkinson's disease.

Automatic Evaluation of Articulatory Disorders in Parkinson's Disease

  • DOI: 10.1109/TASLP.2014.2329734
  • Link: https://doi.org/10.1109/TASLP.2014.2329734
  • Department: Department of Circuit Theory
  • Annotation:
    Although articulatory deficits represent an important manifestation of dysarthria in Parkinson’s disease (PD), the most widely used methods currently available for the automatic evaluation of speech performance are focused on the assessment of dysphonia. The aim of the present study was to design a reliable automatic approach for the precise estimation of articulatory deficits in PD. Twenty-four individuals diagnosed with de novo PD and twenty-two age-matched healthy controls were recruited. Each participant performed diadochokinetic tasks based upon the fast repetition of /pa/-/ta/-/ka/ syllables. All phonemes were manually labeled and an algorithm for their automatic detection was designed. Subsequently, 13 features describing six different articulatory aspects of speech including vowel quality, coordination of laryngeal and supralaryngeal activity, precision of consonant articulation, tongue movement, occlusion weakening, and speech timing were analyzed. In addition, a classification experiment using a support vector machine based on articulatory features was proposed to differentiate between PD patients and healthy controls. The proposed detection algorithm reached approximately 80% accuracy for a 5 ms threshold of absolute difference between manually labeled references and automatically detected positions. When compared to controls, PD patients showed impaired articulatory performance in all investigated speech dimensions (p < 0.05). Moreover, using the six features representing different aspects of articulation, the best overall classification result attained a success rate of 88% in separating PD from controls. Imprecise consonant articulation was found to be the most powerful indicator of PD-related dysarthria. We envisage our approach as the first step towards development of acoustic methods allowing the automated assessment of articulatory features in dysarthrias.

Capture problems with auditory evoked potentials in patients with cochlear implants

  • Authors: Bauer, L., Štrupl, M., prof. Ing. Roman Čmejla, CSc.,
  • Publication: IV. Letní doktorandské dny 2014. Praha: ČVUT FEL, Katedra teorie obvodů, 2014. pp. 70-72. ISBN 978-80-01-05506-9.
  • Year: 2014
  • Department: Department of Circuit Theory
  • Annotation:
    Auditory evoked potentials are used to evaluate the development of auditory centers. The artifact makes impossible to monitor the development of auditory centers, that arises due to cochlear implant roughly 30% of implanted cases. Our goal is to detect and suppress this artifact, but we can not influence the original signal. Measuring of late auditory evoked potentials by HearLab device should serve in the future to establish an objective method of measuring of development of auditory centers after cochlear implantation, which would be beneficial mainly in children who are not able to properly cooperate and give clear feedback. Our proposed detectors artifacts achieve sensitivity 94% and specificity 94% for four adult databases.

Characteristics and occurrence of speech impairment in Huntington's disease: possible influence of antipsychotic medication

  • DOI: 10.1007/s00702-014-1229-8
  • Link: https://doi.org/10.1007/s00702-014-1229-8
  • Department: Department of Circuit Theory
  • Annotation:
    Although motor speech impairment is a common manifestation of Huntington’s disease (HD), its description remains limited. The aim of the current study was therefore to estimate the occurrence and characteristics of speech disorder in HD and to explore the influence of antipsychotic medication on speech performance. Speech samples, including reading passage and monologue, were acquired from 40 individuals diagnosed with HD and 40 age- and sex-matched healthy controls. Objective acoustic analyses were used to evaluate key aspects of speech including vowel articulation, intensity, pitch and timing. A predictive model was constructed to detect the occurrence and most prominent patterns of speech dysfunction in HD. We revealed that 93 % of HD patients manifest some degree of speech impairment. Decreased number of pauses, slower articulation rate, imprecise vowel articulation and excess intensity variations were found to be the most salient patterns of speech dysfunction in HD. We further demonstrated that antipsychotic medication may induce excessive loudness and pitch variations perceptually resembling excess patterns of word stress, and may also accentuate general problems with speech timing. Additionally, antipsychotics induced a slight improvement of vowel articulation. Specific speech alterations observed in HD patients indicate that speech production may reflect the pathophysiology of the disease as well as treatment effects, and may therefore be considered a valuable marker of functional disability in HD.

Comparison of Algorithms for Detection of High Frequency Oscillations in Intracranial EEG

  • Authors: Balach, J., Ing. Petr Ježdík, Ph.D., prof. Ing. Roman Čmejla, CSc., Kršek, P., Marusič, P., Jiruška, P.
  • Publication: 2014 IEEE International Symposium on Medical Measurement and Applications Proceedings. Lisboa: IEEE, 2014. p. 466-469. ISBN 978-1-4799-2920-7.
  • Year: 2014
  • DOI: 10.1109/MeMeA.2014.6860107
  • Link: https://doi.org/10.1109/MeMeA.2014.6860107
  • Department: Department of Circuit Theory, Department of Telecommunications Engineering
  • Annotation:
    The paper is focused on high frequency oscillations and their automated detection. This phenomena occurs in EEG of epileptic patients. Localization of these events lead to epileptogenic tissue. With more information neurologist can make better diagnose. Using detection algorithms can help obtain more information from signals. This should lead to better patient outcome. We discuss the possibility of using automatic detectors of high frequency activity in intracranial EEG as a support for human recording evaluation. Nowadays exist various algorithms for HFO detection with different approach and used methods. We have chosen and implemented several published detectors and developed our own algorithms. We compared their parameters on set of expert labeled records from patients with epilepsy. Results of algorithms are reviewed and discussed their quality output with respect to clinical practice and its utility in the diagnosis of epilepsy.

Detection of High-frequency oscillations in iEEG based on spectral changes

  • Department: Department of Circuit Theory
  • Annotation:
    The detection algorithm described in this paper responses on frequency spectrum changes. Process contains of spectral normalization and classification of each status.

Detection of repetetive patterns in intracranial EEG

  • Department: Department of Circuit Theory
  • Annotation:
    High frequency oscillation (HFO) are one of the biomarkers epileptogení tissue. Visual assessment of long-term EEG recordings is demanding not only time. Therefore it offers the possibility of automatic algorithms for their detection. Out in the previous studies, we examined individual detectors HFO and comparing their quality. In this study, we continue to present our findings and the subsequent possible processing results in order to obtain better results. Our proposed method classifies each detection according to their relative occurrence within the electrodes and the entire recording and also according to their frequency. We assume that false detections will primarily occur sporadically, while the HFO will propagate into surrounding tissues and will be detected within one electrode on multiple channels simultaneously. Using this method, we reassess results from two patients from pediatric neurology and compare the original and the new results rated to neurologist opinion.

Evaluation of disfluent speech by means of automatic acoustic measurements

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA. 2014, 135(3), 1457-1468. ISSN 0001-4966.
  • Year: 2014
  • DOI: 10.1121/1.4863646
  • Link: https://doi.org/10.1121/1.4863646
  • Department: Department of Circuit Theory
  • Annotation:
    An experiment was carried out to determine whether the level of the speech fluency disorder can be estimated by means of automatic acoustic measurements. These measures analyze, for example, the amount of silence in a recording or the number of abrupt spectral changes in a speech signal. All the measures were designed to take into account symptoms of stuttering. In the experiment, 118 audio recordings of read speech by Czech native speakers were employed. The results indicate that the human-made rating of the speech fluency disorder in read speech can be predicted on the basis of automatic measurements. The number of abrupt spectral changes in the speech segments turns out to be the most appropriate measure to describe the overall speech performance. The results also imply that there are measures with good results describing partial symptoms (especially fixed postures without audible airflow).

Functional mapping and testing of excitability of limbic structures in vivo

  • Authors: Kudláček, J., Levčík, D., Vlk, P., Demeterová, L., Otáhal, J., prof. Ing. Roman Čmejla, CSc., Jiruška, P.
  • Publication: IV. Letní doktorandské dny 2014. Praha: ČVUT FEL, Katedra teorie obvodů, 2014, pp. 55-59. ISBN 978-80-01-05506-9.
  • Year: 2014
  • Department: Department of Circuit Theory
  • Annotation:
    Ventral commissure constitutes one of the main connections between the structures of limbic systems of both hemispheres. Anatomical studies proved that the ventral commissure contains pathways interconnecting predominantly hippocampi and entorhinal cortices. In the present study we focused on analysis of functional connectivity and excitability of limbic structures using stimulation ov ventral commissure in vivo. Pilot results show that stimulation of ventral commissure activates mainly dorsal hippocampus and medial entorhinal cortex. Using laminar profile analysis, we determined that evoked responses have character of orthodromicaly induced excitatory post-synaptic potentials. Simultaneously antidromic responses are present. These have character of back-propagated dendritic action potentials. Thus, ventral commissureal stimulation allows not only functional connectivity analysis but also testing of excitability of various neuronal compartments.

Graphical interface for evaluation of high-frequency oscillations detected in intracranial EEG records

  • Department: Department of Circuit Theory
  • Annotation:
    Visual marking of High-frequency oscillations in intracranial EEG records is extremely time demanding. The aproach which make this process faster involve automatic detection followed by visual confirmation. The application created for this purpose is described in this paper.

Hypernasality in dysartric utterances

  • Department: Department of Circuit Theory
  • Annotation:
    Currently, measurement of hypernasality is essential topic in research of patients with cleft lip and palate.

Interactiv GUI for surface EMG analysis

  • Authors: Sedlák, J., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 22nd Annual Conference Proceedings Technical Computing Bratislava 2014. Praha: Humusoft, 2014, pp. 56-57. ISSN 2336-1662. ISBN 978-80-7080-898-6.
  • Year: 2014
  • Department: Department of Circuit Theory
  • Annotation:
    Design and realization of EMG analysis software. The software alows results coparison for the selected EMG onset detection techniques. The software is compiled as Windows Standalone Application because the software does not require the full MATLAB instalation.

Multimodal visualization of the quantitative EEG evaluation to medical images

  • Department: Department of Circuit Theory
  • Annotation:
    The surgical treatment of epilepsy is aimed on removing epileptogenic zone that primary cause seizures.

Muscle activity detection in electromyograms recorded during periodic movements

  • Authors: Špulák, D., prof. Ing. Roman Čmejla, CSc., Bačáková, R., Kračmar, B., Satrapová, L., Novotný, P.
  • Publication: Computers in Biology and Medicine. 2014, 47(1), 93-103. ISSN 0010-4825.
  • Year: 2014
  • DOI: 10.1016/j.compbiomed.2014.01.013
  • Link: https://doi.org/10.1016/j.compbiomed.2014.01.013
  • Department: Department of Circuit Theory
  • Annotation:
    Muscle coordination during periodic movements is often studied using the average envelope of the electromyographic (EMG) signal. We show that this method causes a loss of important information, and potentially gives rise to errors in analysis of muscle activity coordination. We created four simulated two-channel surface EMG signals, in order to compare the results of muscle onset/cessation detection, performed on the average EMG envelope and the EMG envelopes in every single movement cycle. Our results show that the common method using the average EMG envelope is unable to reveal certain important characteristics of the EMG signals, while the analysis performed on individual cycles accentuates this information. This ability was verified on 16-channel surface EMGs obtained during walking and cycling. By detecting muscle activity in individual movement cycles, we could observe fine changes in muscle coordination. Moreover, muscles with questionable reliability of activity detection were distinguished and highlighted in the presented summary figures. In the second part of the paper, our publicly available set of MATLAB files for surface EMG signal processing is described.

Optimalization of algorithms for automated detection of pathological high frequency oscillations in intracranial EEG records

  • Department: Department of Circuit Theory
  • Annotation:
    Evaluation of intracranial EEG recordings is very complex and complicated matter. In long-term monitoring is recorded large amounts of data and for the doctor is almost impossible to view in detail and evaluate the entire record. Finding high-frequency activity, which has a very small signal / noise ratio is then almost impossible. Therefore, it offers the possibility to create a method for automatic detection of this activity. For this purpose, has produced several algorithms for automatic detection of this phenomenon. In our group we also made a few of own algorithms. In this article we try to optimize individual algorithms and compare their improvement or sensitivity with respect to changes of their parameters.

Relationship between stuttering and therapy in patients with Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (PD) is neurodegenerative illness associated with loss of dopaminergic neurons in the substantia nigra.

The design for a long-term multi-electrode recordings of cortical activity in vivo

  • Authors: Vlk, P., Kudláček, J., Ing. Petr Ježdík, Ph.D., Otáhal, J., Demeterová, L., prof. Ing. Roman Čmejla, CSc., Jiruška, P.
  • Publication: IV. Letní doktorandské dny 2014. Praha: ČVUT FEL, Katedra teorie obvodů, 2014. pp. 49-54. ISBN 978-80-01-05506-9.
  • Year: 2014
  • Department: Department of Circuit Theory
  • Annotation:
    Understanding of the mechanisms responsible for the genesis of epilepsy and epileptic seizures requires obtaining comprehensive information on the spatial and temporal dynamics of epileptic activity of neuronal populations. Electrophysiological techniques of multichannel recordings represent some of the basic methods that allow a comprehensive study of neuronal dynamics in the epileptic brain.

Using of acoustic analyses for evaluation of voice and speech in Huntington disease

  • Department: Department of Circuit Theory
  • Annotation:
    Huntington's disease is a neurodegenerative disorder characterized by non-coordinative movements of the body.

Acoustic Findings of Voice Disorders in Huntington's Disease Compared to Parkinson's Disease

  • Department: Department of Circuit Theory
  • Annotation:
    One common finding in Huntington's disease (HD) is related to phonatory disruptions that can be perceptually characterized by harshness, strained strangled voice quality, and pitch fluctuations. These alterations of voice occur mainly as a consequence of underlying involuntary contractions, variable muscle tone, or even tremor of laryngeal musculature. Recently, several new acoustic analysis methods have been introduced to capture different aspects of these phonatory abnormalities. In this report, we summarize objective acoustic metrics suitable for assessment of phonatory dysfunction and provide their classification accuracy in separation between patients with HD and healthy controls. For this purpose, data consists of 272 phonations collected from 34 individuals with HD and 34 healthy controls. As impairment of phonatory function in HD was found across all investigated measurements, voice analysis may potentially serve as a marker of disease progression.

Automatic Detection and Spatial Clustering of Interictal Discharges in Invasive Recordings

  • Department: Department of Circuit Theory
  • Annotation:
    Interictal epileptiform discharges (spikes) represent electrographic marker of epileptogenic brain tissue. Besides ictal onsets, localization of interictal epileptiform discharges provides additional information to plan resective epilepsy surgery. The main goals of this study were: 1) to develop a reliable automatic algorithm to detect high and low amplitude interictal epileptiform discharges in intracranial EEG recordings and 2) to design a clustering method to extract spatial patterns of their propagation. For detection, we used a signal envelope modeling technique which adaptively identifies statistical parameters of signals containing spikes. Application of this technique to human intracranial EEG data demonstrated that it was superior to expert labeling and it was able to detect even small amplitude interictal epileptiform discharges. In the second task, detected spikes were clustered by principal component analysis according to their spatial distribution. Preliminary results showed that this unsupervised approach is able to identify distinct sources of interictal epileptiform discharges and has the potential to increase the yield of presurgical examination by improved delineation of the irritative zone.

Automatic Detection of High-frequency Oscillations In Invasive Recordings

  • DOI: 10.1109/MeMeA.2013.6549741
  • Link: https://doi.org/10.1109/MeMeA.2013.6549741
  • Department: Department of Circuit Theory
  • Annotation:
    High-frequency oscillations (HFOs) represent relatively new electrographic marker of epileptogenic tissue. It is starting to be used in presurgical examination to better plan surgical resection and to improve outcome of epilepsy surgery. Development of new techniques of unsupervised HFOs detection is required to further investigate the role of HFO in the pathophysiology of epilepsy and to increase the yield of presurgical examination. In this study we applied an envelope distribution modelling technique on experimental and human invasive data to detect HFOs. Application to experimental microelectrode recordings demonstrated satisfactory results with sensitivity 89.9% and false positive rate 2.1 per minute. Application of this algorithm to human invasive recordings achieved sensitivity 80%. High numbers of false positive detections required utilization of post-processing steps to eliminate the majority of them. This study shows that envelope distribution modelling represents a promising approach to detect HFOs in intracranial recordings. Advantages of this approach are quick adjustments to changes in background activity and resistance to signal non-stationarities. However, successful application to clinical practice requires development of secondary processing steps that will decrease the rate of false positive detections.

Bayesian changepoint detection for the automatic assessment of fluency and articulatory disorders

  • DOI: 10.1016/j.specom.2012.08.003
  • Link: https://doi.org/10.1016/j.specom.2012.08.003
  • Department: Department of Circuit Theory
  • Annotation:
    The accurate changepoint detection of different signal segments is a frequent challenge in various worldwide application domains. With regard to speech utterances, the changepoints are related to significant spectral changes, mostly represented by the borders between two phonemes. The main aim of this study is to design a novel Bayesian autoregressive changepoint detector (BACD) and test its feasibility in the evaluation of fluency and articulatory disorders. The originality of the proposed method consists in its normalising of a posteriori probability using Bayesian evidence and the designing of a recursive algorithm for reliable practice. For further evaluation of the BACD, we used data from (a) 118 people with various severity of stuttering to assess the extent of speech disfluency using a short reading passage, and (b) 24 patients with early Parkinson’s disease and 22 healthy speakers for evaluation of articulation accuracy using fast syllable repetition. Subsequently, we designed two measures for each type of disorder. While speech disfluency has been related to greater distances between spectral changes, dysarthric inaccurate articulation has instead been associated with lower spectral changes. These findings have been confirmed by statistically significant differences which were achieved in separating several degrees of disfluency and distinguishing healthy from parkinsonian speakers. In addition, a significant correlation between the automatic assessment of speech fluency and the judgment of human experts was obtained. In conclusion, our designed method provides a cost-effective, easy applicable and freely available evaluation of the speech disorders as well as other areas requiring reliable techniques for changepoint detection. In a more modest scope, BACD may be used in diagnosis of disease severity, monitoring treatment, and support for therapists’ evaluation.

Characteristics of utterances of patients with Parkinson's disease extracted from speech diadochokinetic task

  • Department: Department of Circuit Theory
  • Annotation:
    Although the Parkinson's disease is currently on of the most common neurodegenerative illnesses, there is only symptomatic treatment. Moreover, previous studies showed that diagnosis is managed even 10 years after appearance of the first symtoms.

Detection of Pathological High Frequency Activity in Intracranial EEG in Patients with Epilepsy using Hilber-Huang Transformation

  • Department: Department of Circuit Theory, Department of Telecommunications Engineering
  • Annotation:
    High-frequency oscillations are one of the symptoms of epileptic EEG and can take the advantage to accurate diagnosis and localization of epileptogenic tissue. For the processing of epileptic EEG are used several methods. Most of them are working on the basis of Fourier transform. This assumes that the signal is stationary and linear or linear only if the short-time Fourier transform is used. In this article we will focus on the use of not so well-known Hilbert-Huang transform, which is designed specifically for the analysis of nonlinear and non-stationary signals. We propose an algorithm for detecting high-frequency activity and verify whether this method is suitable for those purposes.

Evaluation of speech impairment in early stages of Parkinson's disease: a prospective study with the role of pharmacotherapy

  • DOI: 10.1007/s00702-012-0853-4
  • Link: https://doi.org/10.1007/s00702-012-0853-4
  • Department: Department of Circuit Theory
  • Annotation:
    Despite the initial reports showing beneficial effects of dopaminergic treatment on speech in Parkinson's disease (PD), more recent studies based upon valid measurements have not approved any improvement of speech performance under pharmacotherapy. The aim of this study was to analyze the effect of treatment initiation on the progression of speech impairment in PD, using novel evaluation criteria. Nineteen de novo patients with PD were tested and retested within 2 years after the introduction of antiparkinsonian therapy. As controls, 19 age-matched individuals were recorded. Speech examination included sustained phonation, fast syllable repetition, reading text, and monolog. Quantitative acoustic analyses of the key aspects of speech based on Gaussian kernel distribution, statistical decision-making theory, and healthy speech observation were used to assess the improvement or deterioration of speech. A trend for speech performances to improve was demonstrated after treatment mainly in quality of voice, intensity variability, pitch variability, and articulation. The treatment-related improvement differed in various aspects of speech for individual PD patients. Improvements in vowel articulation and pitch variability correlated with treatment-related changes in bradykinesia and rigidity, whereas voice quality and loudness variability improved independently. Using a novel approach of acoustic analysis and advanced statistics, improvements in speech performance can be demonstrated in PD patients after the introduction of antiparkinsonian therapy. Moreover, changes in speech articulation and pitch variability appear to be related with dopaminergic responsiveness of bradykinesia and rigidity. Therefore, speech may be a valuable marker of disease progression and treatment efficacy in PD.

Formant characteristics during contrastive stress production task in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Prosodic abnormalities have long been associated with hypokinetic dysarthria developed in the course of idiopathic Parkinson's disease (PD). Although reduced word stress (WS) is considered one of the common manifestations of dysprosody, few studies have investigated stress patterns based on objective acoustic analyses in detail.

Graphical user interface for data acquisition

  • Department: Department of Circuit Theory
  • Annotation:
    In the current time the closer interconnection of different fields of science leads to increased need of compatibility of used tools. For this reason a graphical user interface was implemented to make the data acquissition and assesment more efficient and more comfortable.

Imprecise vowel articulation as a potential early marker of Parkinson's disease: Effect of speaking task

  • DOI: 10.1121/1.4816541
  • Link: https://doi.org/10.1121/1.4816541
  • Department: Department of Circuit Theory
  • Annotation:
    The purpose of this study was to analyze vowel articulation across various speaking tasks in a group of 20 early Parkinson’s disease (PD) individuals prior to pharmacotherapy. Vowels were extracted from sustained phonation, sentence repetition, reading passage, and monologue. Acoustic analysis was based upon measures of the first (F1) and second (F2) formant of the vowels /a/, /i/, and /u/, vowel space area (VSA), F2i/F2u and vowel articulation index (VAI). Parkinsonian speakers manifested abnormalities in vowel articulation across F2u, VSA, F2i/F2u, and VAI in all speaking tasks except sustained phonation, compared to 15 age-matched healthy control participants. Findings suggest that sustained phonation is an inappropriate task to investigate vowel articulation in early PD. In contrast, monologue was the most sensitive in differentiating between controls and PD patients, with classification accuracy up to 80%. Measurements of vowel articulation were able to capture even minor abnormalities in speech of PD patients with no perceptible dysarthria. In conclusion, impaired vowel articulation may be considered as a possible early marker of PD. A certain type of speaking task can exert significant influence on vowel articulation. Specifically, complex tasks such as monologue are more likely to elicit articulatory deficits in parkinsonian speech, compared to other speaking tasks.

Measuring of cortical auditory evoked potentials in patients with hearing aids

MORPHOLOGICAL CLUSTERING OF INTERICTAL EPILEPTIFORM DISCHARGES IN INTRACRANIAL ELECTROENCEPHALOGRAPHY

  • Authors: Vlk, P., Ing. Radek Janča, Ph.D., prof. Ing. Roman Čmejla, CSc., Kršek, P., Marusič, P., Jiruška, P.
  • Publication: 21th Annual Conference Proceedings Technical Computing Prague 2013. Praha: Humusoft, 2013. ISSN 2336-1662. ISBN 978-80-7080-863-4.
  • Year: 2013
  • Department: Department of Circuit Theory
  • Annotation:
    Selected patients with refractory epilepsy can benefit from surgical treatment. The main purpose of presurgical examination is to identify and delineate epileptogenic areas of the brain which should be removed. Epileptogenic areas are determined according to the spatial distribution of seizure onsets, interictal epileptiform discharges or high-frequency oscillations. Specificity of interictal epileptiform discharges to mark epileptogenic tissue is decreased by the fact, that they are also observed outside the epileptogenic areas. To improve the localizing yield of interictal discharges, identification of specific features of the discharges generated only within epileptogenic region is required. The main aim of this project was to develop self-clustering algorithm which will discriminate distinct populations of interictal epileptiform discharges according to the morphology of their waveforms. First step of the developed algorithm extracts nine basic morphological features of each interictal epileptiform discharge detected in band-pass filtered (2-60 Hz) intracranial recordings. Principal component analysis is applied on extracted features to reduce their dimension. Only the first principle components with cumulative variance of 80 % or above are used for clustering. Gaussian Mixture Distribution method is utilized to assign each discharge to appropriate morphological cluster. Results of the clustering algorithm are displayed in the form of cortical maps together with medians of the clustered discharge waveform. Developed algorithm was tested in the model of intracranial EEG signal and on data recorded in patients who underwent intracranial monitoring. Results demonstrate the ability of the algorithm to separate interictal epileptiform discharges according to their morphological features.

Objective Acoustic Quantification of Phonatory Dysfunction in Huntington's Disease

  • DOI: 10.1371/journal.pone.0065881
  • Link: https://doi.org/10.1371/journal.pone.0065881
  • Department: Department of Circuit Theory
  • Annotation:
    Purpose: Although speech motor changes are reported as a common sign of Huntington’s disease (HD), the most prominent signs of voice dysfunction remain unknown. The aim of the current study was to explore specific changes in phonatory function in subjects with HD. Method: 34 subjects with HD and 34 age- and sex-matched healthy controls were examined. Participants performed sustained vowel phonation for subsequent analyses of airflow insufficiency, aperiodicity, irregular vibrations of vocal folds, signal perturbations, increased noise, and articulation deficiency. In total, 272 phonations were collected and 12 voice parameters were extracted. Subsequently, a predictive model was built to find the most salient patterns of voice disorders in HD. The results were also correlated with disease severity according to the Unified HD Rating Scale (UHDRS) motor score. Results: Subjects with HD showed deterioration in all investigated phonatory functions. Irregular pitch fluctuations, sudden phonation interruption, increased noise, and misplacement of articulators were found to be most significant patterns of phonatory dysfunction in HD (p < 0.001). The combination of these four dysphonia aspects contributed to the best classification performance of 94.1 % (sensitivity: 95.1 %; specificity: 93.2 %) in the separation of HD patients from healthy participants. Our results further indicated stronger associations between sudden phonation interruption and voluntary components of the UHDRS (r = - 0.48, p < 0.01) and between misplacement of articulators and involuntary components of the UHDRS (r = 0.52, p < 0.01). Conclusions: Our configuration of phonatory features can detect subtle voice abnormalities in subjects with HD. As impairment of phonatory function in HD was found to parallel increasing motor involvement, a qualitative description of voice dysfunction may be helpful to gain better insight into the pathophysiology of the vocal mechanism.

Segmentation of Surface EMG Signals

  • Authors: Sedlák, J., Špulák, D., prof. Ing. Roman Čmejla, CSc., Bačáková, R.B., Chrástková, Martina, Kračmar, B.K.
  • Publication: Applied Electronics - 2013 International Conference on Applied Electronics. Pilsen: University of West Bohemia, 2013. pp. 257-260. ISSN 1803-7232. ISBN 978-80-261-0166-6.
  • Year: 2013
  • Department: Department of Circuit Theory
  • Annotation:
    This paper compares two different approaches to electromyographic (EMG) segmentation for the purpose of muscle activation pattern identification. A widely known linear EMG envelope technique is compared with a newly designed method based on marker detection in a video. The results are evaluated by comparison of the muscle activity intervals. The experiments show that the video-based technique can achieve similar results to the EMG envelope. The EMG segmentation based on the video processing is more robust for segmentation of various types of EMG signal.

Using Hilbert-Huang Transformation in Epileptic EEG Analysis - Case study

  • Department: Department of Circuit Theory
  • Annotation:
    This article discusses the possibilities of using Hilbert-Huang transform in terms of processing EEG data of patients with epilepsy. This method was chosen because it accesses the signal as a non-linear and nonstationary process, which is the EEG recording. The paper focuses primarily on empirical modal decomposition that decomposes a signal into several copmonents with different frequency content. It discusses whether the expected content found graphoelements sought and whether it is possible to use individual components to determine their spread to neighboring channels.

Acoustic analyses of rhythm instability in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Dysprosody in Parkinson's disease (PD) can be characterized by changes in fundamental frequency, signal amplitude, and speech timing.

Acoustic analysis of emotions in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Acoustic analysis of emotional contrast in Parkinson's disease. Speech impairment in Parkinson's disease is traditionally characterized as hypokinetic dysarthria...

Acoustic Analysis of Stress Pattern in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Dysprosody in Parkinson's disease (PD) is characterized by several symptoms. One of these symptoms is disability to produce precise stress pattern. A group of 20 untreated patients in the early stages of PD was compared to a group of 17 healthy age-matched controls to quantify possible acoustic changes in speech production secondary to PD. All participants produced a standardized speaking task of contrastive stress. The standard used measurements of amplitude, fundamental frequency and duration were performed. Moreover, the novel measurement of Stress Pattern Index (SPI) that combines amplitude, fundamental frequency, and duration characteristics was designed. The measurements of F0 range, Intensity range, Duration, and SPI differentiate between PD and healthy controls (HC) across many conditions in the group of stressed and unstressed tokens. On the other hand, only SPI was sufficiently sensitive to distinguish between PD and HC group for stress tokens that was normalized by normal reading tokens. To summarize, acoustic changes of prosodic stress appear to occur from early stages of PD.

Acoustic analysis of stress patterns in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Dysprosody in Parkinson's disease (PD) is characterized by several symptoms. One of these symptoms is disability to produce precise stress pattern. A group of 20 untreated patients in the early stages of PD was compared to a group of 17 healthy age-matched controls to quantify possible acoustic changes in speech production secondary to PD. All participants produced a standardized speaking task of contrastive stress. The standard used measurements of amplitude, fundamental frequency and duration were performed. Moreover, the novel measurement of Stress Pattern Index (SPI) that combines amplitude, fundamental frequency, and duration characteristics was designed. The measurements of F0 range, Intensity range, Duration, and SPI differentiate between PD and healthy controls (HC) across many conditions in the group of stressed and unstressed tokens. On the other hand, only SPI was sufficiently sensitive to distinguish between PD and HC group for stress tokens that was normalized by normal reading tokens. To summarize, acoustic changes of prosodic stress appear to occur from early stages of PD.

Acoustic Analysis of Voice and Speech Impairment in Central Nervous System Diseases

  • Department: Department of Circuit Theory
  • Annotation:
    A number of central nervous system (CNS) ilnessess are manifested by alterations of voice and speech, frequently presented from early stages of disease.

Acoustical parameters for speech disfluency evaluation

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Sborník 85. akustického semináře. Praha: Nakladatelství ČVUT, 2012. pp. 25-30. ISBN 978-80-01-05133-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with an analysis of disfluent speech by means of automatic algorithms. Algorithms are based on the analysis of read audio recordings of stutterers. These parameters are for example regularity of speech energy, average length of silence or number of spectral changes in speech intervals. The results of comparison among control data and algorithms suggest that the parameters are able to estimate the level of speech fluency disorder.

Algorithms for detection of speech pathologies based on sustained phonation recordings

  • Department: Department of Circuit Theory
  • Annotation:
    Algorithms for detection of speech pathologies based on sustained phonation recordings.

Analysis of pathological voice and speech in laboratory of SAMI CTU

  • Department: Department of Circuit Theory
  • Annotation:
    The analysis and processing of speech signal is systemathically solved in Department of Circuit Theory, Faculty Of Electrical Engineering, CTU in Prague from beginning of 1990's.

Assessment of Hoarseness by Means of Analysis of Acoustic Signals

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. pp. 62-63. ISBN 978-80-01-05164-1.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Article deals with parametrisation of acoustic signal by mean of analysis of hoarseness. Several parameters were introduced and correlation analysis with subjective assessment was performed.

Audio signal synthesis

  • Authors: prof. Ing. Roman Čmejla, CSc.,
  • Publication: 20th Annual Conference Proceeding's Technical Computing Bratislava 2012. Praha: Humusoft, 2012. pp. 1-6. ISBN 978-80-970519-4-5.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with digital audio synthesis techniques taught in the subject Synthesis of multimedia signals at the Technical University in Prague. Theoretical concepts from the lectures are supplemented by practical exercises in programming tasks in MATLAB.

Automatic evaluation of articulation disorders based on fast syllables repetition in patients with Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    In the current time is the Parkinson's disease (PD) the second most common neurodegenerative disorder; however the PD becomes clinical evident after the mastifestation of some of four basic motoric symptoms.

Automatic search of significant location in parkinsonian utterances based on fast /pa/-/ta/-/ka/ syllables repetition

  • Department: Department of Circuit Theory
  • Annotation:
    This study is aimed on automatic assessment of pathological utterances in Parkisnon's disease (PD) using Matlab environment.

AUTOMATIC SEGMENTATION OF PHONEMES DURING THE FAST REPETITION OF (/PA/-/TA/-/KA/) SYLLABLES IN A SPEECH AFFECTED BY HYPOKINETIC DYSARTHRIA

  • Department: Department of Circuit Theory
  • Annotation:
    Hypokinetic dysarthria is a common manifestation of Parkinson's disease (PD). Articulation characteristics can provide useful information to distinguish dysarthric speakers from healthy subjects and monitor the severity of disease and treatment effects. The aim of this study was to design an algorithm for automatic segmentation of consonants and vowels based upon a rapid steady /pa/-/ta/-/ka/ syllable repetition.

Design and Construction of Pneumograph

  • Authors: Sedlák, J., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. pp. 59-61. ISBN 978-80-01-05164-1.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Automatic calculation a signals parameters are very often used methods in a biomedical engineering research. This paper describes algorithms in Matlab usable to calculate or detect specified parameters of pneumogram signals. The described algorithms are suitable to objective evaluation of pneumogram signals. The pneumogram signals used to testing proposed algorithms were acquired by equipment own production.

DETECTION OF HIGH FREQUENCY OSCILLATIONS IN CHRONICAL MODEL OF TEMPORAL EPILEPSY

  • Department: Department of Circuit Theory
  • Annotation:
    Vysokofrekvenční oscilace (HFO) hrají důležitou roli jak v normální funkci mozku, tak v patofyziologii neurologických onemocnění, především v epilepsii. Experimentálně a klinicky bylo prokázáno, že specifický typ vysokofrekvenčních oscilací je generován v oblastech mozku, ve kterých vznikají epileptické záchvaty. Tato vlastnost vysokofrekvenčních oscilací je v současnosti využívána v předoperační diagnostice za účelem zlepšení výsledků chirurgické léčby epilepsie. Předpokládá se, že vysokofrekvenční oscilace umožní přesnější určení oblasti mozku, kterou je nezbytné chirurgicky odstranit, aby došlo k vymezení záchvatů. Úspěšné použití vysokofrekvenčních oscilací v klinické praxi však vyžaduje vývoj nových algoritmů, které umožní spolehlivě detekovat tyto oscilace. V této práci bylo vyvinuto a testováno využití automatického detekčního algoritmu založeného na odlišnosti statistických parametrů základní aktivity pozadí a vysokofrekvenčních událostí. Tyto parametry byly určeny použitím Hilbertovy transformace, následované nalezením vhodného modelu logaritmicko-normální distribuce transformovaného signálu. Navržený postup umožňuje identifikovat vysokofrekvenční úseky signálů, neboť jejich distribuce jsou statisticky významně odchýlené od distribucí základní aktivity.

Developmental dysphasia evaluation

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Novinky ve foniatrii. Praha: Nakladatelství Galén, 2012. pp. 115-117. ISBN 978-80-7262-940-4.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Summary of objective evaluation of utterances from children with developmental dysphasia.

Estimation of fundamental frequency in pathological speech

  • Department: Department of Circuit Theory
  • Annotation:
    Automatic detection of speech pathology can help with early detection of speech pathology and with the analysis of the therapy progress. The basic speech parameter is the fundamental frequency (f0), which shows the frequency of vocal cord vibration.

Evaluation of hypokinetic utterances in patients with Parkinson's disease using voice onset time

  • Department: Department of Circuit Theory
  • Annotation:
    Evaluation of hypokinetic utterances in patients with Parkinson's disease using voice onset time.

Evaluation of Parkinsonian speech attributes obtained from utterances based upon the fast /pa/ - /ta/ - /ka/ syllables repetition

  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (PD) is the second most common neurodegenerative disorder after the Alzheimer's disease and with the aging of the worldwide population, the higher incidence is predicted. In 70 % - 90 % of PD cases, the hypokinetic dysarthria is manifested as a prodromal syndrom.

Evaluation of respiratory parameters from pneumogram

  • Authors: Sedlák, J.S., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 20th Annual Conference Proceeding's Technical Computing Bratislava 2012. Praha: Humusoft, 2012. pp. 1-3. ISBN 978-80-970519-4-5.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Automatic calculation a signals parameters are very often used methods in a biomedical engineering research. This paper describes algorithms in Matlab usable to calculate or detect specified parameters of pneumogram signals. The described algorithms are suitable to objective evaluation of pneumogram signals. The pneumogram signals used to testing proposed algorithms were acquired by equipment own production.

Evaluation of severity of developmental dysphasia

  • Authors: Nejepsová, M., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. pp. 54-56. ISBN 978-80-01-05164-1.
  • Year: 2012

Hoarseness assessment based on acoustical signal analysis

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Sborník 85. akustického semináře. Praha: Nakladatelství ČVUT, 2012. pp. 51-58. ISBN 978-80-01-05133-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Article deals with a description of several acoustic parameters suitable for objective assessment of hoarseness. A correlation analysis between these parameters and subjective assessment is also performed.

Hoarseness objectification - comparion of HNR estimated in frequency domain with Praat

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: LETNÍ DOKTORANDSKÉ DNY 2012. Praha: České vysoké učení technické v Praze, 2012. pp. 113-119. ISBN 978-80-01-05050-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Article deals with an estimation of HNR in frequency domain and relation of HNR to severity of hoarseness.

Localization of sources of epileptiform discharges during peroperation corticography

  • Department: Department of Circuit Theory
  • Annotation:
    The topic is focused on posibilities of detection and analysis of epileptiform discharges in peroperation corticogtraphy. Discharges clustering into groups allows definition of spike sources, which should cause seizure onsets. Localization of epileptiform bearings during surgery may be used like diagnostic tool for definition of removing area range.

Muscle Activity Detection Using EMG Envelope Thresholding - Comparison of Various Approaches

  • Authors: Špulák, D., prof. Ing. Roman Čmejla, CSc., Mikulíková, P., Bezoušková Paulů, J., Kračmar, B.
  • Publication: 20th Annual Conference Proceeding's Technical Computing Bratislava 2012. Praha: Humusoft, 2012. pp. 1-5. ISBN 978-80-970519-4-5.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    This paper is focused on various approaches to thresholding of electromyographic (EMG) envelope. We used multichannel surface electromyography to record and evaluate electrical activity of muscles during Nordic walking. Muscle activity detection was performed using thresholding of the EMG envelope. We compared thresholding related to the maximum of the corresponding movement cycle and thresholding related to the average of all maxima of several movement cycles. Our conclusion is that the results of both approaches are generally comparable when using the same threshold level. There are greater differences if the overall amplitude of the EMG envelope tends to change during the whole recording. In these cases, thresholding related to the maximum of the corresponding movement cycle should be preferred.

Objective evaluation of the severity of developmental dysphasia by children' speech

  • Authors: Nejepsová, M., prof. Ing. Roman Čmejla, CSc., Janda, J., Vokřál, J.
  • Publication: LETNÍ DOKTORANDSKÉ DNY 2012. Praha: České vysoké učení technické v Praze, 2012. pp. 102-107. ISBN 978-80-01-05050-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The aim of the research is finding acoustic parameters, which are suitable to determine the severity of developmental dysphasia in children speech. On the basis of subjective listening tests (conducted by experts and laics), patients were divided into three groups according to the severity of developmental dysphasia (low, medium and high severity). After voice and speech analyzes of utterances from healthy children and children with developmental dysphasia are obtained parameters suitable for the categorization into each group.

Parametrization of sustained phonation for assessment of hoarseness

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Novinky ve foniatrii. Praha: Nakladatelství Galén, 2012. pp. 126-128. ISBN 978-80-7262-940-4.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Article deals with a description of several acoustic parameters of sustained vowel /a/ which is suitable for assessment of hoarseness.

Processing of Multichannel Surface Electromyographic Signals

  • Authors: Špulák, D., Sedlák, J., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. ISBN 978-80-01-05164-1.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The contribution is focused on analysis of muscle activity pattern by using multichannel EMG signals. The muscle activity detection methods are desribed in detail. The algorithms are adapted to processing of EMG signals which are recorded during periodic movement.

Relation Between Stuttering Severity and Automatic Algorithms

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. pp. 52-53. ISBN 978-80-01-05164-1.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The paper deals with automatic algorithms which are able to describe the level of speech fluency disorder in read audio recordings of stutterers.

ROBUST EPILEPTIFORM SPIKE DETECTION IN INTRACRANIAL ELECTROENCEPHALOGRAPHY

  • Department: Department of Circuit Theory
  • Annotation:
    The spike-wave complex as marker of epileptical activity is occurred in parts of the brain that may trigger epileptic seizure. Localization of this area is necessary for precisely neuro-surgical treatment. The aim of the project is developing of automatic objective robust spike detector.

Techniques of digital audio synthesis in MATLAB

  • Authors: prof. Ing. Roman Čmejla, CSc.,
  • Publication: Sborník 85. akustického semináře. Praha: Nakladatelství ČVUT, 2012. pp. 11-16. ISBN 978-80-01-05133-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with techniques of digital audio synthesis, which are discussed in the course Synthesis of multimedia signals. This course is designed for students of the 5th semester of undergraduate studies at the Czech Technical University in Prague and the course introduces the fundamentals of sound synthesis algorithms (everyday, music and speech), digital audio effects and sonification. Understanding of theoretical concepts will be consolidated through practical programming assignments in MATLAB.

The characteristics in speech from children with developmental dysphasia

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Sborník 85. akustického semináře. Praha: Nakladatelství ČVUT, 2012. pp. 31-34. ISBN 978-80-01-05133-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    This paper describes a part of research which focus on acoustic parameters in speech from children with developmental dysphasia.

The Comparative Analysis of Free Technique XC-skiing and Skating on Roller Skies

  • Authors: Chrástková, M., Bačáková, R., Špulák, D., Kračmar, B., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Scientia Movens - Sborník příspěvků z mezinárodní studentské vědecké konference konané dne 27. března 2012. Praha: Univerzita Karlova, Fakulta tělesné výchovy a sportu, 2012, pp. 28-36. ISBN 978-80-86317-92-2.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Moving on ski is very old human locomotion on snow. In last century it became very favourite winter and olympic sport, which enjoys from big boom among public in this time. New cross country skiing technique – skating has developed very dynamically from 80th years 20th century. Presented study deals with comparison between skating on ski and roller ski. We studied tree base styles of skating: V-2 on right and left side and V-1. It was measured the beginnings of important activation each of muscles on right leg and their subsequent deactivation during one step´s cycle. The results confirm the idea that we can consider roller skiing as special training device from cross country skiing. It is right from kinetic view and from connection of muscles too.

The evaluation of stuttering

  • Authors: Lustyk, T., prof. Ing. Roman Čmejla, CSc., Bergl, P.
  • Publication: LETNÍ DOKTORANDSKÉ DNY 2012. Praha: České vysoké učení technické v Praze, 2012. pp. 108-112. ISBN 978-80-01-05050-7.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The paper is focused on evaluation of stuttering by means of automatic methods. Experiments are carried out for read audio recordings. The parameters process a recording as a whole. Algorithms' outputs are compared to control data based on the LBDL(the Lidcomne Behavioral Data Language of Stuttering) taxonomy.

THE REAL-TIME VIZUALIZATION OF PNEUMOGRAM SIGNALS

  • Department: Department of Circuit Theory
  • Annotation:
    The real-time visualization of pneumogram signals is described in this contribution. The device created provides simultaneous measurement of pneumogram and voice. The sampled signals are transmitted to PC software in real time via USB bus, which makes it possible to show the curves from the sampled signals in real-time. In addition, data archiving and printing functions are integrated into the software. Since pneumography is one of the oldest methods to provide information about the technique of respiration, it will be possible for doctors in phoniatric practice to use the created device and programmed application for patient diagnosis. Listed in the results is an example of the recordings acquired from the designed device.

THE RELATION BETWEEN SPECTRAL CHANGES DISTANCE AND PROLONGATION IN SPEECH OF STUTTERERS

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 20th Annual Conference Proceeding's Technical Computing Bratislava 2012. Praha: Humusoft, 2012. pp. 1-5. ISBN 978-80-970519-4-5.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The paper deals with an algorithm that could be used to identify prolongations in audio recordings of disfluent speech. The measurement utilizes the detector of abrupt spectral changes and the voice activity detector. The experiment is based on the analysis of read audio recordings of stutterers. The highest correlation coefficient with control data is 0.72. This algorithm could become a base of a parameter in an automatic and objective assessment system.

The severity rating of developmental dysphasia by utterances from 5-7 years old patients

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: 2012 International Conference on Applied Electronics. Pilsen: University of West Bohemia, 2012. pp. 191-194. ISSN 1803-7232. ISBN 978-80-261-0038-6.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The main topic of this article is to compare acoustic parameters of utterances from children with developmental dysphasia and healthy children in age 5-7 years. The goals of this article are establishing the voice and speech characteristics that show differences between children with developmental dysphasia on three levels of this disorder and healthy children. These parameters should be helpful in the future to phoniatrists and clinical specialists to assess the current level of development of speech, the articulation and the severity assessment of speech disorder of these patients.

Time behaviour parametrization for hoarseness assessment

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 20th Annual Conference Proceeding's Technical Computing Bratislava 2012. Praha: Humusoft, 2012. ISBN 978-80-970519-4-5.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Article deals with an additional description of time behaviours of parametrised signal: LRE (Line Regression Error) a CL (Curve Length).

Use of pneumography and pneumogram record for phoniatric examination

  • Authors: Sedlák, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Novinky ve foniatrii. Praha: Nakladatelství Galén, 2012. pp. 124-125. ISBN 978-80-7262-940-4.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Pneumography is one of the oldest methods to provide information about respiration during speaking and singing (there is no equipment manufacturer in the present market). This Article describes design and construction of pneumograph for phoniatric examination. The Article provides a short description of hardware and software design, further, results of measurement obtained by the designed device are presented.

Using of Bayesian changepoint detectors in analysis of speech pathology

  • Authors: prof. Ing. Roman Čmejla, CSc.,
  • Publication: Czech-German Workshop on Speech Pathology and Biological Signals - Proceedings. Praha: ČVUT FEL, Katedra teorie obvodů, 2012. pp. 32-34. ISBN 978-80-01-05164-1.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    The aim of this work has been to suggest and verify a novel method for speech pathology analysis based on recursive Bayesian changepoint detector.

Using of the LBDL taxonomy as a control data for automatic algorithms

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Novinky ve foniatrii. Praha: Nakladatelství Galén, 2012. pp. 112-114. ISBN 978-80-7262-940-4.
  • Year: 2012
  • Department: Department of Circuit Theory
  • Annotation:
    Using of the LBDL taxonomy as a control data for automatic algorithms which evaluate the level of speech disfluency on the basis of stutterrers' audio recordings analysis

Vowel articulation in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Dysarthria in Parkinson's disease is multidimensional speech impairment affecting various aspects such as respiration, phonation, articulation, and prosody.

Acoustic analysis of voice and speech characteristics in early untreated Parkinson's disease

  • Authors: doc. Ing. Jan Rusz, Ph.D., prof. Ing. Roman Čmejla, CSc., Růžičková, H., Klempíř, J., Majerová, V., Picmausová, J., Roth, J., Růžička, E.
  • Publication: Proceedings of the 7th International Workshop on Models and Analysis of Vocal Emissions for Biomedical Applications. Florencie: Universita di Firenze, 2011. p. 181-184. 612.78 (ed. 20). ISBN 978-88-6655-009-9.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (pd) is a neurological illness characterized by progressive lost of dopaminergic neurons, primarily in the substantia nigra pars compacta. changes in speech associated with hypokinetic dysarthria are a common manifestation in patients with idiopathic PD. The aim of this study is to investigate the feasibility of automated acoustic measures for the identification of voice and speech disorders in PD.

Acoustic Assessment of Voice and Speech Disorders in Parkinson's Disease Through Quick Vocal Test

  • DOI: 10.1002/mds.23680
  • Link: https://doi.org/10.1002/mds.23680
  • Department: Department of Circuit Theory
  • Annotation:
    The disorders of voice and speech in Parkinson's disease (PD) result from involvements in several subsystems including respiration, phonation, articulation, and prosody. We investigated the feasibility of acoustic measures for the identification of voice and speech disorders in PD, using a quick vocal test consisting of sustained phonation, diadochokinetic task, and running speech. Various traditional and novel acoustic measurements have been designed in order to be gender independent, represent all speech subsystems, reduce the time required for voice investigation, and provide a reliable automated assessment in practice.

Acoustic markers of speech degradation in early untreated Parkinson's disease

  • Authors: doc. Ing. Jan Rusz, Ph.D., prof. Ing. Roman Čmejla, CSc., Růžičková, H., Klempíř, J., Majerová, V., Picmausová, J., Roth, J., Růžička, E.
  • Publication: Proceedings of Forum Acusticum 2011. Madrid: European Acoustics Association, 2011. p. 2725-2730. Danish Acoustical Society. ISSN 2221-3767. ISBN 978-84-694-1520-7.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (PD) is a neurological illness associated with a variety of motor deficits and non-motor deficits involving areas such as speech, mood, behaviour, thinking, and sensation. The PD-related vocal impairment results from the involvement of various speech subsystems including respiration, phonation, articulation, and prosody. The abnormalities in these speech subsystems are traditionally assessed using several acoustic measurements. Signal processing algorithms, in turn, offer an objective method for precise evaluation of speech performance from recorded signals. In this paper, we study the disordered speech of people with early PD who have not undergone pharmacotherapy treatment.

An Analysis of Iterative Algorithm for Estimation of Harmonics-To-Noise Ratio in Speech

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 19th Annual Conference Proceedings Technical Computing Prague 2011. Praha: Vydavatelství VŠCHT Praha, 2011. pp. 1-7. ISBN 978-80-7080-794-1.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    This paper delas with analysis of iterative algorithm which will be used to determine the HNR in speech. Describes the weknesses and brings a new approach to solve them.

Analysis of intracranial EEG recording

  • Department: Department of Circuit Theory
  • Annotation:
    The aim of the project is to create suitable algorithms for detecting the presence, localization and extent of epileptogenic focus. The methods have been developed for providing an objective view of the diagnostic symptoms, and are based on search and evaluation of relations in the intracranial EEG signals recordings During the course of the project in the past year, we have focused mainly on the methods for detection of interictal spikes, methods using one-channel time-frequency analysis through Bayesian change-point detectors, methods using multi-channel detection based on multidimensional autoregressive models, and methods based on correlation analysis.

Analysis of Signals Induced by Mechanical Activity of Heart

  • Department: Department of Physics, Department of Circuit Theory
  • Annotation:
    Our research is focused on two areas, namely on the heart rate variability (HRV) and on noninvasive continuous blood pressure estimation. In this paper, we describe relationships between HRV and both spontaneous and controlled respiration. In order to improve methods for continuous blood pressure estimation, the influence of averaging on correlation between pulse arrival time (PAT) and blood pressure is also studied.

Assessment of Vocal Parameters in Pathological Voices

  • Department: Department of Circuit Theory
  • Annotation:
    In this paper we present usage of the two novel speech fundamental frequency (f0) detection methods for sustained phonations to assess other vocal parameter such as jitter, shimmer, and harmonics-to-noise ratio (HNR). These voice parameters are commonly used to detect the speaker's voice characteristics and his voice quality. The aim of the work is to design the algorithm for detecting these vocal parameters, regarding to the robustness and automatization with minimisation of external parameters set-up. The results of particular parameters are compared with commonly used software Praat. The comparison of results between Praat and our designed vocal parameters were 83% for jitter, 88% for shimmer and 94% for HNR.

Assessment of voice and speech impairment

  • Department: Department of Circuit Theory
  • Annotation:
    The main aim of this project is the evaluation of voice and speech disorders in the field of several areas including dysarthria of Parkinson's disease, children dysphasia, stuttering, and speech dysfluency. The project deals with creation of speech databases, labelling several types of utterances, acoustic analysis, design and development of signal processing algorithms for automatic evaluation of speech features, design of appropriate classifiers, but also collaboration with speech-language pathologists in assessment of quality of speech and with neurologist in assessment of patients' global motor impairment. The research in the area of speech pathology has a crucial role to play in both theoretical and practical issues of developing and improving speech treatment across a range of diseases and conditions. The theoretical results yield essential knowledge for multidisciplinary science development in areas of acoustic, neurology, and linguistics.

Automatic segmentation of phonemes during fast (/pa/ /ta/ /ka/) syllables repetition in hypokinetic dysarthria

  • Department: Department of Circuit Theory
  • Annotation:
    Hypokinetic dysarthria is a common manifestation of Parkinson's disease (PD). Articulation characteristics can provide useful information to distinguish dysarthric speakers from healthy subjects and monitor the severity of disease and treatment effects.

Change evaluation of Bayesian detector for dysfluent speech assessment

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Applied Electronics 2011. Plzeň: University of West Bohemia, 2011, pp. 231-234. ISSN 1803-7232. ISBN 978-80-7043-987-6.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    This paper is focused on a parameter which could be used for automatic and objective assessment of dysfluent speech. The measurement is based on the analysis of read audio recordings of stutterers. The designed parameter analyzes the signal in the spectral domain, which correlates with the subjective assessment of two phoniatrics experts with a coefficient of -0.789. A Bayesian change-point detector (BACD) has been used for spectral analysis. This parameter could become a part of an automatic and objective assessment system.

Complex measurement of lung functions using spirometry

  • Department: Department of Circuit Theory
  • Annotation:
    Mezi základní parametry neinvazivních měření plicních funkcí patří plicní ventilace (kapacita a objem plic), maximální výdechový objem (usilovný výdech) a maximální volní ventilace (hluboké dýchání). Motivací této práce je laboratorní úloha určená pro výuku studentů biomedicínského inženýrství ČVUT FEL v předmětu Biologické signály. Pro všechna měření byl využit spirometr z profesionálního měřícího zařízení Biopac Student Lab a následně byly vyvinuty vlastní jednoduché algoritmy pro automatickou extrakci parametrů ze signálů ve vývojovém prostředí Matlab.

Objectification of dysarthria in Parkinson's disease using Bayes theorem

  • Authors: doc. Ing. Jan Rusz, Ph.D., prof. Ing. Roman Čmejla, CSc., Růžičková, H., Růžička, E.
  • Publication: Proceedings of the 10th WSEAS International Conference on SIGNAL PROCESSING, ROBOTICS and AUTOMATION. Athens: World Scientific and Engineering Academy and Society, 2011. pp. 165-169. Zoran Bojkovic, Janusz Kacprzyk, nikos Mastorakis, Valeri Mladenov, Roberto Revetria, Lotfi A. Zadeh, Alexandr Zemliak. ISSN 1792-8192. ISBN 978-960-474-276-9.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents an assessment of vocal impairment for separating healthy persons from patients with Parkinson's disease (PD). We have recently shown that deterioration of speech performances in PD speakers is notable from an early stage of the disease, even before starting pharmacotherapy. In this study, we present the potential of the simple Bayes rule to reveal changes in degradable speech performance in the course of PD-related dysarthria. The various speech data were recorded from 23 speakers with recently diagnosed PD and 23 healthy speakers. It has been found that 19 various acoustic measurements are able to differentiate PD significantly from healthy speakers. Subsequently, the Bayes theorem was applied to each of these measurements. As a result, the 21 PD patients and 21 healthy people were correctly classified according to their group. The Bayes theorem thus confirms its feasibility for identifying the features of the impaired voice.

Parameters for Mean Blood Pressure Estimation Based on Electrocardiography and Photoplethysmography

  • Department: Department of Physics, Department of Circuit Theory
  • Annotation:
    The focus of this paper is continuous non-invasive blood pressure monitoring. We used electrocardiography (ECG) and photoplethysmography (PPG) to compute pulse arrival time (PAT) between heart and finger. PAT was measured as the time difference between the R-peak of ECG and a characteristic point of PPG. We defined different characteristic points on PPG and compared their suitability for blood pressure estimation. Measurements were made on a group of 10 volunteers using an invasive blood pressure meter as reference. Generally, the best results were achieved when using time intervals from R-peak of ECG to the relative minimum preceding the rising edge of PPG. However, for some participants, differing parameters gave better results.

Parametrization of the Intracranial Electroencephalography via Wavelet Transformation

  • Department: Department of Circuit Theory
  • Annotation:
    This paper describes a methodology for finding appropriate wavelets to parameterizing of intracranial electroencephalography (iEEG) records. Suitable wavelets were calculated according to mutual energy in defined frequency ranges that correspond with standard distributions of EEG bands. The appropriate wavelets were selecting from 84 types of wavelets. The wavelets will be used to parameterizing iEEG records for classifier that will be localizing the seizure onset zone (SOZ) in the epileptology.

Quantitative acoustic measurements for characterization of speech and voice disorders in early untreated Parkinson's disease

  • DOI: 10.1121/1.3514381
  • Link: https://doi.org/10.1121/1.3514381
  • Department: Department of Circuit Theory
  • Annotation:
    An assessment of vocal impairment is presented for separating healthy people from persons with early untreated Parkinson's disease (PD). This study's main purpose was to (a) determine whether voice and speech disorder are present from early stages of PD before starting dopaminergic pharmacotherapy, (b) ascertain the specific characteristics of the PD-related vocal impairment, (c) identify PD-related acoustic signatures for the major part of traditional clinically used measurement methods with respect to their automatic assessment, and (d) design new automatic measurement methods of articulation.On the basis of objective acoustic measures, statistical decision-making theory, and validation from practicing speech therapists, it has been demonstrated that 78% of early untreated PD subjects indicate some form of vocal impairment. The speech defects thus uncovered differ individually in various characteristics including phonation, articulation, and prosody.

Robust Peak Detection Algorithms for Fundamental Frequency Estimation of Sustained Vowel Phonation in Pathological Voices

  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents design of two new methods of speech fundamental frequency (f0) detection for vowel sustained phonations and the detection method, which use cross-corelation to detect f0, is tested. The algorithm consists of certain preprocessing and processing methods. The first method is based on the detection of maxima and the second method is based on band pass filtration. In comparison with the other commonly used f0 detection methods, our algorithms are designed with respect to speech pathology detection. These methods lead to detection of the other voice parameters such as jitter, shimmer and harmonic-to-noise ratio (HNR). The results of this study are compared with database, which is labeled by the help of Praat algorithm. The results for maximum method succeed at 88.4% and for pass band method at 83.9%. The detection leads to create self-automated method, which detect robustly f0.

Robust pitch detection algorithms for estimation of fundamental frequency of prolonged vowels phonations at pathological voices

  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents a design for two novel methods of speech fundamental frequency (f0) detection in pathological voices. The methods are tested on the database of vowelsustained phonation. One method is based on the detection of maxima and the other on signal filtration with band pass. Compared to other well known f0 detection methods, our algorithms are designed with respect to speech pathology detection. Moreover, the designed methods support detection of the other voice parameters, such as jitter, shimmer and harmonic-to-noise ratio (HNR). The comparison of our results with results generated by the help of the Praat algorithm reaches a high value of correspondence: for the maximum method, the accuracy is 88,4 % and for the pass band method it is 83,9 %.

Rules for Spike Detection in Multichanel Intracranial Electroencephalography

  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents a possibility for quantitative evaluation of the electroencephalography (EEG) through searching for discharge artifacts in the intracranial electrocorticography signals of epileptic patients. Occurrence of the discharges as a spike-wave complex directly relates to affected parts of the brain. The described method uses the spike detector based on comparison of the energy envelope of the filtered signals and the individual settings of thresholds. The settings of the thresholds follow the rule which is dependent on statistical distribution and its MLE approximation of the envelope. Two patient results are introduced as a preview. The proposed algorithm is designed for automatic evaluation of intracranial EEG signals to increase objectivity and save time for neurologist examinations.

Seizure Onset Zone Detection and Localization in iEEG Using DTF

  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents the possibility of early detection and localization of epileptogenic focus in the iEEG (intracranial Electroencephalography) signal using a method based on multidimensional autoregressive models. The work provides the first results of the method in the iEEG signal, and discusses technical aspects in terms of the suitability of the sampling frequency, AR model order and segmentation step.

Spectral Approximations for Sibilant Classification

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 2011 International Conference on Applied Electronics. Plzeň: Západočeská univerzita v Plzni, 2011. pp. 385-388. ISSN 1803-7232. ISBN 978-80-7043-987-6.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    The paper deals with analysis of spectrum of czech sibilants /s/, /sh/, /z/ and /zh/ and tries to find optimal approximation of power spectrum which can be used for sibilant classification. Several polynomial approximations and a Bézier`s approximation in several frequency bands were used. The best classification results are given by the Bézier`s approximation in frequency band 0.3-10 kHz (97,4 %).

Sturge-Weber syndrome: a favourable surgical outcome in a case with contralateral seizure onset and myoclonic-astatic seizures

  • Authors: Jiruska, P., Marusic, P., Jefferys, JGR, Krsek, P., prof. Ing. Roman Čmejla, CSc., Sebronova, V., Komarek, V.
  • Publication: Epileptic Disorders. 2011, 13(1), 76-81. ISSN 1294-9361.
  • Year: 2011
  • DOI: 10.1684/epd.2011.0407
  • Link: https://doi.org/10.1684/epd.2011.0407
  • Department: Department of Circuit Theory
  • Annotation:
    Sturge-Weber syndrome is a neurocutaneous disorder classically characterized by the presence of facial port-wine stain and ipsilateral leptomeningeal angiomatosis. It is often associated with refractory epilepsy which requires surgical treatment. We present a case of a patient who initially presented with partial seizures of temporo-occipital origin, ipsilateral to the pial angiomatosis. During the course of the disease, the patient developed medically refractory epilepsy with partial seizures originating predominantly from the contralateral temporo-occipital area as well as myoclonic and myoclonicastatic seizures. Resection of the occipital and temporal lobe affected by the pial angioma resulted in favourable outcome.

The Number of Spectral Changes in Speech Segments for Evaluation of Dysfluent Speech

  • Authors: Lustyk, T., Bergl, P., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: 19th Annual Conference Proceedings Technical Computing Prague 2011. Praha: Vydavatelství VŠCHT Praha, 2011. pp. 1-6. ISBN 978-80-7080-794-1.
  • Year: 2011
  • Department: Department of Circuit Theory
  • Annotation:
    This paper is focused on a measurement that could be useful for an automatic and objective evaluation of dysfluent speech. The parameter combines the time and frequency domain analysis. The experiment is based on the analysis of read audio recordings of stutterers. The highest correlation coefficient with control data is -0.821. This parameter could become a part of an automatic and objective assessment system.

Acoustic Analysis of Utterances from Children with Developmental Dysphasia

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Akustické listy. 2010, 16(4), 4-8. ISSN 1212-4702.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    The main goal of this article is to find parameters for the severity assessment of children with developmental dysphasia. This parameters will be used for an automatic classifier. Developed algorithm in the form of software applications should be helpful in the future to phoniatrist and clinical specialists to assess the current level of development of speech and articulation. Spectral moments of isolated sibilants and their length, rate and rhytm of speech in specific utterances are only some of suitable parameters for diagnosis of developmental dysphasia.

Acoustic Analysis of Utterances: Children With Developmental Dysphasia

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Škodová, E.Š.
  • Publication: Technical Computing Bratislava 2010. Bratislava: RT systems, s.r.o, 2010. pp. 1-3. 18th Annual Conference Proceedings. ISBN 978-80-970519-0-7.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    This paper describes recording utterances of children with developmental dysphasia and subsequent analysis of captured records. Suitable parameters from this acoustic analysis should be used for automatic classification of utterances. These parameters include for examples DTW isolated words, the spectral moments of isolated utterances sibilants and rate of specific utterances.

Analysis of Fricative Consonant /s/ in Dysarthria Test

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 20th Czech - German Workshop on Speech Processing. Prague: Institute of Photonics and Electronics AS CR, 2010. pp. 58-69. ISBN 978-80-86269-21-4.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    During dysarthria test an assessment of fricative consonant /s/ is commonly performed. This paper introduces the possibilities of automated evaluation of two of the most common tasks: "Ability to sustain /s/ on exhalation" and "Ability to repeat series of /s/". Pitfalls of recording /s/ on commonly used PCs, design of algorithm for classifying /s/ and parameters which can be used to classify speakers with Parkinson disease and healthy speakers based on assessment of /s/ are also discussed. In the end the design of an application useful for real-time assessment of both of the above-mentioned tasks is mentioned.

Analysis of Heart Rate Variability During Respiration

  • Authors: Demchenko, V., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Technical Computing Bratislava 2010. Bratislava: RT systems, s.r.o, 2010. 18th Annual Conference Proceedings. ISBN 978-80-970519-0-7.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with a problem of signal processing of heart rate variability (HRV) with emphasis on a detection of similarity between the HRV and respiration signal. Different heart rate variability tests were used to reach this goal. We also have performed tests with spontaneous respiration signals in order to compare our results. Linear (mean, standard deviation etc.) and nonlinear methods (Poincaré plot), that work in both time and frequency domain, were used. A segmentation of the breathing signal took place within the time domain method. It gives a basic notion of normal and controlled breathing pattern and the information about a time delay of the HRV. Influence of naturalness of breathing pattern was also detected. The known fact that the HRV stay similar to previous pattern during first 20seconds of apnea phase (special respiration pattern with no breathing) was confirmed. A result of this research is a confirmation of the previous results and a basic detection of the time...

Analysis of utterance from children with developmental dysphasia

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Škodová, Š.
  • Publication: 8. ČESKO-SLOVENSKÝ FONIATRICKÝ KONGRES. Bratislava: Samedi s.r.o., 2010. pp. 25. ISSN 1337-2181.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    This article deals with the acoustic analysis of children's speech. Describes a methodology for finding and comparing the differences recorded speeches healthy children and children with developmental dysphasia. Analysis of these utterances are obtained appropriate parameters for future automatic evaluation of children's speech defects.

Assessment of Dysfluency in Stuttered Speech

  • Authors: Bergl, P., Lustyk, T., prof. Ing. Roman Čmejla, CSc., Černý, L., Hrbková, M.
  • Publication: Technical Computing Bratislava 2010. Bratislava: RT systems, s.r.o, 2010. pp. 1-3. 18th Annual Conference Proceedings. ISBN 978-80-970519-0-7.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    The project aims to find methods that would be based on the analysis of audio recordings stutterers and were able to objectively and automatically determine the degree of speech fluency disorders.

Change detection in intracranial EEG signals

  • Department: Department of Circuit Theory
  • Annotation:
    The aim of our research is to create algorithms suitable for detecting the presence, localization and determine the extent of epileptogenic focus. We developed experimentally methods for providing view of the diagnostic symptoms. The methods are based on search and evaluation of relations between the monitoring intracranial EEG signals. This paper shows our first approach to the problematic and also our preliminary results.

Description of the Spectral Properties of Czech Sibilants Using the Line Spectral Pairs Parametrisation

  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with quantitative description of a power spectrum of Czech sibilants /s/, /š/, /z/, /ž/ using Line Spectral Pairs (LSP) parametrisation. A brief introduction to LSP and several experiments with signal description by LSP and signal periodicity was performed. According to the parametrisation, two classifiers were developed - the first classifier that does not use signal periodicity description is able to classify 95.2% of microsegments from the input database correctly, the second classifier using signal periodicity description increases the amount of correctly classified microsegments up to 96.8%.

Distributed computing system for EEG signal processing using MATLAB as activex object in DCOM model

  • Department: Department of Circuit Theory
  • Annotation:
    Automated signal processing of electroencephalographic (EEG) data is mostly very sophisticated and time consuming. Also the data amounts of the patients necessary to process are mostly high. Using a parallel signal processing techniques is suitable for saving the time. When the workgroup computing force consist only standard personal computers with standard equipment, a technique easy to implement is necessary. The one suitable technique is the Distributed Component Object Model (DCOM) mechanism that basically allows ActiveX clients to use remote ActiveX objects over a network. This paper describes one possibility of using a number of Matlab instances as ActiveX objects in school network working in synergy when the consequent is decreasing the computing time of EEG signal processing.

Epileptic high-frequency network activity in a model of non-lesional temporal lobe epilepsy

  • Authors: Jiruska, P., Finnerty, G. T., Powell, A. D., Lofti, N., prof. Ing. Roman Čmejla, CSc., Jefferys, G. R.
  • Publication: Brain. 2010, 133(5), 1380-1390. ISSN 0006-8950.
  • Year: 2010
  • DOI: 10.1093/brain/awq070
  • Link: https://doi.org/10.1093/brain/awq070
  • Department: Department of Circuit Theory
  • Annotation:
    High-frequency cortical activity, particularly in the 250-600 Hz (fast ripple) band, has been implicated in playing a crucial role in epileptogenesis and seizure generation. Fast ripples are highly specific for the seizure initiation zone. However, evidence for the association of fast ripples with epileptic foci depends on animal models and human cases with substantial lesions in the form of hippocampal sclerosis, which suggests that neuronal loss may be required for fast ripples. In the present work, we tested whether cell loss is a necessary prerequisite for the generation of fast ripples, using a non-lesional model of temporal lobe epilepsy that lacks hippocampal sclerosis.

Evaluation of speech dysfuency

  • Authors: Bergl, P., Lustyk, T., prof. Ing. Roman Čmejla, CSc., Černý, L., Hrbková, M.
  • Publication: 8. ČESKO-SLOVENSKÝ FONIATRICKÝ KONGRES. Bratislava: Samedi s.r.o., 2010. pp. 25. ISSN 1337-2181.
  • Year: 2010

Evaluation of stress, emotions, rhythm, articulation rate and regularity in Parkinsons disease

  • Department: Department of Circuit Theory
  • Annotation:
    The purpose of this study is to assess the traditional measure of articulation rate and regularity performed on rapid steady syllable repetition known as diadochokinetic (DDK) task, and to present the new measures of prosody performed on non-standard vocal tasks, include stress, emotions and rhythm. We collected various speech data from 46 Czech native speakers, 23 with Parkinson's disease (PD). We introduce automatical algorithms for all these measures using MATLAB environment. The voice of PD group demonstrated significant differences in measures of DDK rate, pitch and intensity variation in stress, pitch and F2 variations in emotions, and in rhythm.

Experiments with Blood Pressure Monitoring Using ECG and PPG

  • Department: Department of Physics, Department of Circuit Theory
  • Annotation:
    We observed dependencies among blood pressure, pulse arrival time and pulse wave amplitude of photoplethysmogram in our studies. Measurements were made at the intensive care unit in Motol University Hospital in Prague and were compared with invasive blood pressure meter. The correlation coefficients between systolic or diastolic pressure and pulse arrival time achieved up to - 0.97, when averaging over 16 RR intervals. However, the range of the measured systolic and diastolic pressure is very low in some cases, and therefore we cannot do reliable conclusions. Correlations between pressure and amplitude of photoplethysmogram were still somewhat higher, however, a poor repeatability of such a measurement can be expected.

Information system for management and analysis of medical data

  • Department: Department of Circuit Theory
  • Annotation:
    This paper presents the current state of information system for management and analysis of medical data. The main aim of this project was to develop a web based system that could be used to store and easily manage large amount of data and that could provide access to this data for multiple users at the same time. The final system consists of ASP.NET application that forms the user interface and provides connection to SQL and file server. The system provides extensive user management and user access control, to suite variable needs of both small and large projects. It also contains a sophisticated search engine that makes management of large quantities of information easier. Finally, there is a basic analytical tool with potential for further improvement.

Laboratory tasks from voice analysis in the study of biomedical engineering using matlab

  • Department: Department of Circuit Theory
  • Annotation:
    Students of Biomedical Engineering of the Czech Technical University in Prague meet with the laboratory task "Voice Analysis" in the subject Biological Signals. They have the opportunity here to practically learn basics about fundamental glottal tone, formant frequencies, their parameters and detection, frequently used in voice pathology analysis.

Possibilities of Automated Assessment of /s/

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 8. ČESKO-SLOVENSKÝ FONIATRICKÝ KONGRES. Bratislava: Samedi s.r.o., 2010. pp. 25-26. ISSN 1337-2181.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    In dysarthria testing, there are several phonorespiration tasks that evaluate the ability to control exhalation stream. Some of these tasks are related to fricative consonant /s/. In present days these tasks are evaluated subjectively only. The aim of this paper is to introduce the possibilities of automated assessment of this tasks working in real-time. As a part of the presentation the software application is introduced.

Possibilities of Automated Assessment of /s/

  • Authors: Stráník, A., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Technical Computing Bratislava 2010. Bratislava: RT systems, s.r.o, 2010. pp. 1-5. 18th Annual Conference Proceedings. ISBN 978-80-970519-0-7.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    In dysarthria testing, there are several phonorespiration tasks that evaluate the ability to control exhalation stream. Some of these tasks are related to fricative consonant /s/. The aim of this paper is to introduce the possibilities of automated assessment of two of these tasks: "Ability to sustains /s/ on exhalation" and "Ability to repeat series of /s/".

Quantitative acoustic analysis in Parkinsons disease

  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (PD) is a neurodegenerative illness characterized by progressive loss of dopaminergic neurons in substantia nigra. In addition to many motor symptoms, in the course of PD, it has been developed non-motor deficits in speech characterized as hypokinetic dysarthria. The aim of the study is the quantitative acoustic analysis of specific speech subsystems in patients with PD and its comparison with healthy control (HC) group.

Speech characteristics for developmental dysphasia assessment

  • Authors: Nejepsová, M., Janda, J., prof. Ing. Roman Čmejla, CSc., Vokřál, J.
  • Publication: Digital Technologies 2010. Žilina: TU v Žilině, 2010. ISBN 978-80-554-0304-5.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    The purpose of this paper is to compare the appropriate parameters to assess the severity of defects in utterances from children with developmental dysphasia. There are the parameters for age-dependent classification and parameters to distinguish utterances from healthy children and children with developmental dysphasia.

System for automatic evaluation of stuttering

  • Authors: Bergl, P., prof. Ing. Roman Čmejla, CSc., Hrbková, M., Černý, L.
  • Publication: Automatizace. 2010, 53(1-2), 49-52. ISSN 0005-125X.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    This paper describes a method which based on analysis of audio recordings stutterers estimates grade of disorder

Usability of ECG and PPG for Continuous Blood Pressure Monitoring

  • Authors: Špulák, D., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Digital Technologies 2010. Žilina: TU v Žilině, 2010. ISBN 978-80-554-0304-5.
  • Year: 2010
  • Department: Department of Circuit Theory
  • Annotation:
    We used electrocardiography (ECG) and photoplethysmography (PPG) to compute pulse arrival time (PAT) between heart and finger. PAT was measured as time difference between the R-peak of ECG and a characteristic point of PPG. We defined different characteristic points of PPG and compared their suitability for blood pressure (BP) estimation. Measurements were made on a group of 13 young males using sphygmomanometric BP measurement as reference. Generally, best results were achieved when using time intervals from R-peak of ECG to the local minimum preceding the rising edge of PPG. However, for some participants, another parameters gave better results.

Voice analysis in laboratory tasks during education of biomedical engineering students

  • Department: Department of Circuit Theory
  • Annotation:
    Students of Biomedical Engineering CTU meet with the laboratory task "Voice Analysis" in the subject Biological Signals. They have the opportunity here to practically learn basics about fundamental glottal tone, its parameters and detection. Using of simple robust algorithms in MATLAB environment, the students performe analysis of fundamental frequency, show and evaluate fonetogram, and study the influence of formant frequencies to production and perception of vowels with the Klatt's cascade synthetizator.

Acoustic-phonetic measures for evaluation of dysfluent speech

  • Authors: Bergl, P., Černý, L., prof. Ing. Roman Čmejla, CSc., Hrbková, M.
  • Publication: Akustické listy. 2009, 15(2), 9-15. ISSN 1212-4702.
  • Year: 2009
  • Department: Department of Circuit Theory
  • Annotation:
    Several methods for objective evaluation of dysfluent speech are introduced: silence to speech ratio; number of segments per minute; frequency of edges of power envelope; usage of levels in speech signals. These can be grouped into one combined parameter correlating strongly (r = 0.73) to subjective judgment made by two authors (the judgment divided all participants into 5 classes according to dysfluency severity). Higher correlation coefficient is achieved for parameters based on detection of spectral changes - e.g. r = -0.77 for Bayesian detector and r = -0.8 for number of phonemes in HTK. Output of the parameters can be clustered using discriminative analysis. A combination of all described parameters correctly classified 60% of participants in training set (121 utterances in good audio quality) and 67% of participants in testing set (33 utterances in poor audio quality caused by additive interference etc.).

Analysis of Fundamental Frequency and Amplitude Instability of the Vocal Cords in Parkinson's Disease

  • Department: Department of Circuit Theory
  • Annotation:
    This study examined the acoustic measures of fundamental frequency F0, frequency instability (Jitter), amplitude instability (Shimmer) and noise to harmonic ratio (NHR), which are the traditional parameters examined in the acoustic voice analyses and it was used in the Parkinson disease (PD) patients analyses. The acoustic voice characteristics of 14 patients with early stage PD were compared with data from 14 normal control subjects (CS). Voice recordings consisted of two read texts consisting of 80 and 136 words and a free monologue lasting for 90 seconds for F0 variance measurement, one sentence consisting of 28 words and scale singing for F0 range measurement. For the next F0 range measurement, one announce, one imperative and one interrogative sentence were used. Next voice recordings consisted of at least 4 second prolongation of the vowel (i) and prolongation of the nasal (m) for Jitter, Shimmer and NHR measurements.

Formant analysis and new methods for the assessment of articulation rate in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    This study examined the measures of F1 and F2 formantswhich are the frequently used parameters examined in the acoustic voice analyses and the new methods for articulation rate measures. The first method is based on spectral distance change by using of Bayessian detector and the second is measure of F2 periodicity in articulation. Allmeasures were investigated in 16 male patients in early untreated stage of Parkinson disease (PD) and compared to 16 male members of the healthy control group (HC) of comparable age. Voice recordings consisted of at least 5 second prolongation of the vowels /a/, /e/, /i/, /o/, /u/ for formant analyses and fast repeatable syllable sequention for articulation rate measures. All calculations were performed with automatic scripts in MATLAB environment. The voice of PD groups demonstrated no significant difference in vowel space area. Present study also suggested that the PD patients' voices are characterized by lower distinct and periodicity of articulat

Occlusive weakening in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with the analysis of occlusive weakening in the repeated utterances /pa-ta-ka/, which arise as a result of inaccurate articulation in patients with Parkinson's disease (PD) and compares them with the utterances of participants in the healthy control group (HC). From the perspective of the acoustic analysis the noise is present during occlusive, which can be seen in the signal histogram, in energy waveform and difference of energy, and in longtime spectrum. This work provides simple robust statistical parameters that are selected with regard to the diversity of talk PD and HC and can be obtained without labeling. For all eleven parameters are obtained statistical significance p <0,001. The study thus complements the previously published acoustic analysis of patients with Parkinson's disease.

Parameters investigating the regularity of the energy and voicing in dysfluent utterances

  • Authors: Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Analýza a zpracování řečových a biologických signálů - sborník prací 2009. Praha: České vysoké učení technické v Praze, 2009. pp. 16-21. ISBN 978-80-01-04474-2.
  • Year: 2009
  • Department: Department of Circuit Theory
  • Annotation:
    There are two parameters for the study of dysfluent utterances in the paper. The first deals with the regularity of energy output speech signal, the second parameter to examine the periodicity of voicing utterances.

Rerence noise method of removing powerline noise from recorded signals

  • Authors: Jiruska, P., prof. Ing. Roman Čmejla, CSc., Powell, A.D., Chang, W.Ch., Vreugdenhil, M., Jefferys, J.G.R.
  • Publication: Journal of Neuroscience Methods. 2009, 184(1), 110-114. ISSN 0165-0270.
  • Year: 2009
  • DOI: 10.1016/j.jneumeth.2009.07.003
  • Link: https://doi.org/10.1016/j.jneumeth.2009.07.003
  • Department: Department of Circuit Theory
  • Annotation:
    Powerline contamination of recorded signals represents a major source of noise in electrophysiology and impairs the use of recordings for research. In this article we present simple and effective method for cancelling 50 Hz (or 60 Hz) noise using a reference noise signal and average noise cycle subtraction. This method is capable of reliably removing not only the fundamental powerline frequency but also its harmonic frequencies. The efficiency of this method appears to be superior to other commonly used methods such as notch filtering or adaptive filtering. Our experience and results show that this method can be efficiently used with very low signal-to-noise ratios, while preserving original signal waveform.

Rhythm evaluation in early untreated Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Parkinson's disease (PD) is a neurological illness which impairs motor skills, speech, and other functions such as mood, behavior, thinking, and sensation. It causes vocal impairment for approximately 70-90% of the patients. The lower ability of rhythm perception may be one of the deficits in PD speech. Thus, we performed the vocal task of one rhythmically read text according performance of examinator for measuring of the ability to maintain appropriate rhythm. It is based on the modification of the dynamic time warping algorithm called derivative dynamic time warping. The automatical scripts for this measure were performed using Matlab environment. We collected speech data from 46 Czech native speakers, 23 with Parkinson's disease. As an algorithm validation, speech performance of the each participant was evaluated by speech therapist. The voice of PD group demonstrated significant differences in maintaining the appropriate rhythm.

Acoustic voice analysis of intensity and velocity in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with the analysis of Parkinson's disease symptomps as lower intensity, tempo changes, and speech disfluences which belong to early stage of disease.

Biological Signals

  • Authors: prof. Ing. Roman Čmejla, CSc., Bergl, P.
  • Publication: Digital Technologies 2008. Žilina: Žilinská universita, Elektrotechnická fakulta, 2008. ISBN 978-80-8070-953-2.
  • Year: 2008
  • Department: Department of Circuit Theory
  • Annotation:
    The paper deals about new updated course "Biological Signals", which is reading at the Faculty of Electrical Engineering of the Czech Technical Unversity in Prague.

Change Detection with Applications to Speech

  • Authors: Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Digital Technologies 2008. Žilina: Žilinská universita, Elektrotechnická fakulta, 2008. ISBN 978-80-8070-953-2.
  • Year: 2008
  • Department: Department of Circuit Theory
  • Annotation:
    The study describes qualities of Bayesian change point detector and its applications to speech analysis.

Objective and Subjective Evaulation of Dysfluent Speech

  • Authors: Bergl, P., prof. Ing. Roman Čmejla, CSc., Černý, L., Hrbková, M.
  • Publication: Digital Technologies 2008. Žilina: Žilinská universita, Elektrotechnická fakulta, 2008. ISBN 978-80-8070-953-2.
  • Year: 2008
  • Department: Department of Circuit Theory
  • Annotation:
    Three methods for objective evaluation of dysfluent speech are introduced: 1) Silence to Speech ratio which shows that dysfluent speech contains much more silence. 2) Number of Segments per minute demonstrating stutterers are prone to utter a high number of short segments during repetitions. 3) Frequency of Edges of Power Envelope that decreases according to dysfluency severity.

Speech disorders in Parkinson disease - analysis of occurrence and symptoms

  • Authors: Růžičková, H., prof. Ing. Roman Čmejla, CSc., doc. Ing. Jan Rusz, Ph.D., Klempíř, J., Majerová, V., Roth, J., Růžička, E.
  • Publication: Česká a slovenská neurologie a neurochirurgie, 22. český a slovenský neurologický sjezd, sborník abstrakt. Praha: Česká lékařská společnost J. E. Purkyně, 2008, pp. 36. ISSN 1210-7859.
  • Year: 2008
  • Department: Department of Circuit Theory
  • Annotation:
    This paper deals with the analysis of these Parkinson's disease symptoms (lower intensity, tempo changes, speech disfluences, changes in frequency range and amplitude stability) in the beginning of their disease.

Speech fundamental frequency and velocity analysis in Parkinson's disease

  • Department: Department of Circuit Theory
  • Annotation:
    Lower intensity, tempo changes, speech disfluences, changes in both frequency range and amplitude stability belong to early stage of Parkinson's disease (PD) symptoms. This paper deals with the analysis of these PD features in the beginning of their disease. The analyses were performed with the MATLAB computing software in 14 healthy participants (Healthy Control HC) and 16 patient volunteers (Parkinson's Disease PD).

Speech fundamental frequency and velocity analysis in Parkinsons disease

  • Department: Department of Circuit Theory
  • Annotation:
    Lower intensity, tempo changes, speech disfluences, changes in both frequency range and amplitude stability belong to early stage of Parkinson's disease (PD) symptoms. This paper deals with the analysis of these PD features in the beginning of their disease. The analyses were performed with the MATLAB computing software in 14 healthy participants (Healthy Control HC) and 16 patient volunteers (Parkinson's Disease PD).

Speech Rate and Intensity Analysis in Parkinson's Disease

  • Department: Department of Circuit Theory
  • Annotation:
    The traditional parameters examined in the Parkinson disease (PD) are the articulation rate, the number of pauses and the voice intensity. These parameters where investigated in 14 PD patients and compared to 14 members of the healthy control group (HC).

Assessment of Disordered Speech Using Several Speech Characteristics

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc., Žáčková, J., Komárek, V.
  • Publication: Akustické listy. 2007, 13(3-4), 5-9. ISSN 1212-4702.
  • Year: 2007

Disordered Speech Assesment Using Different Speech Parameterizations

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: 19th International Congress on Acoustics. Madrid: Sociedad Espaňola de Acústica, 2007. pp. 1-4. ISBN 84-87985-12-2.
  • Year: 2007
  • Department: Department of Circuit Theory
  • Annotation:
    This article deals with evaluation speech of disordered children with the Landau-Kleffner syndrome (LKS) which rarely occurs in pre-school age children. LKS, also called acquired epileptiform aphasia, is a children's neurological disorder characterized by the sudden or gradual development of aphasia (loss of language) and an abnormal electro-encephalogram.

Improved Detection of Boundaries of Phonemes in Speech Databases

  • Authors: Bergl, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Proceedings of the Fifth IASTED International Conference on Biomedical Engineering. Calgary: IASTED, 2007. p. 171-174. ISBN 978-0-88986-649-2.
  • Year: 2007

Technology of Speech Communication

  • Department: Department of Circuit Theory
  • Annotation:
    The book brings complex information about digital processing of human speech in the fields of transmission, synthesis, and coding. It should serve as studying material for the subject Digital processing of speech signals from master program of structural study and also for the subject Phonetic signals and their coding from doctoral study program. The purpose of the book is also to present general overview of this research field for wide public or to present several research results obtained in this field in the research team at the Department of Circuit Theory at Faculty of Electrical Engineering at CTU in Prague.

Detection of Electrocortical Rhytms Induced by Spinal Cord Stimulation for Neuropathic Pain Patients

  • Authors: Svoboda, L., Stančák, A., Kozák, J., Vrba, I., prof. Ing. Pavel Sovka, CSc., prof. Ing. Roman Čmejla, CSc., Poláček, H., Vrána, J., Tintěra, J., Bukovský, S.
  • Publication: 53. společný sjezd české a slovenské společnosti klinické neurofyziologie. Praha: MH Consulting, 2006. pp. 32. ISBN 80-7254-916-2.
  • Year: 2006

Disordered Speech Evaluation Using the DTW Algorithm

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Analysis of Biomedical Signals and Images - Proceedings of Biosignal 2006. Brno: VUTIUM Press, 2006. pp. 70-72. ISSN 1211-412X. ISBN 80-214-3152-0.
  • Year: 2006

Disordered Speech Evaluation Using the DTW Algorithm

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Analýza a zpracování řečových a biologických signálů - Sborník prací 2006. Praha: České vysoké učení technické v Praze, 2006. pp. 84-89. ISBN 80-01-03621-9.
  • Year: 2006

Evaluation of Children Pathology Speech

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Technical Computing Prague 2006 - Sborník příspěvků 14. ročníku konference. Praha: Humusoft, 2006. pp. 108. ISBN 80-7080-616-8.
  • Year: 2006

Speech Evaluation for Children Afflicted with the Aphasia

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Nové smery v spracovaní signálov VIII. Medzinárodná vedecká konferencia. Tatranské Zruby: Akadémia ozbrojených síl gen. M. R. Štefánika, 2006. pp. 332-336. ISBN 80-8040-294-9.
  • Year: 2006

Application of the DTW Algorithm for Disordered Speech Evaluation

  • Authors: Zlatník, P., prof. Ing. Roman Čmejla, CSc.,
  • Publication: Digital Technologies 2005. Žilina: Slovenská elektrotechnická společnost, 2005. pp. 112-116. ISBN 80-8070-486-4.
  • Year: 2005

Digital Filters

Digital Filters II

Digital Filters III

Audio Signal Segmentation Using Recursive Bayesian Change-point Detectors

Audio Signal Segmentation Using Recursive Bayesian Change-point Detectors

Bayesian and Monte Carlo Change-Point Detection

Bayesian and Monte Carlo Change-Point Detection

Bayesian Detection and Localization of Signal Changes

Family of Recursive Bayesian Detectors and Applications for Signal Segmentation

Family of Recursive Bayesian Detectors and Applications for Signal Segmentation

Model Order Selection Using Penalty Criteria and Bayesian Evidence

Recursive Bayesian Autoregressive Changepoint Detector for Sequential Signal Segmentation

Introduction into Bayesian Classification Techniques

Study of Detection of Changes in EEG Signal

Using of Bayesian Detectors for EEG Signal Changes

Blind Location of Phonetic Boundaries

Estimation of Boundaries between Speech Units Using Bayesian Changepoint Detectors

Head Reconstruction and Localization of Brain Activity Using Bayesian Evidence

  • Authors: Králík, J., prof. Ing. Roman Čmejla, CSc., prof. Ing. Pavel Sovka, CSc., Stančák, A.
  • Publication: Proceeding of the 23rd Annual International Conference of the IEEE,Engineering in Medicine and Biology Society. New York: IEEE Press, 2001. p. 130. ISBN 0-7803-7211-5.
  • Year: 2001

Analysis of the Electrical Engineering Problem Using Computer Tools

AR model order selection in speech processing

Speech Technology Education at the Czech Technical University in Prague

Application of Bayesian Detector in Speech Processing

Modified Spectral Subtraction Based Speech Enhancement

Notes on Bayesian Detector Use

The MATLAB Environment for Circuit Theory Teaching

The Using of Bayesian Autoregressive Detector for Speech Segmentation

Adaptive Filtering for Vowel Description

Adaptive Formant Tracking and Classification of Czech Vowels

Speech Segmentation Using Bayesian Autoregressive Changepoint Detector

Integrated Information System of the Czech Technical University in Prague

Teaching of Electrical Engineering Courses in English in CTU Prague

An Algorithm of Formant Extraction in Speech Signal

  • Authors: prof. Ing. Roman Čmejla, CSc.,
  • Publication: 31st Conference on Acoustics. Praha: České vysoké učení technické v Praze, 1994, pp. 161-163. ISBN 80-01-01146-1.
  • Year: 1994

PCS - VOICE: Software for Czech speech synthesis

Rule Based Speech Recognition

  • Authors: prof. Ing. Roman Čmejla, CSc.,
  • Publication: Proceedings of the Czech, Hungariean, and Polish Workshop on Circuit Theory and Applications. Budapest: Technical University, 1994. pp. 30-34.
  • Year: 1994

Turbotutor-Program System for Circuit Theory Education

Responsible person Ing. Mgr. Radovan Suk