prof. Dr. Ing. Jan Kybic

Automatic caries detection in bitewing radiographs—Part II: experimental comparison

Autoři: Tichý, A., Kunt, L., Nagyová, V., prof. Dr. Ing. Jan Kybic,
Publikace: Clinical Oral Investigations. 2024, 28(2), ISSN 1436-3771.
Rok: 2024

DOI: 10.1007/s00784-024-05528-2
Odkaz: https://doi.org/10.1007/s00784-024-05528-2
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Objective The objective of this study was to compare the detection of caries in bitewing radiographs by multiple dentists with an automatic method and to evaluate the detection performance in the absence of a reliable ground truth. Materials and methods Four experts and three novices marked caries using bounding boxes in 100 bitewing radiographs. The same dataset was processed by an automatic object detection deep learning method. All annotators were compared in terms of the number of errors and intersection over union (IoU) using pairwise comparisons, with respect to the consensus standard, and with respect to the annotator of the training dataset of the automatic method. Results The number of lesions marked by experts in 100 images varied between 241 and 425. Pairwise comparisons showed that the automatic method outperformed all dentists except the original annotator in the mean number of errors, while being among the best in terms of IoU. With respect to a consensus standard, the performance of the automatic method was best in terms of the number of errors and slightly below average in terms of IoU. Compared with the original annotator, the automatic method had the highest IoU and only one expert made fewer errors. Conclusions The automatic method consistently outperformed novices and performed as well as highly experienced dentists. Clinical significance The consensus in caries detection between experts is low. An automatic method based on deep learning can improve both the accuracy and repeatability of caries detection, providing a useful second opinion even for very experienced dentists.

Automatic caries detection in bitewing radiographs: part I---deep learning

Autoři: Kunt, L., prof. Dr. Ing. Jan Kybic, Nagyová, V., Tichý, A.
Publikace: Clinical Oral Investigations. 2023, 27(12), 7463-7471. ISSN 1436-3771.
Rok: 2023

DOI: 10.1007/s00784-023-05335-1
Odkaz: https://doi.org/10.1007/s00784-023-05335-1
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Objective: The aim of this work was to assemble a large annotated dataset of bitewing radiographs and to use convolutional neural networks to automate the detection of dental caries in bitewing radiographs with human-level performance. Materials and methods: A dataset of 3989 bitewing radiographs was created, and 7257 carious lesions were annotated using minimal bounding boxes. The dataset was then divided into 3 parts for the training (70%), validation (15%), and testing (15%) of multiple object detection convolutional neural networks (CNN). The tested CNN architectures included YOLOv5, Faster R-CNN, RetinaNet, and EfficientDet. To further improve the detection performance, model ensembling was used, and nested predictions were removed during post-processing. The models were compared in terms of the [Formula: see text] score and average precision (AP) with various thresholds of the intersection over union (IoU). Results: The twelve tested architectures had [Formula: see text] scores of 0.72-0.76. Their performance was improved by ensembling which increased the [Formula: see text] score to 0.79-0.80. The best-performing ensemble detected caries with the precision of 0.83, recall of 0.77, [Formula: see text], and AP of 0.86 at IoU=0.5. Small carious lesions were predicted with slightly lower accuracy (AP 0.82) than medium or large lesions (AP 0.88). Conclusions: The trained ensemble of object detection CNNs detected caries with satisfactory accuracy and performed at least as well as experienced dentists (see companion paper, Part II). The performance on small lesions was likely limited by inconsistencies in the training dataset. Clinical significance: Caries can be automatically detected using convolutional neural networks. However, detecting incipient carious lesions remains challenging.

Deep learning for laser beam imprinting

Autoři: Chalupský, J., Vozda, V., Hering, J., prof. Dr. Ing. Jan Kybic, Burian, T., Dziarzhytski, S., Frantálová, K., Hájková, V., Jelínek, Š., Juha, L., Keitel, B., Kuglerová, Z., Kuhlmann, M., Petryshak, B., Ruiz-Lope, M., Vyšín, L., Wodzinski, T., Plönjes, E.
Publikace: Optics Express. 2023, 31(12), 19703-19721. ISSN 1094-4087.
Rok: 2023

DOI: 10.1364/OE.481776
Odkaz: https://doi.org/10.1364/OE.481776
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Methods of ablation imprints in solid targets are widely used to characterize focused X-ray laser beams due to a remarkable dynamic range and resolving power. A detailed description of intense beam profiles is especially important in high-energy-density physics aiming at nonlinear phenomena. Complex interaction experiments require an enormous number of imprints to be created under all desired conditions making the analysis demanding and requiring a huge amount of human work. Here, for the first time, we present ablation imprinting methods assisted by deep learning approaches. Employing a multi-layer convolutional neural network (U-Net) trained on thousands of manually annotated ablation imprints in poly(methyl methacrylate), we characterize a focused beam of beamline FL24/FLASH2 at the Free-electron laser in Hamburg. The performance of the neural network is subject to a thorough benchmark test and comparison with experienced human analysts. Methods presented in this Paper pave the way towards a virtual analyst automatically processing experimental data from start to end.

Fully automated imaging protocol independent system for pituitary adenoma segmentation: a convolutional neural network—based model on sparsely annotated MRI

Autoři: Černý, M., prof. Dr. Ing. Jan Kybic, Májovký, M., Sedlák, V., Pirgl, K., Misiorzová, E., Lipina, R., Netuka, D.
Publikace: Neurosurgical Review. 2023, 46(1), ISSN 0344-5607.
Rok: 2023

DOI: 10.1007/s10143-023-02014-3
Odkaz: https://doi.org/10.1007/s10143-023-02014-3
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
This study aims to develop a fully automated imaging protocol independent system for pituitary adenoma segmentation from magnetic resonance imaging (MRI) scans that can work without user interaction and evaluate its accuracy and utility for clinical applications. We trained two independent artificial neural networks on MRI scans of 394 patients. The scans were acquired according to various imaging protocols over the course of 11 years on 1.5T and 3T MRI systems. The segmentation model assigned a class label to each input pixel (pituitary adenoma, internal carotid artery, normal pituitary gland, background). The slice segmentation model classified slices as clinically relevant (structures of interest in slice) or irrelevant (anterior or posterior to sella turcica). We used MRI data of another 99 patients to evaluate the performance of the model during training. We validated the model on a prospective cohort of 28 patients, Dice coefficients of 0.910, 0.719, and 0.240 for tumour, internal carotid artery, and normal gland labels, respectively, were achieved. The slice selection model achieved 82.5% accuracy, 88.7% sensitivity, 76.7% specificity, and an AUC of 0.904. A human expert rated 71.4% of the segmentation results as accurate, 21.4% as slightly inaccurate, and 7.1% as coarsely inaccurate. Our model achieved good results comparable with recent works of other authors on the largest dataset to date and generalized well for various imaging protocols. We discussed future clinical applications, and their considerations. Models and frameworks for clinical use have yet to be developed and evaluated.

Learning to segment from object thickness annotations

Autoři: Ing. Denis Baručić, prof. Dr. Ing. Jan Kybic,
Publikace: 2023 IEEE 20th International Symposium on Biomedical Imaging (ISBI). New Jersey: IEEE Signal Processing Society, 2023. ISSN 1945-8452. ISBN 978-1-6654-7358-3.
Rok: 2023

DOI: 10.1109/ISBI53787.2023.10230621
Odkaz: https://doi.org/10.1109/ISBI53787.2023.10230621
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Measuring object size is fast and a standard part of many radiological evaluation procedures. We describe a deep learning segmentation method that can be trained on a small number of pixel-wise reference segmentation and then fine-tuned from the weak annotations of the object thickness. The difficulty is in the non-differentiability of the thickness function defined using the pixel-wise distance transform. We overcome it by optimizing the expected value of the loss function after the injection of a virtual random noise. Further speedup is possible using the properties of the distance transform. We demonstrate the benefit of the proposed method on ultrasound images of the carotid artery. The fine-tuning improves the performance by about 10% IoU.

Characterization of drug effects on cell cultures from phase-contrast microscopy images

Autoři: Ing. Denis Baručić, Kaushik, S., prof. Dr. Ing. Jan Kybic, Stanková, J., Džubák, P., Hajdúch, M.
Publikace: Computers in Biology and Medicine. 2022, 151 ISSN 0010-4825.
Rok: 2022

DOI: 10.1016/j.compbiomed.2022.106171
Odkaz: https://doi.org/10.1016/j.compbiomed.2022.106171
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
In this work, we classify chemotherapeutic agents (topoisomerase inhibitors) based on their effect on U-2 OS cells. We use phase-contrast microscopy images, which are faster and easier to obtain than fluorescence images and support live cell imaging. We use a convolutional neural network (CNN) trained end-to-end directly on the input images without requiring for manual segmentations or any other auxiliary data. Our method can distinguish between tested cytotoxic drugs with an accuracy of 98%, provided that their mechanism of action differs, outperforming previous work. The results are even better when substance-specific concentrations are used. We show the benefit of sharing the extracted features over all classes (drugs). Finally, a 2D visualization of these features reveals clusters, which correspond well to known class labels, suggesting the possible use of our methodology for drug discovery application in analyzing new, unseen drugs.

Fast learning from label proportions with small bags

Autoři: Ing. Denis Baručić, prof. Dr. Ing. Jan Kybic,
Publikace: 2022 IEEE International Conference on Image Processing (ICIP). Piscataway, NJ: IEEE, 2022. p. 3156-3160. ISSN 2381-8549. ISBN 978-1-6654-9620-9.
Rok: 2022

DOI: 10.1109/ICIP46576.2022.9897895
Odkaz: https://doi.org/10.1109/ICIP46576.2022.9897895
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
In learning from label proportions (LLP), the instances are grouped into bags, and the task is to learn an instance classifier given relative class proportions in training bags. LLP is useful when obtaining individual instance labels is impossible or costly. In this work, we focus on the case of small bags, which allows to design an algorithm that explicitly considers all consistent instance label combinations. In particular, we propose an EM algorithm alternating between optimizing a general neural network instance classifier and incorporating bag-level annotations. Using two different image datasets, we experimentally compare this method with an approach based on normal approximation and two existing LLP methods. The results show that our approach converges faster to a comparable or better solution.

Learning to segment cell nuclei in phase-contrast microscopy from fluorescence images for drug discovery

Autoři: Mertanová, H., prof. Dr. Ing. Jan Kybic, Staňková, J., Džubák, P., Hajdúch, M.
Publikace: Proc. SPIE 12032: Medical Imaging 2022: Image Processing. Bellingham: SPIE, 2022. SPIE. vol. 12032. ISSN 1605-7422. ISBN 978-1-5106-4939-2.
Rok: 2022

DOI: 10.1117/12.2607500
Odkaz: https://doi.org/10.1117/12.2607500
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We describe a method for analyzing geometrical properties of cell nuclei from phase contrast microscopy images. This is useful in drug discovery for quantifying the effect of candidate chemical compounds, bypassing the need for fluorescence imaging. Fluorescence images are then only used for training our nuclei segmentation, avoiding the need for the time consuming expert annotations. Geometry based descriptors are calculated and aggregated and fed into a classifier to distinguish the different types of chemical treatments. The drug treatment can be distinguished from no treatment with accuracy better than 95% from fluorescence images and better than 77% from phase contrast images.

Learning to segment from object sizes

Autoři: Ing. Denis Baručić, prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of the 22nd Conference Information Technologies – Applications and Theory (ITAT 2022). CEUR-WS.org, 2022. p. 55-60. CEUR Workshop Proceedings. vol. 3226. ISSN 1613-0073.
Rok: 2022

Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Deep learning has proved particularly useful for semantic segmentation, a fundamental image analysis task. However, the standard deep learning methods need many training images with ground-truth pixel-wise annotations, which are usually laborious to obtain and, in some cases (e.g., medical images), require domain expertise. Therefore, instead of pixel-wise annotations, we focus on image annotations that are significantly easier to acquire but still informative, namely the size of foreground objects. We define the object size as the maximum Chebyshev distance between a foreground and the nearest background pixel. We propose an algorithm for training a deep segmentation network from a dataset of a few pixel-wise annotated images and many images with known object sizes. The algorithm minimizes a discrete (non-differentiable) loss function defined over the object sizes by sampling the gradient and then using the standard back-propagation algorithm. Experiments show that the new approach improves the segmentation performance.

The Effect of Primary Aldosteronism on Carotid Artery Texture in Ultrasound Images

Autoři: Kaushik, S., Majtan, B., Holaj, R., Ing. Denis Baručić, Kološová, B., Widimský, J., prof. Dr. Ing. Jan Kybic,
Publikace: Diagnostics. 2022, 12(12), ISSN 2075-4418.
Rok: 2022

DOI: 10.3390/diagnostics12123206
Odkaz: https://doi.org/10.3390/diagnostics12123206
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Primary aldosteronism (PA) is the most frequent cause of secondary hypertension. Early diagnoses of PA are essential to avoid the long-term negative effects of elevated aldosterone concentration on the cardiovascular and renal system. In this work, we study the texture of the carotid artery vessel wall from longitudinal ultrasound images in order to automatically distinguish between PA and essential hypertension (EH). The texture is characterized using 140 Haralick and 10 wavelet features evaluated in a region of interest in the vessel wall, followed by the XGBoost classifier. Carotid ultrasound studies were carried out on 33 patients aged 42–72 years with PA, 52 patients with EH, and 33 normotensive controls. For the most clinically relevant task of distinguishing PA and EH classes, we achieved a classification accuracy of 73% as assessed by a leave-one-out procedure. This result is promising even compared to the 57% prediction accuracy using clinical characteristics alone or 63% accuracy using a combination of clinical characteristics and intima-media thickness (IMT) parameters. If the accuracy is improved and the method incorporated into standard clinical procedures, this could eventually lead to an improvement in the early diagnosis of PA and consequently improve the clinical outcome for these patients in future.

Automatic evaluation of human oocyte developmental potential from microscopy images

Autoři: Ing. Denis Baručić, prof. Dr. Ing. Jan Kybic, Teplá, O., Topurko, Z., Kratochvílová, I.
Publikace: 17th International Symposium on Medical Information Processing and Analysis. Bellingham (stát Washington): SPIE, 2021. ISSN 0277-786X. ISBN 978-1-5106-5052-7.
Rok: 2021

DOI: 10.1117/12.2604010
Odkaz: https://doi.org/10.1117/12.2604010
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Infertility is becoming an issue for an increasing number of couples. The most common solution, in vitro fertilization, requires embryologists to carefully examine light microscopy images of human oocytes to determine their developmental potential. We propose an automatic system to improve the speed, repeatability, and accuracy of this process. We first localize individual oocytes and identify their principal components using CNN (U-Net) segmentation. Next, we calculate several descriptors based on geometry and texture. The final step is an SVM classifier. Both the segmentation and classification training is based on expert annotations. The presented approach leads to a classification accuracy of 70%.

Differentiating between stable and progressive carotid atherosclerotic plaques from in-vivo ultrasound images using texture descriptors

Autoři: Kostelanský, M., Manzano-Rodríguez, A., prof. Dr. Ing. Jan Kybic, Hekrdla, M., Dvorský, O., Kozel, J., Baurová, P., Pakizer, D., Školoudík, D.
Publikace: 17th International Symposium on Medical Information Processing and Analysis. Bellingham (stát Washington): SPIE, 2021. ISSN 0277-786X. ISBN 978-1-5106-5052-7.
Rok: 2021

DOI: 10.1117/12.2605795
Odkaz: https://doi.org/10.1117/12.2605795
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We describe an automatic pipeline for processing ultrasound images of the carotid artery, consisting of image type classification, carotid artery localization, segmentation, feature descriptor extraction, and plaque stability classification. The aim is to distinguish between stable (safe) and progressive (dangerous) atherosclerotic plaques from a single standard ultrasound transversal or longitudinal B-mode examination. The processing pipeline uses modern deep CNN techniques, while the descriptors are based on geometry and wavelets to characterize texture. When testing on a large dataset of 28718 images from 413 patients, we found that our automatically calculated descriptors are statistically significantly different between the two classes with a very high significance level, p<0.001. We have also created a random forest-based classifier to distinguish between progressive and stable plaques, although its accuracy remains low (61~62%).

Potential Biomarkers From Positive Definite 4th Order Tensors In Hardi

Autoři: Kaushik, S., prof. Dr. Ing. Jan Kybic, Bansal, A., Tsegey, T., Slovak, J.
Publikace: 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI). New Jersey: IEEE Signal Processing Society, 2021. p. 1003-1006. ISSN 1945-8452. ISBN 978-1-6654-1246-9.
Rok: 2021

DOI: 10.1109/ISBI48211.2021.9434144
Odkaz: https://doi.org/10.1109/ISBI48211.2021.9434144
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
In this paper, we provide a framework to evaluate new scalar quantities for higher order tensors (HOT) appearing in high angular resolution diffusion imaging (HARDI). These can potentially serve as biomarkers. It involves flattening of HOTs and extraction of the diagonal D-components. Experiments performed in the 4th order case reveal that D-components encode geometric information unlike the isometric 6D 2nd order Voigt form. The existing invariants obtained from the Voigt form are considered for comparison. We also notice that D-components can be useful in segmentation of white matter structures in crossing regions and classification. Results on phantom and the synthetic dataset support the conclusions

ANHIR: Automatic Non-rigid Histological Image Registration Challenge

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic, Arganda-Carreras, I.
Publikace: IEEE Transactions on Medical Imaging. 2020, 39(10), 3042-3052. ISSN 1558-254X.
Rok: 2020

DOI: 10.1109/TMI.2020.2986331
Odkaz: https://doi.org/10.1109/TMI.2020.2986331
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Automatic Non-rigid Histological Image Registration (ANHIR) challenge was organized to compare the performance of image registration algorithms on several kinds of microscopy histology images in a fair and independent manner. We have assembled 8 datasets, containing 355 images with 18 different stains, resulting in 481 image pairs to be registered. Registration accuracy was evaluated using manually placed landmarks. In total, 256 teams registered for the challenge, 10 submitted the results, and 6 participated in the workshop. Here, we present the results of 7 well-performing methods from the challenge together with 6 well-known existing methods. The best methods used coarse but robust initial alignment, followed by non-rigid registration, used multiresolution, and were carefully tuned for the data at hand. They outperformed off-the-shelf methods, mostly by being more robust. The best methods could successfully register over 98% of all landmarks and their mean landmark registration accuracy (TRE) was 0.44% of the image diagonal. The challenge remains open to submissions and all images are available for download.

Generalized Multiple Instance Learning for Cancer Detection in Digital Histopathology

Autoři: Hering, J., prof. Dr. Ing. Jan Kybic,
Publikace: Image Analysis and Recognition,17th International Conference, ICIAR 2020, Póvoa de Varzim, Portugal, June 24–26, 2020, Proceedings, Part II. Cham: Springer, 2020. p. 274-282. LNCS. vol. 12132. ISSN 0302-9743. ISBN 978-3-030-50515-8.
Rok: 2020

DOI: 10.1007/978-3-030-50516-5_24
Odkaz: https://doi.org/10.1007/978-3-030-50516-5_24
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We address the task of detecting cancer in histological slide images based on training with weak, slide- and patch-level annotations, which are considerably easier to obtain than pixel-level annotations. we use CNN based patch-level descriptors and formulate the image classification task as a generalized multiple instance learning (MIL) problem. The generalization consists of requiring a certain number of positive instances in positive bags, instead of just one as in standard MIL. The descriptors are learned on a small number of patch-level annotations, while the MIL layer uses only image-level patches for training. We evaluate multiple generalized MIL methods on the H&E stained images of lymphatic nodes from the CAMELYON dataset and show that generalized MIL methods improve the classification results and outperform no-MIL methods in terms of slide-level AUC. Best classification results were achieved by the MI-SVM(k) classifier in combination with simple spatial Gaussian aggregation, achieving AUC 0.962. However, MIL did not outperform methods trained on pixel-level segmentations.

Multiple Instance Learning Via Deep Hierarchical Exploration for Histology Image Classification

Autoři: Hering, J., prof. Dr. Ing. Jan Kybic,
Publikace: 17th IEEE International Symposium on Biomedical Imaging. Los Alamitos: IEEE Computer Society, 2020. p. 235-238. vol. 2020. ISSN 1945-7928. ISBN 978-1-5386-9330-8.
Rok: 2020

DOI: 10.1109/ISBI45749.2020.9098616
Odkaz: https://doi.org/10.1109/ISBI45749.2020.9098616
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We present a fast hierarchical method to detect a presence of cancerous tissue in histological images. The image is not examined in detail everywhere but only inside several small regions of interest, called glimpses. The final classification is done by aggregating classification scores from a CNN on leaf glimpses at the highest resolution. Unlike in existing attention-based methods, the glimpses form a tree structure, low resolution glimpses determining the location of several higher resolution glimpses using weighted sampling and a CNN approximation of the expected scores. We show that it is possible to perform the classification with just a small number of glimpses, leading to an important speedup with only a small performance deterioration. Learning is possible using image labels only, as in the multiple instance learning (MIL) setting.

AbloCAM: a versatile optomechanical, semi-automated tool for an in situ characterization and optimization of focused XUV/x-ray laser beam (Conference Presentation)

Autoři: Juráňová, K., Burian, T., Hájková, V., Hošek, J., Juha, L., prof. Dr. Ing. Jan Kybic, Němcová, Š., Petryshak, B., Vozda, V., Vyšín, L., Chalupský, J.
Publikace: Proc. SPIE 11035, Optics Damage and Materials Processing by EUV/X-ray Radiation VII. Bellingham (stát Washington): SPIE, 2019. Proceedings of SPIE. vol. 11035. ISSN 1996-756X. ISBN 978-1-5106-2737-6.
Rok: 2019

DOI: 10.1117/12.2524883
Odkaz: https://doi.org/10.1117/12.2524883
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Technically, the AbloCAM tool consists of a vacuum compatible motorized positioning system executing a series of well-defined irradiations of a chosen slab target according to algorithms fulfilling requirements of the combined F(z)-scan procedure. Damage patterns formed in that way should then be visualized in situ by means of Nomarski (DIC – Differential Interference Contrast) microscope equipped with the software which indicates and processes pattern outer contours. There is a feedback established between positioning and inspecting components and functions of the tool. The software helps to align and characterize any focused beam in the interaction chamber semi-automatically in a reasonable time.

Benchmarking of image registration methods for differently stained histological slides

Autoři: Borovec, J., Munoz-Barrutia, A., prof. Dr. Ing. Jan Kybic,
Publikace: International Conference on Image Processing. IEEE (Institute of Electrical and Electronics Engineers), 2018. p. 3368-3372. ISSN 2381-8549. ISBN 978-1-4799-7061-2.
Rok: 2018

DOI: 10.1109/ICIP.2018.8451040
Odkaz: https://doi.org/10.1109/ICIP.2018.8451040
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Image registration is a common task for many biomedical analysis applications. The present work focuses on the benchmarking of registration methods on differently stained histological slides. This is a challenging task due to the differences in the appearance model, the repetitive texture of the details and the large image size, between other issues. Our benchmarking data is composed of 616 image pairs at two different scales - average image diagonal 2.4k and 5k pixels. We compare eleven fully automatic registration methods covering the widely used similarity measures (and optimization strategies with both linear and elastic transformation). For each method, the best parameter configuration is found and subsequently applied to all the image pairs. The performance of the algorithms is evaluated from several perspectives - the registrations (in)accuracy on manually annotated landmarks, the method robustness and its processing computation time.

Detecting multiple myeloma via generalized multiple-instance learning

Autoři: Hering, J., prof. Dr. Ing. Jan Kybic, Lambert, L.
Publikace: SPIE Medical Imaging 2018. Baltimore: SPIE, 2018. Image Processing. vol. 10574. ISSN 0277-786X. ISBN 978-1-5106-1638-7.
Rok: 2018

DOI: 10.1117/12.2293112
Odkaz: https://doi.org/10.1117/12.2293112
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We address the task of automatic detection of lesions caused by multiple myeloma (MM) in femurs or other long bones from CT data. Such detection is already an important part of the multiple myeloma diagnosis and staging. However, it is so far performed mostly manually, which is very time consuming. We formulate the detection as a multiple instance learning (MIL) problem, where instances are grouped into bags and only bag labels are available. In our case, instances are regions in the image and bags correspond to images. This has the advantage of requiring only subject-level annotation (ground truth), which is much easier to get than voxel-level manual segmentation. We consider a generalization of the standard MIL formulation where we introduce a threshold on the number of required positive instances in positive bags. This corresponds better to the classification procedure used by the radiology experts and is more robust with respect to false positive instances. We extend several existing MIL algorithms to solve the generalized case by estimating the threshold during learning. We compare the proposed methods with the baseline method on a dataset of 220 subjects. We show that the generalized MIL formulation outperforms standard MIL methods for this task. For the task of distinguishing between healthy controls and MM patients with infiltrations, our best method makes almost no mistakes with a mean AUC of 0.982 and F1 = 0.965. We outperform the baseline method significantly in all conducted experiments

Fast registration by boundary sampling and linear programming

Autoři: prof. Dr. Ing. Jan Kybic, Borovec, J.
Publikace: Medical Image Computing and Computer Assisted Intervention, Part I. Springer, Cham, 2018. p. 783-791. Lecture Notes in Computer Science. vol. 11070. ISSN 0302-9743. ISBN 978-3-030-00927-4.
Rok: 2018

DOI: 10.1007/978-3-030-00928-1_88
Odkaz: https://doi.org/10.1007/978-3-030-00928-1_88
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We address the problem of image registration when speed is more important than accuracy. We present a series of simplification and approximations applicable to almost any pixel-based image similarity criterion. We first sample the image at a set of sparse keypoints in a direction normal to image edges and then create a piecewise linear convex approximation of the individual contributions. We obtain a linear program for which a global optimum can be found very quickly by standard algorithms. The linear program formulation also allows for an easy addition of regularization and trust-region bounds. We have tested the approach for affine and B-spline transformation representation but any linear model can be used. Larger deformations can be handled by multiresolution. We show that our method is much faster than pixel- based registration, with only a small loss of accuracy. In comparison to standard keypoint based registration, our method is applicable even if individual keypoints cannot be reliably identiffied and matched.

Incremental B-spline deformation model for geometric graph matching

Autoři: Dos Santos Pinheiro, M., prof. Dr. Ing. Jan Kybic,
Publikace: 2018 IEEE 15th IEEE International Symposium on Biomedical Imaging (ISBI). New Jersey: IEEE Signal Processing Society, 2018. p. 1079-1082. ISSN 1945-7928. ISBN 978-1-5386-3636-7.
Rok: 2018

DOI: 10.1109/ISBI.2018.8363758
Odkaz: https://doi.org/10.1109/ISBI.2018.8363758
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We propose a B-spline deformation model, which can be efficiently updated from sequential measurements. Our incremental update is based on the Kalman filtering, is optimal in the least squares sense, and is very fast, an order of magnitude faster than the direct methods, while converging to the same solution. While this method can be applied to any multidimensional scattered data interpolation task, our main application is a geometric graph matching algorithm, which is used for registering large images from the biomedical domain or remote sensing, based on matching linear structures such as blood vessels or roads. The B-spline transformation model needs to be updated incrementally as more points are added to the match and using the proposed Kalman-like update yields substantial speed gains over the previously used bi-Lipschitz model.

Reconstructing Evolving Tree Structures in Time Lapse Sequences by Enforcing Time-Consistency

Autoři: Glowacki, P., Dos Santos Pinheiro, M., Mosinska, A., Turetken, E., Lebrecht, D., Sznitman, R., Holtmaat, A., prof. Dr. Ing. Jan Kybic, Fua, P.
Publikace: IEEE Transactions on Pattern Analysis and Machine Intelligence. 2018, 40(3), 755-761. ISSN 0162-8828.
Rok: 2018

DOI: 10.1109/TPAMI.2017.2680444
Odkaz: https://doi.org/10.1109/TPAMI.2017.2680444
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We propose a novel approach to reconstructing curvilinear tree structures evolving over time, such as road networks in 2D aerial images or neural structures in 3D microscopy stacks acquired in vivo. To enforce temporal consistency, we simultaneously process all images in a sequence, as opposed to reconstructing structures of interest in each image independently. We formulate the problem as a Quadratic Mixed Integer Program and demonstrate the additional robustness that comes from using all available visual clues at once, instead of working frame by frame. Furthermore, when the linear structures undergo local changes over time, our approach automatically detects them.

Volume estimation from single images: an application to pancreatic islets

Autoři: Dvořák, J., doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Radochová, B., Janáček, J., Kukal, J., Borovec, J., Habart, D.
Publikace: Image Analysis and Stereology. 2018, 37(3), 191-204. ISSN 1580-3139.
Rok: 2018

DOI: 10.5566/ias.1869
Odkaz: https://doi.org/10.5566/ias.1869
Pracoviště: Katedra radioelektroniky, Algoritmy pro biomedicínské zobrazování
Anotace:
The present paper deals with the problem of volume estimation of individual objects from a single 2D view. Our main application is volume estimation of pancreatic (Langerhans) islets and the single 2D view constraint comes from the time and equipment limitations of the standard clinical procedure. Two main approaches are followed in this paper. First, two regression-based methods are proposed, using a set of simple shape descriptors of the segmented image of the islet. Second, two example-based methods are proposed, based on a database of islets with known volume. For training and evaluation, islet volumes were determined by OPT microscopy and a stereological volume estimation using the so-called Fakir probes. The performance of the single image volume estimation methods is studied on a set of 99 islets from human donors. Further experiments were also performed on a stone dataset and on synthetic 3D shapes, generated using a flexible stochastic particle model. The proposed methods are fast and the experimental results show that in most situations the proposed methods perform significantly better than the methods currently used in clinical practice, which are based on simple spherical or ellipsoidal models.

Binary pattern dictionary learning for gene expression representation in drosophila imaginal discs

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic,
Publikace: Computer Vision – ACCV 2016 Workshops. Cham: Springer International Publishing, 2017. p. 555-569. Lecture Notes in Computer Science. vol. 10117. ISSN 0302-9743. ISBN 978-3-319-54426-7.
Rok: 2017

DOI: 10.1007/978-3-319-54427-4_40
Odkaz: https://doi.org/10.1007/978-3-319-54427-4_40
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We present an image processing pipeline which accepts a largenumber of images, containing spatial expression information for thousands of genes in Drosophila imaginal discs. We assume that the geneactivations are binary and can be expressed as a union of a small setof non-overlapping spatial patterns, yielding a compact representationof the spatial activation of each gene. This lends itself well to furtherautomatic analysis, with the hope of discovering new biological relationships. Traditionally, the images were labeled manually, which was verytime consuming. The key part of our work is a binary pattern dictionarylearning algorithm, that takes a set of binary images and determinesa set of patterns, which can be used to represent the input images witha small error. We also describe the preprocessing phase, where input images are segmented to recover the activation images and spatially alignedto a common reference. We compare binary pattern dictionary learningto existing alternative methods on synthetic data and also show resultsof the algorithm on real microscopy images of the Drosophila imaginaldiscs.

Detection and Localization of Drosophila Egg Chambers in Microscopy Images

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic, Nava Velazco, U.
Publikace: 8th International Workshop on Machine Learning in Medical Imaging (with MICCAI 2017). Cham: Springer International Publishing, 2017. p. 19-26. Lecture Notes of Computer Science. vol. 10541. ISSN 0302-9743. ISBN 978-3-319-67388-2.
Rok: 2017

DOI: 10.1007/978-3-319-67389-9_3
Odkaz: https://doi.org/10.1007/978-3-319-67389-9_3
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Drosophila melanogaster is a well-known model organism that can be used for studying oogenesis (egg chamber development) including gene expression patterns. Standard analysis methods require manual segmentation of individual egg chambers, which is a difficult and time-consuming task. We present an image processing pipeline to detect and localize Drosophila egg chambers that consists of the following steps: (i) superpixel-based image segmentation into relevant tissue classes; (ii) detection of egg center candidates using label histograms and ray features; (iii) clustering of center candidates and; (iv) area-based maximum likelihood ellipse model fitting. Our proposal is able to detect 96% of human-expert annotated egg chambers at relevant developmental stages with less than 1% false-positive rate, which is adequate for the further analysis.

Geometric Graph Matching Using Monte Carlo Tree Search

Autoři: Pinheiro, M.A., prof. Dr. Ing. Jan Kybic, Fua, P.
Publikace: IEEE Transactions on Pattern Analysis and Machine Intelligence. 2017, 39(11), 2171-2185. ISSN 0162-8828.
Rok: 2017

DOI: 10.1109/TPAMI.2016.2636200
Odkaz: https://doi.org/10.1109/TPAMI.2016.2636200
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
We present an efficient matching method for generalized geometric graphs. Such graphs consist of vertices in space connected by curves and can represent many real world structures such as road networks in remote sensing, or vessel networks in medical imaging. Graph matching can be used for very fast and possibly multimodal registration of images of these structures. We formulate the matching problem as a single player game solved using Monte Carlo Tree Search, which automatically balances exploring new possible matches and extending existing matches. Our method can handle partial matches, topological differences, geometrical distortion, does not use appearance information and does not require an initial alignment. Moreover, our method is very efficient — it can match graphs with thousands of nodes, which is an order of magnitude better than the best competing method, and the matching only takes a few seconds.

Langerhans islet volume estimation from 3D optical projection tomography

Autoři: doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Habart, D., Hlushak, H., Dvořák, J., Radochová, B.
Publikace: Computer Vision – ACCV 2016 Workshops. Cham: Springer International Publishing, 2017. p. 583-594. Lecture Notes in Computer Science. vol. 10117. ISSN 0302-9743. ISBN 978-3-319-54426-7.
Rok: 2017

DOI: 10.1007/978-3-319-54427-4_42
Odkaz: https://doi.org/10.1007/978-3-319-54427-4_42
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
This paper concerns the comparison of automatic volume estimation methods for isolated pancreatic islets. The estimated islet volumes are needed during the process of assessing the islet sample quality prior to the islet transplantation. We study several different methods for automatic volume estimation. For this purpose we acquired a set of projections using optical tomography for a sample of an islet population. Based on these projections we estimated the islet volumes using two stereological methods (the automatic Wulfsohn’s method and the manual fakir method, considered to be the ground truth in this study), together with the filtered back projection followed by 3D segmentation. We have also employed two simple methods, currently used in medical practice, based on fitting a sphere or a prolate ellipsoid to a single binarized 2D islet projections.

Left ventricle Hermite-based segmentation

Autoři: Olveres, J., Nava Velazco, U., Escalante-Ramirez, B., Vallejo, E., prof. Dr. Ing. Jan Kybic,
Publikace: Computers in Biology and Medicine. 2017, 87 236-249. ISSN 0010-4825.
Rok: 2017

DOI: 10.1016/j.compbiomed.2017.05.025
Odkaz: https://doi.org/10.1016/j.compbiomed.2017.05.025
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
In recent years, computed tomography (CT) has become a standard technique in cardiac imaging because it provides detailed information that may facilitate the diagnosis of the conditions that interfere with correct heart function. However, CT-based cardiac diagnosis requires manual segmentation of heart cavities, which is a difficult and time-consuming task. Thus, in this paper, we propose a novel technique to segment endocardium and epicardium boundaries based on a 2D approach. The proposal computes relevant information of the left ventricle and its adjacent structures using the Hermite transform. The novelty of the work is that the information is combined with active shape models and level sets to improve the segmentation. Our database consists of mid-third slices selected from 28 volumes manually segmented by expert physicians. The segmentation is assessed using Dice coefficient and Hausdorff distance. In addition, we introduce a novel metric called Ray Feature error to evaluate our method. The results show that the proposal accurately discriminates cardiac tissue. Thus, it may be a useful tool for supporting heart disease diagnosis and tailoring treatments.

Region growing using superpixels with learned shape prior

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic, Sugimoto, A.
Publikace: Journal of Electronic Imaging. 2017, 26(6), ISSN 1017-9909.
Rok: 2017

DOI: 10.1117/1.JEI.26.6.061611
Odkaz: https://doi.org/10.1117/1.JEI.26.6.061611
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Region growing is a classical image segmentation method based on hierarchical region aggregation using local similarity rules. Our proposed method differs from classical region growing in three important aspects. First, it works on the level of superpixels instead of pixels, which leads to a substantial speed-up. Second, our method uses learned statistical shape properties that encourage plausible shapes. In particular, we use ray features to describe the object boundary. Third, our method can segment multiple objects and ensure that the segmentations do not overlap. The problem is represented as an energy minimization and is solved either greedily or iteratively using graph cuts. We demonstrate the performance of the proposed method and compare it with alternative approaches on the task of segmenting individual eggs in microscopy images of Drosophila ovaries.

Supervised and unsupervised segmentation using superpixels, model estimation, and Graph Cut

Autoři: Borovec, J., doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Habart, D.
Publikace: Journal of Electronic Imaging. 2017, 26(6), ISSN 1017-9909.
Rok: 2017

DOI: 10.1117/1.JEI.26.6.061610
Odkaz: https://doi.org/10.1117/1.JEI.26.6.061610
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Image segmentation is widely used as an initial phase of many image analysis tasks. It is often advantageous to first group pixels into compact, edge-respecting superpixels, because these reduce the size of the segmentation problem and thus the segmentation time by an order of magnitudes. In addition, features calculated from superpixel regions are more robust than features calculated from fixed pixel neighborhoods. We present a fast and general multiclass image segmentation method consisting of the following steps: (i) computation of superpixels; (ii) extraction of superpixel-based descriptors; (iii) calculating image-based class probabilities in a supervised or unsupervised manner; and (iv) regularized superpixel classification using graph cut. We apply this segmentation pipeline to five real-world medical imaging applications and compare the results with three baseline methods: pixelwise graph cut segmentation, supertexton-based segmentation, and classical superpixel-based segmentation. On all datasets, we outperform the baseline results. We also show that unsupervised segmentation is surprisingly efficient in many situations. Unsupervised segmentation provides similar results to the supervised method but does not require manually annotated training data, which is often expensive to obtain.

Automated Analysis of Microscopic Images of Isolated Pancreatic Islets

Autoři: Habart, D., doc. Ing. Jan Švihlík, Ph.D., Schier, J., Cahová, M., Girman, P., Zacharovová, K., Berková, Z., Kříž, J., Fabryová, E., Kosinová, L., Papáčková, Z., prof. Dr. Ing. Jan Kybic, Saudek, F.
Publikace: Cell Transplantation. 2016, 25(12), 2145-2156. ISSN 0963-6897.
Rok: 2016

DOI: 10.3727/096368916X692005
Odkaz: https://doi.org/10.3727/096368916X692005
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Clinical islet transplantation programs rely on the capacities of individual centers to quantify isolated islets. Current computer-assisted methods require input from human operators. Here we describe two machine learning algorithms for islet quantification: the trainable islet algorithm (TIA) and the nontrainable purity algorithm (NPA). These algorithms automatically segment pancreatic islets and exocrine tissue on microscopic images in order to count individual islets and calculate islet volume and purity. References for islet counts and volumes were generated by the fully manual segmentation (FMS) method, which was validated against the internal DNA standard. References for islet purity were generated via the expert visual assessment (EVA) method, which was validated against the FMS method. The TIA is intended to automatically evaluate micrographs of isolated islets from future donors after being trained on micrographs from a limited number of past donors. Its training ability was first evaluated on 46 images from four donors. The pixel-to-pixel comparison, binary statistics, and islet DNA concentration indicated that the TIA was successfully trained, regardless of the color differences of the original images. Next, the TIA trained on the four donors was validated on an additional 36 images from nine independent donors. The TIA was fast (67 s/image), correlated very well with the FMS method (R 2 = 1.00 and 0.92 for islet volume and islet count, respectively), and had small REs (0.06 and 0.07 for islet volume and islet count, respectively). Validation of the NPA against the EVA method using 70 images from 12 donors revealed that the NPA had a reasonable speed (69 s/image), had an acceptable RE (0.14), and correlated well with the EVA method (R 2 = 0.88). Our results demonstrate that a fully automated analysis of clinical-grade micrographs of isolated pancreatic islets is feasible. The algorithms described herein will be freely available as a Fiji platform plugin.

Automated separation of merged Langerhans islets

Autoři: doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Habart, D.
Publikace: Medical Imaging 2016: Image Processing. Bellingham: SPIE, 2016. SPIE. vol. 9784. ISSN 1605-7422. ISBN 978-1-5106-0019-5.
Rok: 2016

DOI: 10.1117/12.2216798
Odkaz: https://doi.org/10.1117/12.2216798
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
This paper deals with separation of merged Langerhans islets in segmentations in order to evaluate correct histogram of islet diameters. A distribution of islet diameters is useful for determining the feasibility of islet transplantation in diabetes. First, the merged islets at training segmentations are manually separated by medical experts. Based on the single islets, the merged islets are identified and the SVM classifier is trained on both classes (merged/single islets). The testing segmentations were over-segmented using watershed transform and the most probable back merging of islets were found using trained SVM classifier. Finally, the optimized segmentation is compared with ground truth segmentation (correctly separated islets).

Characterization of Hematologic Malignancies based on Discrete Orthogonal Moments

Autoři: Nava Velazco, U., Gonzalez, G., prof. Dr. Ing. Jan Kybic, Escalante-Ramirez, B.
Publikace: Image Processing Theory Tools and Applications. Piscataway: IEEE, 2016. ISSN 2154-512X. ISBN 978-1-4673-8910-5.
Rok: 2016

DOI: 10.1109/IPTA.2016.7821039
Odkaz: https://doi.org/10.1109/IPTA.2016.7821039
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
During the last decade leukemia and lymphomas have been a hot topic in the biomedical area. Their diagnosis is a time-consuming task that, in many cases, delays treatments. On the other hand, discrete orthogonal moments (DOMs) are a tool recently introduced in biomedical image analysis. Here, we propose a combination of DOMs to help in the diagnosis of leukemia and lymphomas. We classify the IICBU2008-lymphoma dataset that includes three hematologic malignancies: chronic lymphocytic leukemia, follicular lymphoma, and mantle cell lymphoma. Our methodology analyzes such diseases in the hematoxylin and eosin color space. We also include feature analysis to preserve the most discriminating characteristics of the malignant tissues. Finally, the classification of the samples is performed with kernel Fisher discriminant analysis. The accuracy is 93.85%. The results show the proposal could be useful in different biomedical applications.

Classification of tumor epithelium and stroma in colorectal cancer based on discrete tchebichef moments

Autoři: Nava Velazco, U., González, G., prof. Dr. Ing. Jan Kybic, Escalante-Ramírez, B.
Publikace: CLIP 2015: Proceedings of the 4th MICCAIA Workshop on Clinical Image-Based Procedures. Cham: Springer, 2016. p. 79-87. LNCS. vol. 9401. ISSN 0302-9743. ISBN 978-3-319-31807-3.
Rok: 2016

DOI: 10.1007/978-3-319-31808-0_10
Odkaz: https://doi.org/10.1007/978-3-319-31808-0_10
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Colorectal cancer is a major cause of mortality. As the disease progresses, adenomas and their surrounding tissue are modified. Therefore, a large number of samples from the epithelial cell layer and stroma must be collected and analyzed manually to estimate the potential evolution and stage of the disease. In this study, we propose a novel method for automatic classification of tumor epithelium and stroma in digitized tissue microarrays. To this end, we use discrete Tchebichef moments (DTMs) to characterize tumors based on their textural information. DTMs are able to capture image features in a non-redundant way providing a unique description. A support vector machine was trained to classify a dataset composed of 1376 tissue microarrays from 643 patients with colorectal cancer. The proposal achieved 97.62% of sensitivity and 95% of specificity showing the effectiveness of the methodology.

Comparison of volume estimation methods for pancreatic islet cells

Autoři: Dvořák, J., doc. Ing. Jan Švihlík, Ph.D., Habart, D., prof. Dr. Ing. Jan Kybic,
Publikace: Medical Imaging 2016: Biomedical Applications in Molecular, Structural, and Functional Imaging. Bellingham: SPIE, 2016. SPIE. vol. 9788. ISSN 1605-7422. ISBN 978-1-5106-0023-2.
Rok: 2016

DOI: 10.1117/12.2216783
Odkaz: https://doi.org/10.1117/12.2216783
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
In this contribution we study different methods of automatic volume estimation for pancreatic islets which can be used in the quality control step prior to the islet transplantation. The total islet volume is an important criterion in the quality control. Also, the individual islet volume distribution is interesting it has been indicated that smaller islets can be more effective. A 2D image of a microscopy slice containing the islets is acquired. The input of the volume estimation methods are segmented images of individual islets. The segmentation step is not discussed here. We consider simple methods of volume estimation assuming that the islets have spherical or ellipsoidal shape. We also consider a local stereological method, namely the nucleator. The nucleator does not rely on any shape assumptions and provides unbiased estimates if isotropic sections through the islets are observed. We present a simulation study comparing the performance of the volume estimation methods in different scenarios and an experimental study comparing the methods on a real dataset.

Fully automated classification of bone marrow infiltration in low-dose CT of patients with multiple myeloma based on probabilistic density model and supervised learning

Autoři: Martínez-Martínez, F., prof. Dr. Ing. Jan Kybic, Lambert, L., Meckova, Z.
Publikace: Computers in Biology and Medicine. 2016, 71 57-66. ISSN 0010-4825.
Rok: 2016

DOI: 10.1016/j.compbiomed.2016.02.001
Odkaz: https://doi.org/10.1016/j.compbiomed.2016.02.001
Pracoviště: Katedra kybernetiky
Anotace:
This paper presents a~fully-automated method for the identification of bone marrow infiltration in femurs in low-dose CT of patients with multiple myeloma. We automatically find the femurs and the bone marrow within them. In the next step, we create a~probabilistic, spatially-dependent density model of normal tissue. At test time, we detect unexpectedly high density voxels which may be related to bone marrow infiltration, as outliers to this model. Based on a~set of global, aggregated features representing all detections from one femur, we classify the subjects as being either healt hy or not. This method was validated on a~dataset of 127 subjects with ground truth created from a consensus of two expert radiologists, obtaining an AUC of 0.996 for the task of distinguishing healthy controls and patients with bone marrow infiltration. To the best of our knowledge, no other automatic image-based method for this task has been published before.

Texel-based image classification with orthogonal bases

Autoři: Carbajal-Degante, E., Nava Velazco, U., Olveres, J., Escalante-Ramírez, B., prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of SPIE - The International Society for Optical Engineering. Bellingham (stát Washington): SPIE, 2016. SPIE. vol. 9896. ISSN 1996-756X. ISBN 978-1-5106-0141-3.
Rok: 2016

DOI: 10.1117/12.2228694
Odkaz: https://doi.org/10.1117/12.2228694
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Periodic variations in patterns within a group of pixels provide important information about the surface of interest and can be used to identify objects or regions. Hence, a proper analysis can be applied to extract particular features according to some specific image properties. Recently, texture analysis using orthogonal polynomials has gained attention since polynomials characterize the pseudo-periodic behavior of textures through the projection of the pattern of interest over a group of kernel functions. However, the maximum polynomial order is often linked to the size of the texture, which implies in many cases, a complex calculation and introduces instability in higher orders leading to computational errors. In this paper, we address this issue and explore a pre-processing stage to compute the optimal size of the window of analysis called "texel." We propose Haralick-based metrics to find the main oscillation period, such that, it represents the fundamental texture and captures the minimum information, which is sufficient for classification tasks. This procedure avoids the computation of large polynomials and reduces substantially the feature space with small classification errors. Our proposal is also compared against different fixed-size windows. We also show similarities between full-image representations and the ones based on texels in terms of visual structures and feature vectors using two different orthogonal bases: Tchebichef and Hermite polynomials. Finally, we assess the performance of the proposal using well-known texture databases found in the literature.

Automatic detection of bone marrow infiltration by multiple myeloma detection in low-dose CT

Autoři: Martínez-Martínez, F., prof. Dr. Ing. Jan Kybic, Lambert, L.
Publikace: ICIP '15: Proceedings of the 2015 IEEE International Conference on Imag e Processing. Piscataway: IEEE, 2015. pp. 4813-4817. ISSN 1522-4880. ISBN 978-1-4799-8339-1.
Rok: 2015

DOI: 10.1109/ICIP.2015.7351721
Odkaz: https://doi.org/10.1109/ICIP.2015.7351721
Pracoviště: Katedra kybernetiky
Anotace:
Multiple myeloma is a disease primarily affecting bone marrow. This paper presents a novel method that can automatically detect infiltration of the bone marrow by multiple myeloma in diseased femurs (thigh bones) from low-dose CT images. This detection is done by evaluating two aspects of the CT images: bone marrow infiltrations (increased density values compared to regular fatty bone marrow) and scalloping (indentation of the inner margin of the cortical bone). First, bone marrow and cortical bone are automatically segmented from CT images. Afterwards, a probabilistic model of the bone marrow density is created in order to automatically detect islands of infiltrations. Finally, scalloping is detected by means of the quantification of the roughness of the boundary between the bone marrow and bony tissue. We have experimentally tested both infiltration and scalloping detection methods, obtaining a sensitivity of 74.9pct and 69.2pct and a specificity of 75pct and 61pct respectively.

Color normalization for robust evaluation of microscopy images

Autoři: doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Habart, D.
Publikace: Optics and Photonics 2015: Applications of Digital Image Processing XXXVIII. Bellingham: SPIE, 2015. ISSN 0277-786X. ISBN 978-1-62841-765-4.
Rok: 2015

DOI: 10.1117/12.2188236
Odkaz: https://doi.org/10.1117/12.2188236
Pracoviště: Katedra kybernetiky
Anotace:
This paper deals with color normalization of microscopy images of Langerhans islets in order to increase robustness of the islet segmentation to illumination changes. The main application is automatic quantitative evaluation of the islet parameters, useful for determining the feasibility of islet transplantation in diabetes. First, background illumination inhomogeneity is compensated and a preliminary foreground/background segmentation is performed. The color normalization itself is done in either $l alpha beta$ or logarithmic RGB color spaces, by comparison with a reference image. The color-normalized images are segmented using color-based features and pixel-wise logistic regression, trained on manually labeled images. Finally, relevant statistics such as the total islet area are evaluated in order to determine the success likelihood of the transplantation. The robustness of the evaluation is increased dramatically by the color normalization, decreasing the image-wise failure rate from 17pct to 0pct and the equal error rate from 30pct to 1pct.

Fast registration of segmented images by normal sampling

Autoři: prof. Dr. Ing. Jan Kybic, Dolejší, M., Borovec, J.
Publikace: CVPRW2015: IEEE Conference on Computer Vision and Pattern Recognition Workshops. Piscataway (New Jersey): IEEE, 2015. pp. 11-19. ISSN 2160-7508. ISBN 978-1-4673-6759-2.
Rok: 2015

DOI: 10.1109/CVPRW.2015.7301311
Odkaz: https://doi.org/10.1109/CVPRW.2015.7301311
Pracoviště: Katedra kybernetiky
Anotace:
It is known that image registration is mostly driven by image edges. We have taken this idea to the extreme. In segmented images, we ignore the interior of the components and focus on their boundaries only. Furthermore, by assuming spatial compactness of the components, the similarity criterion can be approximated by sampling only a small number of points on the normals passing through a sparse set of keypoints. This leads to an order-of-magnitude speed advantage in comparison with classical registration algorithms. Surprisingly, despite the crude approximation, the accuracy is comparable. By virtue of the segmentation and by using a suitable similarity criterion such as mutual information on labels, the method can handle large appearance differences and large variability in the segmentations. The segmentation does not need not be perfectly coherent between images and over-segmentation is acceptable. We demonstrate the performance of the method on a range of different datasets, including histological slices and Drosophila imaginal discs, using rigid transformations.

Feature Ensemble for Quantitative Analysis of Emphysema in CT imaging

Autoři: Nava Velazco, U., Olveres, J., prof. Dr. Ing. Jan Kybic, Escalante, B., Cristóbal, G.
Publikace: The 5th IEEE International Conference on E-Health and Bioengineering. Iasi: Gr. T. Popa University of Medicine and Pharmacy, 2015. ISBN 978-1-4673-7545-0.
Rok: 2015

DOI: 10.1109/EHB.2015.7391573
Odkaz: https://doi.org/10.1109/EHB.2015.7391573
Pracoviště: Algoritmy pro biomedicínské zobrazování
Anotace:
Chronic obstructive pulmonary disease is a non-reversible disorder characterized primarily by a dominant emphysema or bronchitis. Since early treatments can help to control the symptoms, the quantification of emphysema has become an important topic. Here, we introduce a novel procedure to quantify emphysematous lesions using an ensemble of features based on log-Gabor filters, mean difference technique, and intensity values. This set captures both spatial and frequency variations and provides a suitable description of lung tissue. Leave-patient-out cross-validation was employed on the computed tomography emphysema database to validate our proposal. The sensitivity and specificity achieved were 91.22% and 95.48%, respectively. Such results have demonstrated that the proposed methodology could assist in quantification of emphysema.

Geometrical Graph Matching using Monte Carlo Tree Search

Autoři: dos Santos Pinheiro, M., prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of the IEEE International Conference in Image Processing (ICIP). Los Alamitos: IEEE Computer Society Press, 2015. pp. 3145-3149. ISSN 1522-4880. ISBN 978-1-4799-8339-1.
Rok: 2015

DOI: 10.1109/ICIP.2015.7351383
Odkaz: https://doi.org/10.1109/ICIP.2015.7351383
Pracoviště: Katedra kybernetiky
Anotace:
Many medical images contain graph-like geometrical structures such as blood vessels and neuronal networks. We present an algorithm for matching geometrical graphs, in order to quickly and robustly align such images. We use a sampling-based curve descriptor to prune dissimilar edges. The matching is modeled as a single-player game, growing the matching from a random initial correspondence. The coarse global solution is found using a Monte Carlo Tree Search and then refined locally. We show experimentally that our approach finds the correct matching in all tested datasets and is the fastest of all global methods.

Non-Rigid Graph Registration using Active Testing Search

Autoři: Serradell, E., dos Santos Pinheiro, M., Sznitman, R., prof. Dr. Ing. Jan Kybic, Moreno-Noguer, F., Fua, P.
Publikace: IEEE Transactions on Pattern Analysis and Machine Intelligence. 2015, 37(3), 625-638. ISSN 0162-8828.
Rok: 2015

DOI: 10.1109/TPAMI.2014.2343235
Odkaz: https://doi.org/10.1109/TPAMI.2014.2343235
Pracoviště: Katedra kybernetiky
Anotace:
We present a new approach for matching sets of branching curvilinear structure s that form graphs embedded in $mathbb{R}^2$ or $mathbb{R}^3$ and may be subject to deformations. Unlike earlier method s, ours does not rely on local appearance similarity nor does require a good initial alignment. Furthermore, it can cope with non-linear deformatio ns, topological differences, and partial graphs. To handle arbitrary non-linear deformations, we use Gaussian Processes to represen t the geometrical mapping relating the two graphs. In the absence of appearance information, we iteratively establish correspondences between points, update the mapping accordingly, and use it to estimate where to find the most likely correspondences that will be used in the next step. To make the computation tractable for large graphs, the set of new potential matches consider ed at each iteration is not selected at random as in many RANSAC-based algorithms. Instead, we introduce a so-called Active Testin g Search strategy that performs a priority search to favor the most likely matches and speed-up the process. We demonstrat e the effectiveness of our approach first on synthetic cases and then on angiography data, retinal fundus images, and microsco py image stacks acquired at very different resolutions.

Supertexton-based segmentation in early Drosophila oogenesis

Autoři: Nava Velazco, U., prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of the IEEE International Conference in Image Processing (ICIP). Los Alamitos: IEEE Computer Society Press, 2015. pp. 2656-2659. ISSN 1522-4880. ISBN 978-1-4799-8339-1.
Rok: 2015

DOI: 10.1109/ICIP.2015.7351284
Odkaz: https://doi.org/10.1109/ICIP.2015.7351284
Pracoviště: Katedra kybernetiky
Anotace:
Studies concerning gene expression patterns of Drosophila are o f paramount importance in basic biological research because many genes are conserved across o rganisms providing information of fundamental activity. However, mapping a gene requires anal yzing hundreds of objects that have been segmented previously. Hence, a reliable segmentation is a crucial step. Here, we introduce the concept of supertextons and propose a novel segmen tation procedure for localized Drosophila ovaries. First, a pre-segmentation step is performe d using superpixels; each superpixel that belongs to a single class is transform into a featu re vector. Then, a dictionary is built by clustering representative feature vectors per class , such clusters are called supertextons. Finally, during the classification stage, new superp ixels are assigned to certain classes using the k-NN classifier and the supertexton dictionar y. This proposal has been applied to segmentation of cells in Drosophila oogenesis where the results have shown the effectiveness of our approach.

Automatic Analysis of Spatial Gene Expression Patterns

Autoři: Nava Velazco, U., Malinka, F., prof. Dr. Ing. Jan Kybic, prof. Ing. Filip Železný, Ph.D.,
Publikace: Proceedings of International Workshop Modeling in Engineering. Buštěhrad: UCEEB, 2014,
Rok: 2014

Pracoviště: Katedra kybernetiky, Katedra počítačů
Anotace:
Gene functionality is of paramount importance in basic biological research with possible therapeutic applications in medicine. Genes turn on in specific cells and at a specific time-point (gene expression pattern).

Automatic simultaneous segmentation and fast registration of histological images

Autoři: prof. Dr. Ing. Jan Kybic, Borovec, J.
Publikace: International Symposium on Biomedical Imaging (ISBI). Piscataway: IEEE, 2014. p. 774-777. ISBN 978-1-4673-1959-1.
Rok: 2014

DOI: 10.1109/ISBI.2014.6867985
Odkaz: https://doi.org/10.1109/ISBI.2014.6867985
Pracoviště: Katedra kybernetiky
Anotace:
We describe an automatic method for fast registration of images with very different appearances. The images are jointly segmented into a small number of classes, the segmented images are registered, and the process is repeated.

Classification of microscopy images of Langerhans islets

Autoři: doc. Ing. Jan Švihlík, Ph.D., prof. Dr. Ing. Jan Kybic, Habart, D., Berková, Z., Girman, P., Kříž, J., Zacharovová, K.
Publikace: SPIE Medical Imaging. Bellingham: SPIE, 2014. ISSN 1605-7422. ISBN 978-0-8194-9827-4.
Rok: 2014

DOI: 10.1117/12.2043621
Odkaz: https://doi.org/10.1117/12.2043621
Pracoviště: Katedra kybernetiky
Anotace:
Classification of images of Langerhans islet is crucial procedure for optimization of diabetes treatment. Hence, this paper deals with segmentation of microscopy images of Langerhans islets and evaluation of islet parameters such as area, islet diameter, islet equivalent (IE) etc. For all the available images, the ground truth and the islet parameters were independently evaluated by four medical experts in a blinded manner. We utilized linear classifier (perceptron algorithm) and SVM (support vector machine) for image segmentation. All the available image data were segmented and compared with corresponding ground truth. The islet parameters were also evaluated and compared with parameters evaluated by medical experts. The presented fully automatic algorithm analyzes the microscopy images as good as medical experts.

jSLIC : superpixels in ImageJ

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic,
Publikace: CVWW2014: Proceedings of the 19th Computer Vision Winter Workshop. Praha: Czech Society for Cybernetics and Informatics, 2014, pp. 14-18. ISBN 978-80-260-5641-6.
Rok: 2014

Pracoviště: Katedra kybernetiky
Anotace:
This paper presents the implementation and particular improvements on the superpixel clustering algorithm - SLIC (Simple Linear Iterative Clustering). The main contribution of the jSLIC is a significant speed-up of the original clustering method, transforming the compactness parameter such that the value is image independent, and a new post-processing step (after clustering) which now gives more reliable superpixels - the newly established segments are more homogeneous. The improvements of speed and quality are shown on real images. We implemented the new jSLIC in Java and made the source code publicly available. Also we created a plug-in in ImageJ/Fiji which is commonly used as a research and development platform in biology and medical imaging.

Neuromuscular fiber segmentation through particle filtering and discrete optimization

Autoři: Dietenbeck, T., Varray, F., prof. Dr. Ing. Jan Kybic, Basset, O., Cachard, C.
Publikace: SPIE Medical Imaging. Bellingham: SPIE, 2014. ISSN 1605-7422. ISBN 978-0-8194-9827-4.
Rok: 2014

DOI: 10.1117/12.2043257
Odkaz: https://doi.org/10.1117/12.2043257
Pracoviště: Katedra kybernetiky
Anotace:
We present an algorithm to segment a set of parallel, intertwined and bifurcatin g fibers from 3D images, targeted for the identification of neuronal fibers in very large sets of 3D confocal microscopy images. The method consists of preprocessing, local calcula tion of fiber probabilities, seed detection, tracking by particle filtering, global superv ised seed clustering and final voxel segmentation.

Organ-Focused Mutual Information for Nonrigid Multimodal Registration of Liver CT and Gd-EOB-DTPA-enhanced MRI

Autoři: Fernandez-de-Manuel, L., Wollny, G., prof. Dr. Ing. Jan Kybic, Jimenez-Carretero, D., Tellado, J.M., Ramon, E., Desco, M., Santos, A., Pascau, J., Ledesma-Carbayo, M.
Publikace: Medical Image Analysis. 2014, 18(1), 22-35. ISSN 1361-8415.
Rok: 2014

DOI: 10.1016/j.media.2013.09.002
Odkaz: https://doi.org/10.1016/j.media.2013.09.002
Pracoviště: Katedra kybernetiky
Anotace:
Accurate detection of liver lesions is of great importance in hepatic surgery planning. Recent studies have shown that the detection rate of liver lesions is significantly higher in gadoxetic acid-enhanced magnetic resonance imagin than in contrast-enhanced portal-phase computed tomography (CT); however, thelatter remains essential because of its high specificity, good performance in estimating liver volumes and better vessel visibility. To characterize liver lesions using both the above image modalities, we propose a multimodal nonrigid registration framework using organ-focused mutual information (OF-MI). This proposal tries to improve mutual information (MI) based registration by adding spatial information, benefiting from the availability of expert liver segmentation in clinical protocols.

Path Descriptors for Geometric Graph Matching and Registration

Autoři: dos Santos Pinheiro, M., prof. Dr. Ing. Jan Kybic,
Publikace: Image Analysis and Recognition: 11th International Conference (ICIAR 2014). Berlin: Springer-Verlag, 2014, pp. 3-11. Lecture Notes in Computer Science. ISSN 0302-9743. ISBN 978-3-319-11757-7. Available from: ftp://cmp.felk.cvut.cz/pub/cmp/articles/amavemig/Pinheiro-ICIAR2014.pdf
Rok: 2014

DOI: 10.1007/978-3-319-11758-4_1
Odkaz: https://doi.org/10.1007/978-3-319-11758-4_1
Pracoviště: Katedra kybernetiky
Anotace:
Graph and tree-like structures such as blood vessels and neuronal networks are abundant in medical imaging. We present a method to calculate path descriptors in geometrical graphs, so that the similarity between paths in the graphs can be determined efficiently. We show experimentally that our descriptors are more discriminative than existing alternatives. We further describe how to match two geometric graphs using our path descriptors. Our main application is registering images for which standard techniques are inefficient, because the appearance of the images is too different, or there is not enough texture and no uniquely identifiable keypoints to be found. We show that our approach can register these images with better accuracy than previous methods.

Reconstructing Evolving Tree Structures in Time Lapse Sequences

Autoři: Glowacki, P., dos Santos Pinheiro, M., Turetken, E., Sznitman, R., Lebrecht, D., prof. Dr. Ing. Jan Kybic, Hotlmaat, A., Fua, P.
Publikace: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Los Alamitos: IEEE Computer Society Press, 2014. pp. 3035-3042. ISSN 1063-6919. ISBN 978-1-4799-5117-8.
Rok: 2014

DOI: 10.1109/CVPR.2014.388
Odkaz: https://doi.org/10.1109/CVPR.2014.388
Pracoviště: Katedra kybernetiky
Anotace:
We propose an approach to reconstructing tree structures that evolve over time in 2D images and 3D image stacks such as neuronal axons or plant branches. Instead of reconstructing structures in each image independently, we do so for all images simultaneously to take advantage of temporal-consistency constraints. We show that this problem can be formulated as a Quadratic Mixed Integer Program and solved efficiently. The outcome of our approach is a framework that provides substantial improvements in reconstructions over traditional single time-instance formulations. Furthermore, an added benefit of our approach is the ability to automatically detect places where significant changes have occurred over time, which is challenging when considering large amounts of data.

Registration of segmented histological images using thin plate splines and belief propagation

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: SPIE Medical Imaging. Bellingham: SPIE, 2014. ISSN 1605-7422. ISBN 978-0-8194-9827-4.
Rok: 2014

DOI: 10.1117/12.2042317
Odkaz: https://doi.org/10.1117/12.2042317
Pracoviště: Katedra kybernetiky
Anotace:
We register images based on their multiclass segmentations, for cases when correspondence of local features cannot be established. A discrete mutual information is used as a similarity criterion. It is evaluated at a sparse set of location on the interfaces between classes. A thin-plate spline regularization is approximated by pairwise interactions. The problem is cast into a discrete setting and solved efficiently by belief propagation. Further speedup and robustness is provided by a multiresolution framework. Preliminary experiments suggest that our method can provide similar registration quality to standard methods at a fraction of the computational cost.

Active Testing Search for Point Cloud Matching

Autoři: Pinheiro, M., Sznitman, R., Serradell, E., prof. Dr. Ing. Jan Kybic, Moreno-Noguer, F., Fua, P.
Publikace: Information Processing in Medical Imaging. Heidelberg: Springer, 2013, pp. 572-583. ISSN 0302-9743. ISBN 978-3-642-38867-5.
Rok: 2013

Pracoviště: Katedra kybernetiky
Anotace:
We present a general approach for solving the~point-cloud matching problem for the~case of mildly nonlinear transformations. Our method quickly finds a coarse approximation of the solution by exploring a reduced set of partial matches using an approach to which we refer to as Active Testing Search (ATS). We apply the method to registration of graph structures by branching point matching. It is based solely on the geometric position of the points, no additional information is used nor the knowledge of an initial alignment. In the second stage, we use dynamic programming to refine the solution. We tested our algorithm on angiography, retinal fundus, and neuronal data gathered using electron and light microscopy. We show that our method solves cases not solved by most approaches, and is faster than the remaining ones.

Line Filtering for Surgical Tool Localization in 3D Ultrasound Images

Autoři: Uherčík, M., prof. Dr. Ing. Jan Kybic, Zhao, Y., Cachard, C., Liebgott, H.
Publikace: Computers in Biology and Medicine. 2013, 43(12), 2036-2045. ISSN 0010-4825.
Rok: 2013

DOI: 10.1016/j.compbiomed.2013.09.020
Odkaz: https://doi.org/10.1016/j.compbiomed.2013.09.020
Pracoviště: Katedra kybernetiky
Anotace:
We present a method for automatic surgical tool localization in 3D ultrasound images based on line filtering, voxel classification and model fitting. This could possibly provide assistance for biopsy needle or micro-electrode insertion, or a robotic system performing this insertion. The line-filtering method is first used to enhance the contrast of the 3D ultrasound image, then a classifier is chosen to separate the tool voxels, in order to reduce the number of outliers. The last step is Random Sample Consensus (RANSAC) model fitting. Experimental results on several different polyvinyl alcohol (PVA) cryogel data sets demonstrate that the failure rate of the method proposed herein is improved by at least 86% compared to the model-fitting RANSAC algorithm with axis accuracy better than 1 mm, at the expense of only a modest increase in computational effort. The results of this experiment show that this system could be useful for clinical applications.

Registration of multiple stained histological sections

Autoři: Borovec, J., prof. Dr. Ing. Jan Kybic, Bušta, M., Ortiz-de-Solorzano, C., Munoz-Barrutia, A.
Publikace: IEEE International Symposium on Biomedical Imaging: From Nano to Macro 2013 (ISBI2013). New York: IEEE, 2013. p. 1034-1037. ISSN 1945-7928. ISBN 978-1-4673-6454-6.
Rok: 2013

DOI: 10.1109/ISBI.2013.6556654
Odkaz: https://doi.org/10.1109/ISBI.2013.6556654
Pracoviště: Katedra kybernetiky
Anotace:
The analysis of protein-level multigene expression signature maps computed from the fusion of differently stained immunohistochemistry images is an emerging tool in cancer management. Creating these maps requires registering sets of histological images, a challenging task due to their large size, the non-linear distortions existing between consecutive sections and to the fact that the images correspond to different histological stains and thus, may have very different appearance. In this manuscript, we present a novel segmentation-based registration algorithm that exploits a multi-class pyramid and optimizes a fuzzy class assignment specially designed for this task. Compared to a standard non-rigid registration, the proposed method achieves an improved matching on both synthetic as well as real histological images of cancer lesions.

A Multi-Frequency Approach to Increase the Native Resolution of Ultrasound Images

Autoři: Varray, F., Cachard, C., prof. Dr. Ing. Jan Kybic, Novell, A., Bouakaz, A., Basset, O.
Publikace: EUSIPCO 2012: 20th European Signal Processing Conference. Piscataway: IEEE, 2012. p. 2733-2737. ISSN 2076-1465. ISBN 978-1-4673-1068-0.
Rok: 2012

Pracoviště: Katedra kybernetiky
Anotace:
Using large bandwidth transducer, such as cMUT, and the possibility to transmit several frequencies in one single transmission, we propose a multi-frequency scheme based on the frequency compounding (FC) technique to improve the native resolution of the US image and the DOF.

Approximate All Nearest Neighbor Search for High Dimensional Entropy Estimation for Image Registration

Autoři: prof. Dr. Ing. Jan Kybic, Vnučko, I.
Publikace: Signal Processing. 2012, 92(5), 1302-1316. ISSN 0165-1684.
Rok: 2012

DOI: 10.1016/j.sigpro.2011.11.027
Odkaz: https://doi.org/10.1016/j.sigpro.2011.11.027
Pracoviště: Katedra kybernetiky
Anotace:
Information theoretic criteria such as mutual information are often used as similarity measures for inter-modality image registration. We propose to use anearest neighbor based Kozachenko-Leonenko entropy estimator. Here we address the issue of determining a suitable all nearest neighbor (NN) search algorithm for this relatively specific task.

Automated segmentation of a motion mask to preserve sliding motion in deformable registration of thoracic CT

Autoři: Vandemeulebroucke, J., Bernard, O., Rit, S., prof. Dr. Ing. Jan Kybic, Clarysse, P., Sarrut, D.
Publikace: Medical Physics. 2012, 39(2), 1006-1015. ISSN 0094-2405.
Rok: 2012

DOI: 10.1118/1.3679009
Odkaz: https://doi.org/10.1118/1.3679009
Pracoviště: Katedra kybernetiky
Anotace:
Deformable registration generally relies on the assumption that the sought spatial transformation is smooth. Breathing motion involves sliding motion of the lung with respect to the chest wall. In the case of sliding motion, a discontinuity is present in the motion field and the smoothness assumption can lead to poor matching accuracy. Many authors have proposed alternative registration methods to preserve sliding motion, several of which rely on prior segmentations. We focus on a particular, subanatomical segmentation, called a motion mask, because it is advanta- geous for subsequent registration. The motion mask separates moving from less-moving regions, conveniently allowing to simultaneously estimate the motion for similarly moving tissue. We propose an original method for automatically extracting a motion mask from a CT image of the thorax. The obtained segmentation is useful for any registration method relying on a prior segmentation to account for sliding motion. The method is based on the level set framework, which allows to include geometric priors in the definition of the motion mask. To improve robustness, the original images are simplified and only clear anatomical features are retained, with respect to which the segmentation is defined. The resulting procedure comes down to a monitored level set segmentation of binary images. The method is applied to six inhale-exhale image pairs, and produced satisfying results for all patients, consistent with respect to patient anatomy. We show that the obtained motion masks can facilitate deformable registration of the thorax. By preserving the sliding motion, the complexity of the spatial transformation can be reduced considerably while maintaining matching accuracy.

Neuromuscular fiber segmentation using particle filtering and discrete optimization

Autoři: Varray, F., prof. Dr. Ing. Jan Kybic, Basset, O., Cachard, C.
Publikace: MICCAI 2012: Histopathology Image Analysis workshop. 2012, pp. 48-59.
Rok: 2012

Pracoviště: Katedra kybernetiky
Anotace:
'We present an algorithm to segment a set of parallel, intertwined and bifurcating fibers from 3D images, targeted for identification of neuronal fibers in very large sets of 3D confocal microscopy images. The method consists of preprocessing, local calculation of fiber probabilities, seed detection, local tracking by particle filtering, global supervised seed clustering, and final voxel segmentation. The preprocessing uses a novel random local probability filtering segmentation. The global segmentation is solved by discrete optimization. The combination of global and local approaches makes the segmentation robust, yet the individual data blocks can be processed sequentially, limiting memory consumption. The method is automatic but efficient manual interaction is possible if needed. Initial promising results on a neuromuscular projection fiber dataset as well as on simulated data are presented.

Robust elastic 2D/3D geometric graph matching

Autoři: Serradell, E., prof. Dr. Ing. Jan Kybic, Moreno-Noguer, F., Fua, P.
Publikace: SPIE: Medical Imaging 2012. Bellingham: SPIE, 2012. pp. 1-8. Proceedings of SPIE. ISSN 0277-786X. ISBN 978-0-8194-8963-0.
Rok: 2012

DOI: 10.1117/12.910573
Odkaz: https://doi.org/10.1117/12.910573
Pracoviště: Katedra kybernetiky
Anotace:
We present an algorithm for geometric matching of graphs embedded in 2D or 3D space. It is applicable for registering any graph-like structures appearing in biomedical images, such as blood vessels, pulmonary bronchi, nerve fibers, or dendritic arbors. Our approach does not rely on the similarity of local appearance features, so it is suitable for multimodal registration with a large difference in appearance. Unlike earlier methods, the algorithm uses edge shape, does not require an initial pose estimate, can handle partial matches, and can cope with nonlinear deformations and topological differences

Robust Non-Rigid Registration of 2D and 3D Graphs

Autoři: Serradell, E., Glowacki, P., prof. Dr. Ing. Jan Kybic, Moreno-Noguer, F., Fua, P.
Publikace: CVPR 2012: Proceedings of the 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. New York: IEEE Computer Society Press, 2012, pp. 996-1003. ISSN 1063-6919. ISBN 978-1-4673-1228-8.
Rok: 2012

DOI: 10.1109/CVPR.2012.6247776
Odkaz: https://doi.org/10.1109/CVPR.2012.6247776
Pracoviště: Katedra kybernetiky
Anotace:
'We present a new approach to matching graphs embedded in R2 or R3. Unlike earlier methods, our approach does not rely on the similarity of local appearance features, does not require an initial alignment, can handle partial matches, and can cope with non-linear deformations and topological differences.',NULL); keyword('registration');

Automatic Colposcopy Video Tissue Classification Using Higher Order Entropy Based Registration

Autoři: Garcia Arteaga, J., prof. Dr. Ing. Jan Kybic, Li, W.
Publikace: Computers in Biology and Medicine. 2011, 41(10), 960-970. ISSN 0010-4825.
Rok: 2011

DOI: 10.1016/j.compbiomed.2011.07.010
Odkaz: https://doi.org/10.1016/j.compbiomed.2011.07.010
Pracoviště: Katedra kybernetiky
Anotace:
Colposcopy is a well-established method to detect and diagnose intraepithelial lesions and uterine cervical cancer in early stages.During the exam color and texture changes are induced by the application of a contrast agent.Our aim is to densely quantify the change in the acetowhite decay level for a sequence of images captured during a colposcopy examination.

Bootstrap Optical Flow Confidence and Uncertainty Measure

Autoři: prof. Dr. Ing. Jan Kybic, Nieuwenhuis, C.
Publikace: Computer Vision and Image Understanding. 2011, 115(10), 1449-1462. ISSN 1077-3142.
Rok: 2011

DOI: 10.1016/j.cviu.2011.06.008
Odkaz: https://doi.org/10.1016/j.cviu.2011.06.008
Pracoviště: Katedra kybernetiky
Anotace:
We address the problem of estimating the uncertainty of optical flow algorithm results. Our method estimates the error magnitude at all points in the image. It can be used as a confidence measure. It is based on bootstrap resampling, which is a computational statistical inference technique based on repeating the optical flow calculation several times for different randomly chosen subsets of pixel contributions. As few as 10 repetitions are enough to obtain useful estimates of geometrical and angular errors. We use the combined local global optical flow method (CLG) which generalizes both Lucas-Kanade and Horn-Schunck type methods. However, the bootstrap method is very general and can be applied to almost any optical flow algorithm that can be formulated as a minimization problem. We show experimentally on synthetic as well as real video sequences with known ground truth that the bootstrap method performs better than all other confidence measures tested.

Spatio-Temporal Motion Estimation for Respiratory-Correlated Imaging of the Lungs

Autoři: Vandemeulebroucke, J., Rit, S., prof. Dr. Ing. Jan Kybic, Clarysse, P., Sarrut, D.
Publikace: Medical Physics. 2011, 38(1), 166-178. ISSN 0094-2405.
Rok: 2011

DOI: 10.1118/1.3523619
Odkaz: https://doi.org/10.1118/1.3523619
Pracoviště: Katedra kybernetiky
Anotace:
We explore the benefit of a spatio-temporal approach for deformable registration of respiratory-correlated imaging of the thorax. Our goal is obtain a more compact, restrictive parametrization of the deformation model in an attempt to reduce sen sitivity to artifacts and noise.

3D Ultrasound real-time monitoring of surgical tools

Autoři: Gaufillet, F., Liebgott, H., Uherčík, M., Cervenansky, F., prof. Dr. Ing. Jan Kybic, Cachard, C.
Publikace: IEEE Ultrasonics Symposium. Piscataway: IEEE, 2010, pp. 2360-2363. ISSN 1948-5719. ISBN 978-1-4577-0382-9.
Rok: 2010

DOI: 10.1109/ULTSYM.2010.5935783
Odkaz: https://doi.org/10.1109/ULTSYM.2010.5935783
Pracoviště: Katedra kybernetiky
Anotace:
We propose to assist the surgeon by providing a real-time application that takes the 3D data from the ultrasound scanner, automatically localizes the surgical tool and displays the section plane containing the tool. The axis of the tool is estimated with a method based on model fitting using a Random Sample Consensus (RANSAC)

Bootstrap Resampling for Image Registration Uncertainty Estimation without Ground Truth

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: IEEE Transactions on Image Processing. 2010, 19(1), 64-73. ISSN 1057-7149.
Rok: 2010

DOI: 10.1109/TIP.2009.2030955
Odkaz: https://doi.org/10.1109/TIP.2009.2030955
Pracoviště: Katedra kybernetiky
Anotace:
We address the problem of estimating the uncertainty of pixel based image registration algorithms, given just the two images to be registered, for cases when no ground truth data is available. Our novel method uses bootstrap resampling. It is very general, applicable to almost any registration method based on minimizing a pixel-based similarity criterion; we demonstrate it using the SSD, SAD, correlation, and mutual information criteria. We show experimentally that the bootstrap method provides better estimates of the registration accuracy than the state-of-the-art Cramer-Rao bound method. Additionally, we evaluate also a fast registration accuracy estimation (FRAE) method which is based on quadratic sensitivity analysis ideas and has a negligible computational overhead. FRAE mostly works better than the Cramer-Rao bound method but is outperformed by the bootstrap method.

Model Fitting Using RANSAC for Surgical Tool Localization in 3-D Ultrasound Images

Autoři: Uherčík, M., prof. Dr. Ing. Jan Kybic, Liebgott, H., Cachard, C.
Publikace: IEEE Transactions on Biomedical Engineering. 2010, 57(8), 1907-1916. ISSN 0018-9294.
Rok: 2010

DOI: 10.1109/TBME.2010.2046416
Odkaz: https://doi.org/10.1109/TBME.2010.2046416
Pracoviště: Katedra kybernetiky
Anotace:
Ultrasound guidance is used for many surgical interventions such as biopsy and electrode insertion. We present a~method to localize a thin surgical tool such as a biopsy needle or a micro-electrode in a 3D ultrasound image. The proposed method starts with thresholding and model fitting using RANSAC for robust localization of the axis. Subsequent local optimization refines its position. Two different tool image models are presented, one simple and fast, the second using learned a priori information on the tool's voxel intensities and the background. Finally, the tip of the tool is localized by finding an intensity drop along the axis. The simulation study shows that our algorithm can localize the tool at nearly real-time speed, even using a MATLAB implementation, with accuracy better than 1mm. In an experimental comparison to several alternative localization methods, our method appears to be the fastest and the most robust one.

Non-rigid consistent registration of 2D image sequences

Autoři: Arganda-Carreras, I., Sorzano, C.O.S., Thévenaz, P., Munoz-Barutia, A., prof. Dr. Ing. Jan Kybic, Marabini, R., Carazo, J.M., Ortiz-de-Solorzano, C.
Publikace: Physics in Medicine and Biology. 2010, 55(20), 6215-6242. ISSN 0031-9155.
Rok: 2010

DOI: 10.1088/0031-9155/55/20/012
Odkaz: https://doi.org/10.1088/0031-9155/55/20/012
Pracoviště: Katedra kybernetiky
Anotace:
We present a novel algorithm for the registration of 2D image sequences that combines the principles of multiresolution B-spline based elastic registration and those of bidirectional consistent registration. In our method, consecutive triples of images are iteratively registered to gradually extend through the set of images the information of the whole sequence. The intermediate results are reused as initialization of the corresponding next triple registration. We interpolate the images and model the deformation fields using B-spline multiresolution pyramids.

Semi-automated segmentation of Symptomatic Exudate-Associated Derangements (SEADs) in 3D OCT using layer segmentation

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic, Abramoff, M., Šonka, M.
Publikace: Analysis of Biomedical Signals and Images - Proceedings of Biosignal 2010. Brno: VUTIUM Press, 2010. pp. 1-6. ISSN 1211-412X. ISBN 978-80-214-4105-7.
Rok: 2010

Pracoviště: Katedra kybernetiky
Anotace:
We present a novel two step method to segment retinal pathologies related to the so called wet age-related macular degeneration from 3D spectral optical coherence tomography images. In the first step we segment three retinal layers by an optimal surface algorithm. The identified layers are used in the second step to constrain the segmentation of fluid filled retinal regions using GraphCut. We propose a new regularization energy term for GraphCut, that permits long range effect of the manual initialization

Discrete curvature calculation for fast level set segmentation

Autoři: prof. Dr. Ing. Jan Kybic, Krátký, J.
Publikace: ICIP: International Conference on Image Processing. Piscataway: IEEE, 2009. p. 2981-2984. ISBN 978-1-4244-5653-6.
Rok: 2009

Pracoviště: Katedra kybernetiky
Anotace:
Fast level set methods replace continuous PDEs by a discrete formulation, improving the execution times. The regularization in fast level set methods was so far handled indirectly via level set function smoothing. We propose to incorporate standard curvature based regularization into fast level set methods and address the problem of efficiently estimating local curvature of a discretized interface in 2D or 3D based on local partial volume.

Line filtering for detection of microtools in 3D ultrasound data

Autoři: Uherčík, M., prof. Dr. Ing. Jan Kybic, Cachard, C., Liebgott, H.
Publikace: Ultrasonics Symposium (IUS), 2009 IEEE International. Piscataway: IEEE Ultrasonics, Ferroelectrics, and Frequency Control Society, 2009. pp. 594-597. ISSN 1948-5719. ISBN 978-1-4244-4390-1.
Rok: 2009

DOI: 10.1109/ULTSYM.2009.5441538
Odkaz: https://doi.org/10.1109/ULTSYM.2009.5441538
Pracoviště: Katedra kybernetiky
Anotace:
We propose a robust method for localization of elongated surgical tools in 3D ultrasound data based on shape analysis. A new model of a surgical tool appearance in 3D ultrasound image is proposed which exploits its tubular shape. The tool axis is estimated with robust model fitting using a randomized RANSAC procedure. The tool model requires the voxels close to the axis to have a high intensity, high tubularness, and the local principal directions to be consistent with the tool axis. The visual contrast of the tool can be enhanced four-fold using line filtering. We demonstrate that classification rate is improved by 25-40% when adding the tubularness attribute. The comparison to other state-of-the-art localization methods shows that the proposed method is the most robust for data with high level of noise at the expense of additional time for pre-processing (less than 10 seconds for volume of size 53x71x260 voxels).

Respiratory Motion Estimation from Cone-Beam Projections Using a Prior Model

Autoři: Vandemeulebroucke, J., prof. Dr. Ing. Jan Kybic, Clarysse, P., Sarrut, D.
Publikace: Medical Image Computing and Computer-Assisted Intervention - MICCAI 2009. Heidelberg: Springer, 2009. p. 365-372. Lecture Notes in Computer Science. ISSN 0302-9743. ISBN 978-3-642-04270-6.
Rok: 2009

Pracoviště: Katedra kybernetiky
Anotace:
A method for estimating 3D breathing motion from a 2D+T cone-beam projection sequence by making use of a prior model and temporal regularisation, keywords = radiotherapy, lung cancer, breathing motion, deformable registration, cone-beam computed tomography.

The Lung TIME-Annotated Lung Nodule Dataset and Nodule Detection Framework

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic, Polovinčák, M., Tůma, S.
Publikace: Proceedings of SPIE. Bellingham: SPIE, 2009. p. 1-8. Medical Imaging 2009: Computer-Aided Diagnosis. ISBN 978-0-8194-7511-4.
Rok: 2009

Pracoviště: Katedra kybernetiky
Anotace:
The Lung Test Images from Motol Environment (Lung TIME) is a new publicly available dataset of thoracic CT scans with manually annotated pulmonary nodules. It is larger than other publicly available datasets. Pulmonary nodules are lesions in the lungs, which may indicate lung cancer. Their early detection significantly improves survival rate of patients. Automatic nodule detecting systems using CT scans are being developed to reduce physicians' load and to improve detection quality. Besides presenting our own nodule detection system, in this article, we mainly address the problem of testing and comparison of automatic nodule detection methods. Our publicly available 157 CT scan dataset with 394 annotated nodules contains almost every nodule types (pleura attached, vessel attached, solitary, regular, irregular) with 2-10mm in diameter, except ground glass opacities (GGO). Annotation was done consensually by two experienced radiologists.

2-D Locally Regularized Tissue Strain Estimation From Radio-Frequency Ultrasound Images: Theoretical Developments and Results on Experimental Data

Autoři: Brusseau, E., prof. Dr. Ing. Jan Kybic, Déprez, J., Basset, O.
Publikace: IEEE Transactions on Medical Imaging. 2008, 27(2), 145-160. ISSN 0278-0062.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
2-D Locally Regularized Tissue Strain Estimation From Radio-Frequency Ultrasound Images: Theoretical Developments and Results on Experimental Data

Algoritmy CAD systému pro CT vyšetření plic na plicní uzly

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic, Polovinčák, M., Tůma, S.
Publikace: Radiační zátěž pacientů při CT vyšetření a problematika jejího stanovení. Praha: 2. lékařská fakulta Univerzity Karlovy, 2008. p. 29-31.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
Popis dvoustupňového algoritmu pro detekci plicních uzlů. V prvním kroku jsou pomocí prostorových filtrů detekováni kandidáti, v druhém kroku jsou z kandidátů vybrány skutečné uzly.

Estimating Respiratory Motion from Cone-Beam Projections

Autoři: Vandemeulebroucke, J., Clarysse, P., prof. Dr. Ing. Jan Kybic, Sarrut, D.
Publikace: Proceedings of the First International Workshop on Pulmonary Image Analysis. Morrisville: Lulu.com, 2008, pp. 83-92. ISBN 978-1-4357-5952-7.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
We propose a method for estimating breathing motion from cone-beam projections by making use of a prior patient-specific model.

Fast no ground truth image registration accuracy evaluation: Comparison of bootstrap and Hessian approaches

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of 2008 IEEE International Symposium on Biomedical Imaging: From Nano to Macro. New York: IEEE, 2008. p. 792-795. ISSN 1945-7928. ISBN 978-1-4244-2002-5.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
Fast no ground truth image registration accuracy evaluation: Comparison of bootstrap and Hessian approaches.

Multi-resolution Parallel Integral Projection for Fast Localization of a Straight Electro de in 3D Ultrasound Images

Autoři: Uherčík, M., prof. Dr. Ing. Jan Kybic, Liebgott, H., Cachard, C.
Publikace: Proceedings of 2008 IEEE International Symposium on Biomedical Imaging: From Nano to Macro. New York: IEEE, 2008. p. 33-36. ISSN 1945-7928. ISBN 978-1-4244-2002-5.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
We address the problem of fast and accurate localization of miniature surgical instrument s like needles or electrodes using 3D ultrasound (US). An algorithm based on maximizing a~Parallel Integra l Transform (PIP) can automatically localize line-shaped objects in 3D US images with accuracy on the orde r of hundreds of micrometers. Here we propose to use a multi-resolution to accelerate the algorithm signif icantly. We use a maximum function for downsampling to preserve the high intensity voxels of a~thin electr ode. We integrate the multi-resolution pyramid into a~hierarchical mesh-grid search of PIP. The experiment s with a~tissue mimicking phantom and breast biopsy data show that proposed method works well on real US i mages. The speed-up is threefold compared to original PIP method with the same accuracy 0.4~mm. A~further speed-up up to 16 times is reached by an~early stopping of the optimization, at the expense of some loss of accuracy.

Opacity Quantification In Cardiac Angiogram Sequences

Autoři: Kazmar, T., prof. Dr. Ing. Jan Kybic,
Publikace: Analysis of Biomedical Signals and Images; Biosignal 2008 proceedings. Brno: VUTIUM Press, 2008. ISSN 1211-412X. ISBN 978-80-214-3613-8.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
The level of perfusion is routinely analyzed from the level of opacity in cardiac Xray angiograms. We propose an image enhancement method for angiogram sequences by motion compensation and background subtraction. Moreover, we extract a time-series describing the opacification in a given region-of-interest to enable opacity quantification.

Parallel Integral Projection Transform for Straight Electrode Localization in 3-D Ultrasound Images

Autoři: Barva, M., Uherčík, M., Mari, J., prof. Dr. Ing. Jan Kybic, Duhamel, J., Liebgott, H., Hlaváč, V., Cachard, C.
Publikace: IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control (UFFC). 2008, 55(7), 1559-1569. ISSN 0885-3010.
Rok: 2008

DOI: 10.1109/TUFFC.2008.833
Odkaz: https://doi.org/10.1109/TUFFC.2008.833
Pracoviště: Katedra kybernetiky
Anotace:
In surgical practice small metallic instruments are frequently used to perform various tasks inside human body. We address the problem of their accurate localization in the tissue. Recent experiments using medical ultrasound have shown that this modality is suitable for real-time visualization of anatomical structures as well as the position of surgical instruments. We propose an image processing algorithm that permits to automatically estimate the position of a line-segment shaped object. This method was applied to the localization of a thin metallic electrode in biological tissue. We show that the electrode axis can be found through maximizing the Parallel Integral Projection transform which is a form of the Radon transform. To accelerate this step hierarchical mesh-grid algorithm is implemented. Once the axis position is known, localization of the electrode tip is performed. The method was tested on simulated images, on ultrasound images of a tissue mimicking phantom containing a me

Přesnost počítačem asistované CT diagnostiky plicních uzlů ve srovnání s klasickými postupy hodnocení

Autoři: Tůma, S., prof. Dr. Ing. Jan Kybic, Dolejší, M., Polovinčák, J., Neuwirth, J., Adla, T., Čumlivská, E.
Publikace: Česko-slovenská pediatrie. 2008, 63(7/8), 31-52. ISSN 0069-2328.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
Accuracy of computer assisted CT diagnosis of lung nodules.

Reducing false positive responses in lung nodule detector system by asymmetric AdaBoost

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic, Polovinčák, M., Tůma, S.
Publikace: Proceedings of 2008 IEEE International Symposium on Biomedical Imaging: From Nano to Macro. New York: IEEE, 2008. pp. 656-659. ISSN 1945-7928. ISBN 978-1-4244-2002-5.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
We are developing a complex computer aided diagnosis (CAD) system to detect small pulmonary nodules from helical CT scans. Here we present a classifier to reduce the number of false positive responses of the primary detector. Our approach is based on an asymmetric Adaboost which enables us to give different weights to missed nodules (false negatives, FNs) and incorrectly detected structures (false positives, FPs). This is useful because there are noticeably more negative examples in the nodule candidate set than real nodules-true positives (TPs). The whole system is meant as a second opinion for a human radiologist to speed up reading the examination. That is why we should detect as many true nodules as possible, while a certain number of FPs is acceptable. The system was tested on 147 cases (36559 slices) containing 357 nodules marked by an expert radiologist. The new classifier significantly reduced the number of false positives, while only a few nodules were incorrectly omitted.

Regional Image Similarity Criteria Based on the Kozachenko-Leonenko Entropy Estimator

Autoři: Garcia Arteaga, J., prof. Dr. Ing. Jan Kybic,
Publikace: CVPRW: Computer Vision and Pattern Recognition Workshops. Piscataway: IEEE, 2008. pp. 542-549. ISSN 1063-6919. ISBN 978-1-4244-2339-2.
Rok: 2008

DOI: 10.1109/CVPRW.2008.4563022
Odkaz: https://doi.org/10.1109/CVPRW.2008.4563022
Pracoviště: Katedra kybernetiky
Anotace:
Regional Image Similarity Criteria Based on the Kozachenko-Leonenko Entropy Estimator

Three-dimensional segmentation of bones from CT and MRI using fast level sets

Autoři: Krátký, J., prof. Dr. Ing. Jan Kybic,
Publikace: SPIE MI 2008: Proceedings of the SPIE Medical Imaging 2008 Conference. Washington: SPIE, 2008. ISSN 0277-786X. ISBN 978-0-8194-7098-0.
Rok: 2008

Pracoviště: Katedra kybernetiky
Anotace:
Our task is to segment bones from 3D CT and MRI images. The main application is creation of 3D mesh models for finite element modeling. These surface and volume vector models can be used for further biomechanical processing and analysis. We selected a novel fast level set method because of its high computational efficiency, while preserving all advantages of traditional level set methods. Unlike in traditional level set methods, we are not solving partial differential equations (PDEs). Instead, the contours are represeted by two sets of points, corresponding to the inner and outer edge of the object boundary. We have extended the original implementation in 3D, where the speed advantage over classical level set segmentation are even more pronounced. We can segment a CT image of 512x512x125 in less than 20s by this method. It is approximately two orders of magnitude faster than standard narrow band algorithms. Our experiments with real 3D CT and MRI images presented in this paper showed

Automatic landmark detection for cervical image registration validation

Autoři: Garcia Arteaga, J., prof. Dr. Ing. Jan Kybic,
Publikace: SPIE 2007, Medical Imaging 2007: Computer-Aided Diagnosis. Bellingham: SPIE, 2007. p. 54-63. ISBN 978-0-8194-6632-7.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Many cervical Computer-Aided Diagnosis (CAD) methods rely on measuring gradual appearance changes on the cervix after the application of a contrast agent. Image registration has been used to ensure pixel correspondence to the same tissue location throughout the whole temporal sequence but, to date, there is no reliable mean of testing its accuracy to compensate for patient and tissue movement. We present an independent system to use automatically extracted and matched features from a colposcopic image sequence in order to generate position landmarks. These landmarks may be used either to measure the accuracy of a registration method to align any pair of images from the colposcopic sequence or as a cue for registration. The algorithm selects sets of matched features that extend through the whole image sequence allowing to locate, in a reliable and unbiased way, a tissue point throughout the whole image sequence.

Automatic two-step detection of pulmonary nodules

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic,
Publikace: SPIE 2007, Medical Imaging 2007: Computer-Aided Diagnosis. Bellingham: SPIE, 2007. p. 1-12. ISBN 978-0-8194-6632-7.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
We present a computer-aided diagnosis (CAD) system to detect small-size (from 2mm to around 10mm) pulmonary nodules from helical CT scans. A pulmonary nodule is a small, round (parenchymal nodule) or worm (juxta-pleural) shaped lesion in the lungs. Both have greater radio density than lungs parenchyma. Lung nodules may indicate a lung cancer and its detection in early stage improves survival rate of patients. CT is considered to be the most accurate imaging modality for detection of nodules. However, the large amount of data per examination makes the interpretation difficult. This leads to omission of nodules by human radiologist. CAD system presented is designed to help lower the number of omissions. Our system uses two different schemes to locate juxtapleural nodules and parenchymal nodules. For juxtapleural nodules, morphological closing and thresholding is used to find nodule candidates. To locate non-pleural nodule candidates, 3D blob detector uses multiscale filtration. Ellipsoid

Continuous criterion for parallel MRI reconstruction using B-spline approximation (PROBER)

Autoři: Petr, J., prof. Dr. Ing. Jan Kybic,
Publikace: SPIE 2007, Medical Imaging 2007: Image Processing. Bellingham: SPIE, 2007. pp. 43-53. ISBN 978-0-8194-6630-3.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Parallel MRI is a way to use multiple receiver coils with distinct spatial sensitivities to increase the speed of the MRI acquisition. The acquisition is speeded up by undersampling in the phase-encoding direction and the resulting data loss and aliasing is compensated for by the use of the additional information obtained from several receiver coils. The task is to reconstruct an unaliased image from a series of aliased images. We have proposed an algorithm called PROBER that takes advantage of the smoothness of the reconstruction transformation in space. B-spline functions are used to approximate the reconstruction transformation. Their coefficients are estimated at once minimizing the total expected reconstruction error. This makes the reconstruction less sensitive to noise in the reference images and areas without signal in the image. We show that this approach outperforms the SENSE and GRAPPA reconstruction methods for certain coil configurations. In this article, we propose anothe

Cortical mapping by Laplace-Cauchy transmission using a boundary element method

Autoři: Clerc, M., prof. Dr. Ing. Jan Kybic,
Publikace: Journal on Inverse Problems. 2007, 23(6), 2589-2601. ISSN 0266-5611.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
The Laplace-Cauchy problem of propagating Dirichlet and Neumann data from a portion to the rest of the boundary is an ill-posed inverse problem. Many regularizing algorithms have been recently proposed, in order to stabilize the solution with respect to noisy or incomplete data. Our main application is in electro-encephalography (EEG) where potential measurements available at part of the scalp are used to reconstruct the potential and the current on the inner skull surface. This problem, known as cortical mapping, and other applications - in fields such as nondestructive testing, or biomedical engineering - require to solve the problem in realistic, three-dimensional geometry. The goal of this article is to present a new boundary element based method for solving the Laplace-Cauchy problem in three dimensions, in a multilayer geometry. We validate the method experimentally on simulated data.

Geometric and Information Constraints for Automatic Landmark Selection in Colposcopy Sequences

Autoři: Garcia Arteaga, J., prof. Dr. Ing. Jan Kybic, Gu, J., Li, W.
Publikace: VISAPP 2007: Proceedings of the Second International Conference on Computer Vision Theory and Applications. Setúbal: INSTICC Press, 2007. pp. 333-338. ISBN 978-972-8865-73-3.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Colposcopy is a diagnostic method to visually detect cancerous and pre-cancerous tissue regions in the uterine cervix. A typical result is a sequence of cervical images captured at different times after the application of a contrast agent that must be spatially registered to compensate for patient, camera and tissue movement and on which progressive color and texture changes may be seen. We present a method to automatically select correct landmarks for non-consecutive sequence frames captured at long time intervals from a group of candidate matches. Candidate matches are extracted by detecting and matching feature points in consecutive images. Selection is based on geometrical constraints and a local rigid registration using Mutual Information.

High-Dimensional Entropy Estimation for Finite Accuracy Data: R-NN Entropy Estimator

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: IPMI2007: Information Processing in Medical Imaging, 20th International Conference. Heidelberg: Springer, 2007. p. 569-580. ISBN 3-540-73272-1.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
High-Dimensional Entropy Estimation for Finite Accuracy Data: R-NN Entropy Estimator.

Image Processing, Analysis and Machine Vision - A MATLAB Companion

Autoři: prof. Ing. Tomáš Svoboda, Ph.D., prof. Dr. Ing. Jan Kybic, Hlaváč, V.
Publikace: Toronto: Thomson, 2007. ISBN 0-495-29595-7.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Comprehensive textbook about Matlab programming in image processing and computer vision

Localization of surgical instruments in biological tissue from 3D ultrasound images

Autoři: Barva, M., prof. Dr. Ing. Jan Kybic, Hlaváč, V.
Publikace: Proceedings of Workshop 2007. Praha: České vysoké učení technické v Praze, 2007, pp. 480-481. ISBN 978-80-01-03667-9.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Localization of surgical instruments in biological tissue from 3D ultrasound images

Možnost snižovánídávky ionizujícího zářenípři vyšetřenínádorových onemocněníplic u dětía dorostu

Autoři: Tůma, S., Neuwirth, J., Dolejší, M., prof. Dr. Ing. Jan Kybic, Daníčková, K., Polovinčák, M., Šanda, J., Čumlivská, E., Mališ, J.
Publikace: Čes.-slov. Pediatrie. Praha: Česká lékařská společnost J. E. Purkyně, 2007, pp. 357-358. ISSN 0069-2328.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Reduction of dose in CT acquisition for lung tumor detection.

Neural mass model parameter identification for MEG/EEG

Autoři: prof. Dr. Ing. Jan Kybic, Faugeras, O., Clerc, M., Papadopoulo, T.
Publikace: SPIE 2007, Medical Imaging 2007: Physiology, Function, and Structure from Medical Imaging. Bellingham: SPIE, 2007. pp. 1-9. ISBN 978-0-8194-6629-7.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Neural mass model parameter identification for MEG/EEG.

Parallel Image Reconstruction Using B-Spline Approximation (PROBER)

Autoři: Petr, J., prof. Dr. Ing. Jan Kybic, Bock, M., Müller, S., Hlaváč, V.
Publikace: Magnetic Resonance in Medicine. 2007, 58(9), 582-591. ISSN 0740-3194.
Rok: 2007

DOI: 10.1002/mrm.21366
Odkaz: https://doi.org/10.1002/mrm.21366
Pracoviště: Katedra kybernetiky
Anotace:
Parallel MRI (pMRI) is a way to increase the speed of the MRI acquisition by com bining data obtained simultaneously from several receiver coils with distinct sp atial sensitivities. We propose an algorithm th at uses B-spline functions to approximate the reconstruction map which reduces t he number of parameters to estimate and makes the reconstruction faster and less sensitive to noise. The proposed method is tested on both phantom and in vivo images. The results ar e compared with commercial implementation of GRAPPA and SENSE algorithms in term s of time complexity and quality of the reconstruction.

Zpracování obrazu počítačem v medicíně

Autoři: Hlaváč, V., prof. Dr. Ing. Jan Kybic,
Publikace: Zdravotnická informatika. Praha: Karolinum, 2007. p. 105-114. ISBN 978-80-246-1378-9.
Rok: 2007

Pracoviště: Katedra kybernetiky
Anotace:
Kapitola ve skriptech základy zpracování obryzu počítačem v medicíně

Comparison of Methods for Tool Localization in Biological Tissue from 3D Ultrasound Data

Autoři: Barva, M., prof. Dr. Ing. Jan Kybic, Liebgott, H., Cachard, C., Hlaváč, V.
Publikace: Proceedings of the 2006 IEEE International Ultrasonics Symposium. New Jersey: IEEE, 2006. ISBN 1-4244-0201-8.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
In medical applications, miniature surgical instruments such as needles, or electrodes are introduced into human body. The position of instrument in tissue can be estimated using 3D ultrasound. In previous publications, we introduced two novel algorithms for automatic electrode localization from 3D ultrasound images. The first method is based on the Parallel Integral Projection (PIP) transform, a modification of the Radon transform. We showed that the axis of the electrode can be estimated from the maximum of PIP transformation. To accelerate search for the maximum, a hierarchical mesh-grid algorithm is implemented. In second method, the electrode axis is described by a cubic polynomial. The distribution of voxel intensities inside the electrode region is a priori estimated from acquired data. The model parameters are robustly estimated using the RANSAC estimator. In this paper, their performance in terms of accuracy is compared. A series of tests on numerical phantoms created with the FIELD II simulation program were performed to quantitatively evaluate the localization accuracy. We observed a decrease in accuracy when artificial noise was added to the input data. The algorithms were also tested on real ultrasound data of a cryogel phantom comprising metallic electrode.

Consistent and Elastic Registration of Histological Sections

Autoři: Arganda-Carreras, I., Sorzano, C., Marabini, R., Carazo, J., Ortiz-de-Solorzano, C., prof. Dr. Ing. Jan Kybic,
Publikace: CVAMIA: Computer Vision Approaches to Medical Image Analysis. Heidelberg: Springer, 2006. p. 85-95. ISBN 3-540-46257-0.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Consistent and Elastic Registration of Histological Sections.

Detection of Pulmonary Nodules in CT Scans

Autoři: Dolejší, M., prof. Dr. Ing. Jan Kybic,
Publikace: Analysis of Biomedical Signals and Images - Proceedings of Biosignal 2006. Brno: VUTIUM Press, 2006, pp. 251-253. ISSN 1211-412X. ISBN 80-214-3152-0.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Detection of Pulmonary Nodules in CT Scans

Elastic Image Registration for Movement Compensation in Digital Colposcopy

Autoři: Garcia Arteaga, J., prof. Dr. Ing. Jan Kybic, Li, W.
Publikace: Analysis of Biomedical Signals and Images - Proceedings of Biosignal 2006. Brno: VUTIUM Press, 2006, pp. 236-238. ISSN 1211-412X. ISBN 80-214-3152-0.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Colposcopy is a widespread and reliable diagnostic method for the early detection of cervical cancer. The basis of the diagnosis is the comparison of the progressive color and texture changes in a sequence of images of the uterine cervix. Before a computer based multiple image analysis is made for these images, camera and tissue movement must be compensated for. We use an elastic registration algorithm, representing the problem as an optimization over a set of continuous deformation vector fields. The regularization has been modelled after a linearized 2D elasticity operator describing equilibrium in an elastic material. Final results show that this registration algorithm works reliably and thus permits the subsequent temporal and multiple image analysis to be performed correctly.

Fast Parallel MRI Reconstruction Using B-spline Approximation (PROBER)

Autoři: Petr, J., prof. Dr. Ing. Jan Kybic, Hlaváč, V., Muller, S., Bock, M.
Publikace: Proceedings of SPIE Vol. 6142. Washington: SPIE, 2006. pp. 1251-1262. ISBN 0-8194-6185-7.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Parallel MRI (pMRI) is a way to increase the speed of the MRI acquisition by combining data obtained simultaneously from several receiver coils with distinct spatial sensitivities. The idea is to speed up the acquisition by sampling more sparsely in the k-space and to compensate the data loss using the additional information obtained by a higher number of receiver coils. We propose an algorithm that uses B-spline functions to approximate the reconstruction map which reduces the number of parameters to estimate and makes the reconstruction faster and less sensitive to noise.

Generalized Head Models for MEG/EEG: BEM beyond Nested Volumes

Autoři: prof. Dr. Ing. Jan Kybic, Clerc, M., Faugeras, O., Keriven, R., Papadopoulo, T.
Publikace: Physics in Medicine and Biology. 2006, 51(5), 1333-1346. ISSN 0031-9155.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Generalized topology extension for the symmetric MEG/EEG boundary element method.

Image Registration Accuracy Estimation without Ground Truth using Bootstrap

Autoři: prof. Dr. Ing. Jan Kybic, Smutek, D.
Publikace: CVAMIA: Computer Vision Approaches to Medical Image Analysis. Heidelberg: Springer, 2006. p. 61-72. ISBN 3-540-46257-0.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Image Registration Accuracy Estimation without Ground Truth using Bootstrap.

Incremental updating of nearest neighbor-based high-dimensional entropy estimation

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: ICASSP 2006 - Conference Proceedings. Piscataway: IEEE, 2006. ISSN 1520-6149. ISBN 1-4244-0469-X.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Incremental updating of nearest neighbor-based high-dimensional entropy estimation

Nízkodávková technika spirální CT plic v diagnostice metastatických ložiskových nálezů u dětí a v dorostovém věku

Autoři: Tůma, S., Neuwirth, J., Čumlivská, E., Mališ, J., prof. Dr. Ing. Jan Kybic, Šanda, J., Fricová-Poulová, M., Alda, T., Plovinčák, M.
Publikace: Česko-slovenská pediatrie. 2006, 61(4), 179-185. ISSN 0069-2328.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
CT plic pro diagnostiku u dětí.

Parallel Magnetic Resonance Imaging Reconstruction

Autoři: Petr, J., prof. Dr. Ing. Jan Kybic, Hlaváč, V.
Publikace: WORKSHOP 2006. Praha: České vysoké učení technické v Praze, 2006, pp. 134-136. ISBN 80-01-03439-9.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Reconstruction of images from parallel magnetic resonance.

Solitární a mnohocestné plicní uzly - analýza morfologických vlastností na průkaz jejich etiologie, pravidla sledování malých uzlů, jejich detekce pomocí semiautomatické analýzy CZ obrazu -

Autoři: Neuwirth, J., Polovincák, M., prof. Dr. Ing. Jan Kybic, Tůma, S., Čumlivská, E., Suchánek, V., Adla, T., Hlaváč, V., Dolejší, M.
Publikace: Česká radiologie. 2006, 60(5), 311-320. ISSN 1210-7883.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
Analyza morfologickych vlastnosti plicnich uzlu.

Ultrasound Image of Chronic Thyroiditis and its Relation to Antithyroid Antibodies

Autoři: Smutek, D., doc. Dr. Ing. Radim Šára, Holinka, Š., prof. Dr. Ing. Jan Kybic, Tesař, L., Jiskra, J., Maruna, P.
Publikace: Proceedings of the 11th Congress of the World Federation for Ultrasound in Medicine and Biology. Amsterdam: Elsevier Science, 2006, pp. 120. ISSN 0301-5629.
Rok: 2006

Pracoviště: Katedra kybernetiky
Anotace:
We study the relation between ultrasound image of thyroid gland with autoimmune inflammation and levels of antithyroid antibodies.

A Common Formalism for the Integral Formulations of the Forward EEG Problem

Autoři: prof. Dr. Ing. Jan Kybic, Clerc, M., Abboud, T., Faugeras, O., Keriven, R., Papadopoulo, T.
Publikace: IEEE Transactions on Medical Imaging. 2005, 24(1), 12-28. ISSN 0278-0062.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
We first present a dual approach which involves a single-layer potential. Then, we propose a symmetric formulation, which combines single and double-layer potentials, and which is new to the field of EEG, although it has been applied to other problems in electromagnetism.

Automatic Localization of Curvilinear Object in 3D Ultrasound Images

Autoři: Barva, M., prof. Dr. Ing. Jan Kybic, Mari, J., Cachard, C., Hlaváč, V.
Publikace: Medical Imaging 2005: Ultrasonic Imaging and Signal Processing. Washington: SPIE, 2005. pp. 455-462. ISSN 1605-7422.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Utilization of tools during surgical interventions sets the problem of their accurate localization within biological tissue. The ultrasound imaging represents an inexpensive and a flexible approach for a real-time image acquisition of tissue structure with metal instruments. There are several difficulties involving processing of ultrasound images: Their noisy nature makes the localization task difficult; the objects appear irregular and incomplete. Our task is to determine the position of a curvilinear electrode in biological tissue from a three-dimensional ultrasound image. Initially, the data are segmented by thresholding and processed with the randomized version of the RANSAC (R-RANSAC) algorithm. The curvilinear electrode is modeled by a three-dimensional cubic curve. Its shape is subject to check using a curvature measure in the hypothesis evaluation step of the R-RANSAC algorithm. Subsequently, we perform the least squares curve fitting to the data that have been marked by the R-

Computational Elastography from Standard Ultrasound Image Sequences by Global Trust Region Optimization

Autoři: prof. Dr. Ing. Jan Kybic, Smutek, D.
Publikace: Proceedings of IPMI, Lecture Notes in Computer Science. Heidelberg: Springer, 2005. p. 299-310. ISBN 3-540-26545-7.
Rok: 2005

Pracoviště: Katedra kybernetiky

Estimating Elastic Properties of Tissues from Standard 2D Ultrasound Images

Autoři: prof. Dr. Ing. Jan Kybic, Smutek, D.
Publikace: Medical Imaging 2005: Ultrasonic Imaging and Signal Processing. Washington: SPIE, 2005. p. 184-195. ISSN 1605-7422.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
We propose a way of measuring elastic properties of tissues in-vivo, using standard medical image ultrasound machine without any special hardware. Images are acquired while the tissue is being deformed by a varying pressure applied by the operator on the hand-held ultrasound probe. The local elastic shear modulus u is either estimated from a local displacement field reconstructed by an elastic registration algorithm, or both modulus u and the displacement are estimated simultaneously. The relation between modulus and displacement is calculated using a finite element method (FEM). The estimation algorithms were tested on both synthetic, phantom and real subject data.

Fast Multipole Acceleration of the MEG/EEG Boundary Element Method

Autoři: prof. Dr. Ing. Jan Kybic, Clerc, M., Faugeras, O., Keriven, R., Papadopoulo, T.
Publikace: Physics in Medicine and Biology. 2005, 50(19), 4695-4710. ISSN 0031-9155.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Fast multipole acceleration of the symmetric MEG/EEG boundary element method.

In vivo conductivity estimation with symmetric boundary elements

Autoři: Clerc, M., Adde, G., prof. Dr. Ing. Jan Kybic, Papadopoulo, T., Badier, J.M.
Publikace: NFSI2005: 5th International Conference on Bioelectromagnetism and 5th International Symposium on Noninvasive Functional Source Imaging. Tampere: International Society for Bioelectromagnetism, 2005, pp. 307-310. ISSN 1456-7857.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
In vivo conductivity estimation with symmetric boundary elements.

Localizing Metal Electrode from 3D Ultrasound Data Using RANSAC and Intensity Priors

Autoři: Barva, M., prof. Dr. Ing. Jan Kybic, Mari, J.M., Cachard, C., Hlaváč, V.
Publikace: IFMBE Proceedings EMBEC'05, 3rd European Medical and Biological Engineering Conference. Praha: International Federation for Medical and Biological Engineering, 2005, ISSN 1727-1983.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
In surgical interventions, miniature metallic tools such as thin electrodes and needles are introduced in tissue. Tracking systems are used to estimate their precise position. In this paper we describe an algorithm that exploits a three-dimensional ultrasound image and raw radio-frequency (RF) signal to determine the position of a thin, metallic electrode in biological tissue. We assume that electrode appears in 3D ultrasound image as a bright, elongated region. To estimate its position, a mathematical model of the region was established. It approximates the electrode axis with a polynomial curve. The voxel intensity distribution near the ray axis was determined from acquired RF signals. The model parameters are estimated by the RANSAC estimator. Finally, the electrode endpoints are located. The method was tested on real ultrasound data of a phantom with a thin tungsten electrode inserted. The results of experiments show that the method is stable even if the data are very noisy.

Longitudinal and Radial Regional Strain Obtained From Gray-Scale Conventional Echocardiography

Autoři: Ledesma, M., Santos, A., Mahia, P., Garcia Fernandez, M.A., prof. Dr. Ing. Jan Kybic, Malpica, N., Perez-David, E., Desco, M.
Publikace: Journal of the American College of Cardiology. 2005, 45(3), 255A. ISSN 0735-1097.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Longitudinal and Radial Regional Strain Obtained From Gray-Scale Conventional Echocardiography

Parallel MRI Reconstruction Using B-Spline Approximation

Autoři: Petr, J., prof. Dr. Ing. Jan Kybic, Müller, S., Bock, M., Hlaváč, V.
Publikace: ISMRM '05: Proceedings of the 13th Scientific Meeting and Exhibition of International Society for Magnetic Resonance in Medicine. Berkeley: International Society for Magnetic Resonance in Medicine, 2005, pp. 2421. ISSN 1545-4428.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Parallel MRI (pMRI) is a way to increase the speed of the MRI acquisition by sampling more sparsely in the k-space. The decrease of the data amount causing aliasing is compensated by a simultaneous use of several coils with distinct spatial sensitivities. We propose a new algorithm that estimates the coefficients of the linear reconstruction transformation directly without estimating the coil sensitivity maps first. The coefficients are approximated by B-spline functions. This reduces the number of variables and thus speeds up the estimation. It also makes the method more robust to noise. Tests were performed on phantom images. The proposed method had lower square difference error than two commercially used methods GRAPPA and mSENSE.

Quantitative Analysis of Microarrays

Autoři: Muresan, L., Heise, B., prof. Dr. Ing. Jan Kybic, Klement, E.
Publikace: ICIP 2005: Proceedings of 12th IEEE International Conference on Image Processing. Piscataway: IEEE, 2005. p. 1274-1277. ISSN 1522-4880. ISBN 0-7803-9134-9.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Quantitative Analysis of Microarrays

Spatio-Temporal Non-Rigid Registration for Ultrasound Cardiac Motion Estimation

Autoři: Ledesma-Carbayo, M., prof. Dr. Ing. Jan Kybic, Desco, M., Santos, A., Sühling, M., Hunziker, P., Unser, M.
Publikace: IEEE Transactions on Medical Imaging. 2005, 24(9), 1113-1126. ISSN 0278-0062.
Rok: 2005

Pracoviště: Katedra kybernetiky
Anotace:
Spatio-Temporal Non-Rigid Registration for Ultrasound Cardiac Motion Estimation.

Accurate Boundary Element Method for the Electro- and Magnetoencephalography Forward Problem

Autoři: prof. Dr. Ing. Jan Kybic, Clerc, M., Faugeras, O., Adde, G., Keriven, R., Papadopoulo, T.
Publikace: BIOSIGNAL 2004: Analysis of Biomedical Signals and Images. Brno: VUTIUM Press, 2004, pp. 121-123. ISBN 80-214-2633-0.
Rok: 2004

Pracoviště: Katedra kybernetiky

High-Dimensional Mutual Information Estimation for Image Registration

Autoři: prof. Dr. Ing. Jan Kybic,
Publikace: ICIP'04: Proceedings of the 2004 IEEE International Conference on Image Processing. Piscataway: IEEE, 2004. p. 4. ISBN 0-7803-8555-1.
Rok: 2004

Pracoviště: Katedra kybernetiky

Longitudinal and radial regional strain obtained from gray-scale conventional echocardiography

Autoři: Ledesma-Carbayo, M.J., Santos, A., Mahia, S., Garcia, F.M.A., prof. Dr. Ing. Jan Kybic, Malpica, N., David, P., Desco, M.
Publikace: Abstracts of EUROECHO, the Eighth Annual Meeting of the European Association of Echocardiography. Les Templiers: European Society of Cardiology, 2004, pp. 151.
Rok: 2004

Pracoviště: Katedra kybernetiky
Anotace:
We show how to obtain strain information from conventional heart ultrasound sequences.

Myocardial strain analysis of echocardiographic sequences using non-rigid registration

Autoři: Ledesma-Carbayo, M., Santos, A., prof. Dr. Ing. Jan Kybic, Casado, P.M., Fernadez, M.A.G., Malpica, N., David, E.P., Desco, M.
Publikace: Computers in Cardiology. Chicago: University of Chicago Press, 2004. pp. 313-316.
Rok: 2004

Pracoviště: Katedra kybernetiky

Non rigid Registration for Myocardial Motion Analysis

Autoři: Ledesma-Carbayo, M.J., Desco, M., Malpica, M., prof. Dr. Ing. Jan Kybic, Santos, A.
Publikace: 4th European Symposium on Biomedical Engineering. University of Patras, 2004,
Rok: 2004

Pracoviště: Katedra kybernetiky
Anotace:
We are using an elastic registration algorithm to analyze heart ultrasound sequencies.

Obtencion de las componentes axial y longitudinal del strain miocardico a traves de ecocardiografia convencional en scala de grises

Autoři: Ledesma-Carbayo, M.J., Santos, A., Mahia Casado, P., Garcia Fernandez, M.A., prof. Dr. Ing. Jan Kybic, Perez David, E., Desco, M., Sarnago Cebada, F.
Publikace: XX Congreso Nacional de la SEC, Revista Espanol de Cardiologia. Madrid: Sociedad Espanola de Cardiologia, 2004, pp. 59.
Rok: 2004

Pracoviště: Katedra kybernetiky

Radial Radon Transform dedicated to Micro-object Localization from Radio Frequency Ultrasound Signal

Autoři: Barva, M., prof. Dr. Ing. Jan Kybic, Mari, J.M., Cachars, C.
Publikace: UFFC '04: Proceedings of the IEEE International Ultrasonics, Ferroelectrics and Conference. Piscataway: IEEE, 2004. pp. 1836-1839. ISBN 0-7803-8412-1.
Rok: 2004

Pracoviště: Katedra kybernetiky
Anotace:
In this paper we describe a method for automatic electrode localization in soft tissue from radio-frequency signal. The method is exploiting a property of the Radon Transform (RT) that allows to localize a line-segment in 3D data. The method directly processes the radio-frequency (RF) signal provided by the ultrasound (US) probe. Thus, there's no need for ultrasound image reconstruction. The method is able to detect line-segments that are discontinuous. The low computational cost allows to process data in real-time. The algorithm was tested on a radiofrequency signal acquired by scanning a phantom containing a thin metal electrode of 150um in diameter. The experiments show that the developed technique is capable of reliably finding an arbitrarily positioned line-segment in a 3D image.

Radial Radon Transorm for Electrode Localization in Biological Tissue

Autoři: Barva, M., Mari, J., prof. Dr. Ing. Jan Kybic, Cachard, C.
Publikace: BIOSIGNAL 2004: Analysis of Biomedical Signals and Images. Brno: VUTIUM Press, 2004, pp. 299-301. ISBN 80-214-2633-0.
Rok: 2004

Pracoviště: Katedra kybernetiky

Variational, geometric, and statistical methods for modeling brain anatomy and function

Autoři: Faugeras, O., Adde, G., Charpiat, G., Chefd'Hotel, C., Clerc, M., Deneux, T., Deriche, R., Hermosillo, G., Keriven, R., Kornprobst, P., prof. Dr. Ing. Jan Kybic, Lenglet, C., Lopez-Perez, L., Papadopoulo, T., Pons, J.P., Segonne, F., Thirion, B., Tschumperle, D., Vieville, T., Wotawa, N.
Publikace: Neuroimage. 2004, 23(S1), 46-55. ISSN 1053-8119.
Rok: 2004

Pracoviště: Katedra kybernetiky

Regularization for the Inverse EEG/MEG Problem Using the Symmetric Boundary Element Method

Autoři: Adde, G., Clerc, M., Keriven, R., prof. Dr. Ing. Jan Kybic,
Publikace: Proceedings of the 4th International Symposium on Noninvasive Functional Source Imaging (NFSI). Berlin: Organ of the German Society for Biomedical Engineering in VDE, 2003. p. 287-289. ISSN 0939-4990.
Rok: 2003

Pracoviště: Katedra kybernetiky

A level set method for the inverse EEG/MEG problem

Autoři: Clerc, M., Faugeras, O., Keriven, R., prof. Dr. Ing. Jan Kybic, Papadopoulo, Th.
Publikace: Human Brain Mapping -- 8th International Conference on Functional Mapping of the Human Brain. Sendai: Human Brain Mapping, 2002.
Rok: 2002

Pracoviště: Katedra kybernetiky

Comparison of BEM and FEM methods for the E/MEG problem

Autoři: Clerc, N., Dervieux, A., Keriven, R., Faugeras, O., prof. Dr. Ing. Jan Kybic, Papadopoulo, Th.
Publikace: Biomag 2002: Proceedings of the 13th International Conference on Biomagnetism. Berlin: VDE Verlag, 2002. p. 688-690.
Rok: 2002

Pracoviště: Katedra kybernetiky

The Fast Multipole Method for the direct E/MEG problem

Autoři: Clerc, M., Keriven, R., Faugeras, O., prof. Dr. Ing. Jan Kybic, Papadopoulo, Th.
Publikace: Proceedings of the First 2002 IEEE International Symposium on Biomedical Imaging: Macro to Nano (ISBI'02). New York: IEEE Press, 2002. p. 1023-1026. ISBN 0-7803-7585-8.
Rok: 2002

Pracoviště: Katedra kybernetiky

Multidimensional Elastic Registration of Images Using Splines

Autoři: prof. Dr. Ing. Jan Kybic, Unser, M.
Publikace: ICIP2000, Proceedings of International Conference on Image Processing. Piscataway: IEEE, 2000, pp. 455-458. ISBN 07-8036-299-3.
Rok: 2000

Pracoviště: Katedra kybernetiky

Extended Spectral Subtraction

Autoři: prof. Ing. Pavel Sovka, CSc., doc. Ing. Petr Pollák, CSc., prof. Dr. Ing. Jan Kybic,
Publikace: EUSIPCO '96 Eight European Signal Processing Conference. Trieste: LINT, 1996. pp. 963-966. ISBN 88-86179-83-9.
Rok: 1996

Pracoviště: Katedra teorie obvodů, Katedra kybernetiky

prof. Dr. Ing. Jan Kybic

Všechny publikace

Automatic caries detection in bitewing radiographs—Part II: experimental comparison

Automatic caries detection in bitewing radiographs: part I---deep learning

Deep learning for laser beam imprinting

Fully automated imaging protocol independent system for pituitary adenoma segmentation: a convolutional neural network—based model on sparsely annotated MRI

Learning to segment from object thickness annotations

Characterization of drug effects on cell cultures from phase-contrast microscopy images

Fast learning from label proportions with small bags

Learning to segment cell nuclei in phase-contrast microscopy from fluorescence images for drug discovery

Learning to segment from object sizes

The Effect of Primary Aldosteronism on Carotid Artery Texture in Ultrasound Images

Automatic evaluation of human oocyte developmental potential from microscopy images

Differentiating between stable and progressive carotid atherosclerotic plaques from in-vivo ultrasound images using texture descriptors

Potential Biomarkers From Positive Definite 4th Order Tensors In Hardi

ANHIR: Automatic Non-rigid Histological Image Registration Challenge

Generalized Multiple Instance Learning for Cancer Detection in Digital Histopathology

Multiple Instance Learning Via Deep Hierarchical Exploration for Histology Image Classification

AbloCAM: a versatile optomechanical, semi-automated tool for an in situ characterization and optimization of focused XUV/x-ray laser beam (Conference Presentation)

Benchmarking of image registration methods for differently stained histological slides

Detecting multiple myeloma via generalized multiple-instance learning

Fast registration by boundary sampling and linear programming

Incremental B-spline deformation model for geometric graph matching

Reconstructing Evolving Tree Structures in Time Lapse Sequences by Enforcing Time-Consistency

Volume estimation from single images: an application to pancreatic islets

Binary pattern dictionary learning for gene expression representation in drosophila imaginal discs

Detection and Localization of Drosophila Egg Chambers in Microscopy Images

Geometric Graph Matching Using Monte Carlo Tree Search

Langerhans islet volume estimation from 3D optical projection tomography

Left ventricle Hermite-based segmentation

Region growing using superpixels with learned shape prior

Supervised and unsupervised segmentation using superpixels, model estimation, and Graph Cut

Automated Analysis of Microscopic Images of Isolated Pancreatic Islets

Automated separation of merged Langerhans islets

Characterization of Hematologic Malignancies based on Discrete Orthogonal Moments

Classification of tumor epithelium and stroma in colorectal cancer based on discrete tchebichef moments

Comparison of volume estimation methods for pancreatic islet cells

Fully automated classification of bone marrow infiltration in low-dose CT of patients with multiple myeloma based on probabilistic density model and supervised learning

Texel-based image classification with orthogonal bases

Automatic detection of bone marrow infiltration by multiple myeloma detection in low-dose CT

Color normalization for robust evaluation of microscopy images

Fast registration of segmented images by normal sampling

Feature Ensemble for Quantitative Analysis of Emphysema in CT imaging

Geometrical Graph Matching using Monte Carlo Tree Search

Non-Rigid Graph Registration using Active Testing Search

Supertexton-based segmentation in early Drosophila oogenesis

Automatic Analysis of Spatial Gene Expression Patterns

Automatic simultaneous segmentation and fast registration of histological images

Classification of microscopy images of Langerhans islets

jSLIC : superpixels in ImageJ

Neuromuscular fiber segmentation through particle filtering and discrete optimization

Organ-Focused Mutual Information for Nonrigid Multimodal Registration of Liver CT and Gd-EOB-DTPA-enhanced MRI

Path Descriptors for Geometric Graph Matching and Registration

Reconstructing Evolving Tree Structures in Time Lapse Sequences

Registration of segmented histological images using thin plate splines and belief propagation

Active Testing Search for Point Cloud Matching

Line Filtering for Surgical Tool Localization in 3D Ultrasound Images

Registration of multiple stained histological sections

A Multi-Frequency Approach to Increase the Native Resolution of Ultrasound Images

Approximate All Nearest Neighbor Search for High Dimensional Entropy Estimation for Image Registration

Automated segmentation of a motion mask to preserve sliding motion in deformable registration of thoracic CT

Neuromuscular fiber segmentation using particle filtering and discrete optimization

Robust elastic 2D/3D geometric graph matching

Robust Non-Rigid Registration of 2D and 3D Graphs

Automatic Colposcopy Video Tissue Classification Using Higher Order Entropy Based Registration

Bootstrap Optical Flow Confidence and Uncertainty Measure

Spatio-Temporal Motion Estimation for Respiratory-Correlated Imaging of the Lungs

3D Ultrasound real-time monitoring of surgical tools

Bootstrap Resampling for Image Registration Uncertainty Estimation without Ground Truth

Model Fitting Using RANSAC for Surgical Tool Localization in 3-D Ultrasound Images

Non-rigid consistent registration of 2D image sequences

Semi-automated segmentation of Symptomatic Exudate-Associated Derangements (SEADs) in 3D OCT using layer segmentation

Discrete curvature calculation for fast level set segmentation

Line filtering for detection of microtools in 3D ultrasound data

Respiratory Motion Estimation from Cone-Beam Projections Using a Prior Model

The Lung TIME-Annotated Lung Nodule Dataset and Nodule Detection Framework

2-D Locally Regularized Tissue Strain Estimation From Radio-Frequency Ultrasound Images: Theoretical Developments and Results on Experimental Data

Algoritmy CAD systému pro CT vyšetření plic na plicní uzly

Estimating Respiratory Motion from Cone-Beam Projections

Fast no ground truth image registration accuracy evaluation: Comparison of bootstrap and Hessian approaches