The study, entitled ‘Global remapping of the sensory homunculus emerges early in childhood development’, was produced in collaboration with several research teams specialising in neuroscience and developmental psychology from the University of Cambridge, Durham University and University College London. Representing FEE CTU were Associate Professor Matěj Hoffmann and Dr Zdeněk Straka from the humanoid robotics group at the Department of Cybernetics. Roboticists from EPFL Lausanne also took part in the experiments; they developed a special pneumatic device enabling the stimulation of touch whilst the brain is being scanned in an MRI scanner.
The brain’s body map
The brain contains what is known as the somatosensory map of the body, sometimes referred to as the sensory homunculus. This is an arrangement of areas in the cerebral cortex that process touch and other sensory information from various parts of the body.
This map was first described in the mid-20th century by the Canadian neurosurgeon Wilder Penfield during operations on patients with epilepsy. Different parts of the body are represented in the brain by areas of varying sizes depending on their sensitivity – for example, the hands or face occupy significantly more space than the back. However, this map is far more dynamic than was long assumed.
How the brain adapts to a missing limb
Using functional magnetic resonance imaging (fMRI), researchers monitored brain activity in children aged 5–7 and in adults – both in people with an upper limb difference and in a control group.
The results showed that in people born without an arm, there is extensive reorganisation of the brain’s body map. The area of the brain that would normally represent the arm does not remain inactive – instead, it begins to respond to signals from other parts of the body. Moreover, the changes are not limited to the immediate vicinity of this area, but extend to a wider part of the somatosensory cortex.
Homeostatic plasticity: the brain maintains balance
The key to understanding these changes is a mechanism known as homeostatic plasticity. "Every neuron in the brain ´monitors´ itself to some extent to ensure it remains active – that it receives an adequate amount of input signals. When it loses some inputs, it begins to amplify others to maintain this level of activity," explains Matěj Hoffmann from the Faculty of Electrical Engineering at the Czech Technical University in Prague.
So if a neuron in the area of the brain dedicated to the hand does not receive signals from the missing limb, it begins to amplify inputs from other parts of the body – for example, from the arm, wrist or sometimes even the face.
"The brain is actually actively seeking balance across the entire network. It is not just a simple ´use it or lose it´ rule, but a more complex regulatory mechanism," adds Dr Zdeněk Straka.
Computational model from the Faculty of Electrical Engineering, CTU
It was precisely this mechanism that helped explain the computational model developed at the Faculty of Electrical Engineering, CTU. The model created by Dr Zdeněk Straka simulated the behaviour of neurons in the somatosensory cortex. Each neuron in the model followed a simple rule of homeostatic plasticity – it sought to maintain a stable level of activity.
When the researchers ´removed´ the inputs from the hand in the model – that is, simulated a congenital limb difference – activity in the network began to reorganise in a way that corresponded very closely to the results measured in brain scans.
"It turned out that even a relatively simple model can explain the main experimental results very well. The article’s reviewers rated this approach as one of its strengths," says Dr Straka.
The intersection of robotics and neuroscience
The involvement of the team from the Faculty of Electrical Engineering at the Czech Technical University in Prague is linked to their long-term research into the representation of the body and touch in the brain – and also in robotic systems.
For example, the researchers previously developed a model that enabled the humanoid robot iCub to create its own ´body map’ based on tactile stimuli from artificial skin.
"In robotics, we’re tackling a similar question: how does a system learn where a touch occurred on the body and how does it create a map of its own body? These principles are surprisingly close to what happens in the human brain," says Dr Hoffmann.
One more piece of the puzzle
For the humanoid robotics research group, this study is one step towards a deeper understanding of how the brain represents the body and how this representation develops.
"The somatosensory map is actually the first place in the brain where touch appears. But we’re also interested in how this information connects with motor skills, vision and other senses – for example, when a child feels a touch and then tries to reach out to that spot," explains Matěj Hoffmann.
It is precisely these processes that researchers are currently studying through a combination of experiments with children and modelling on humanoid robots.
Video: The robot homunculus: learning of artificial skin representation inspired by the brain
Photo Credit: Petr Neugebauer
