
-
Information and structure from neural code through the lens of new analysis techniques
Under the MNESYS project two strands of research were carried out in parallel to achieve complementary results. In particular, various studies based on the decoding technique, that is, the interpretation through algorithms of neural activity to obtain information regarding behavior or external variables, have shown that the posterior parietal cortex processes information regarding the spatial location of a target to be reached with the upper limbs or the timing of movement with great precision. At the same time, we began to study the geometry of neural representations to understand how information that can be decoded is actually encoded by the parietal cortex. For that purpose, dimensionality reduction techniques have been applied to both simulated and real data to extract the geometry of population activity and understand how it organizes and varies over time.
Decoding and the study of the geometry of neural representations open up concrete perspectives in application areas. The integration of high-precision decoding algorithms with low-dimensional models can give rise to faster, more robust and intuitive human-machine interfaces, and return natural movements to patients with spinal cord injuries, strokes or severe motor deficits, significantly improving their autonomy and quality of life. Furthermore, the analysis of the geometric structure of population activity in subjects suffering from neurodegenerative diseases or motor disorders can allow the identification of specific neural biomarkers, monitoring the pathological evolution and effectiveness of rehabilitation or pharmacological therapies. These approaches can encourage the development of personalized and predictive adaptive neuromodulation protocols, in which electrical or magnetic stimulations can be calibrated in real time based on network dynamics.
Healthcare, Lifescience
Settori Scientifico Disciplinari
Spoke 4 : Perception and Brain-Body Interaction



