Toward the identification of neural replay in humans using sEEG

Neural replay is the repetition during rest of sequential neural activity patterns present during practice. After cooking, or playing a piano partition, the brain may automatically replay part of the neural activity corresponding to the newly learned movement. It has been suggested in rodent studies that neural replay plays an important role in memory and learning consolidation.
Studies demonstrating neural replay in humans are limited, both because of the recording techniques and the analytical methods. We proposed a new method using intracranial electroencephalographic (iEEG) recordings over 24h and machine-learning analyses to address this gap and better understand the learning capabilities of the brain.
Preliminary results show that we are able to detect the learned sequence during the task from the iEEG signal, indicating that we can trust our classifier's predictions. We then applied this framework overnight and obtained promising preliminary results in identifying neural replay.

Présentation faite dans la même séance que celle de Benoît Chatard. 

Romain Quentin et Benoît Chatard sont invités par Michael Pereira.