Ivan Zubarev

This paper introduces a deep learning framework, FC-CNN, to predict brain states from frequency-resolved MEG functional connectivity and shows it outperforms conventional methods in age prediction.

This paper develops an EEG-guided robotic TMS system that controls both pulse timing and location in real time, improving closed-loop neurostimulation for research and clinical applications.

This study demonstrates the feasibility of timing transcranial magnetic stimulation (TMS) to interictal epileptiform discharges (IEDs) in children with SeLECTS using a CNN-based EEG pipeline.

The paper identifies three distinct subtypes of somatostatin-expressing neurons in the mouse ventral tegmental area, revealing their heterogeneity in electrophysiological, morphological, and molecular properties.