Skip to main content

Machine Learning for Healthcare and Life Sciences

IBM Research - Haifa

FRR project in Neuroscience


This research project, carried out in collaboration with Professor Talma Hendler and her team at the Tel Aviv Sourasky Medical Center and Tel Aviv University, focuses on developing measurable neuromarkers of emotion that stem from activities of multiple regions in the brain. The project focuses on two major challenges: First, how can we use fMRI and EEG, recorded together, to improve our ability to measure emotional states from EEG alone? This work [1] may lead to more effective neurofeedback treatments of emotional regulation. Second, can we find reliable measures of emotional intensity in dynamic functional connectivity measures? Previous work by Professor Hendler's team have shown that such measures, defined for manually specified networks, indeed reflect emotional intensity at the group level. Our work [2] focuses on learning brain networks whose connectivity dynamics reflect various features of emotional state. Such measures may, in the future, be used as targets for neurofeedback treatments or for better pre-operating mapping.


[1] Estimating BOLD Signals of Deep Brain Networks from EEG Using Canonical Correlation Analysis. T El Hay, S Kinreich, N Slonim, I Podlipsky, T Hendler, L Shpigelman. Proceedings of OHBM 2012 (Abstract and Oral Talk), Beijing, China.

[2] Learning Cohesion Networks. L Shpigelman, G Raz, T Hendler. OHBM 2013 (workshop oral presentation), Seattle, USA.

Contact: Lavi Shpigelman