Search Papers | Poster Sessions | All Posters
Poster C73 in Poster Session C - Friday, August 9, 2024, 11:15 am – 1:15 pm, Johnson Ice Rink
Decomposition of brain calcium signals in a Pavlovian learning task
Farzan Nadim1, Junichi Yoshida2, Kamran Khodakhah3, Germán Heim4, Mina Eskandar1, Horacio Rotstein (1; 1New Jersey Institute of Technology, 2Yale School of Medicine, 3Albert Einstein College of Medicine, 4Universidad Nacional del Sur
Learning tasks, even the simplest ones, involve multiple brain regions (involved in sensory processing, evaluation of outcomes and generation of movement). Experimental protocols combined with modern techniques (e.g., fiber photometry) allow for the simultaneous recording of these observed values (or predictors, e.g., movement, sensory signals, etc.) and the activities of multiple brain regions (e.g., cerebellum, substantia nigra pars compacta) during a learning task. However decomposing the neural signals into components that encode different predictors and the neural signal they produce during the learning process remains challenging. Existing methods (e.g., generalized linear models) that allow for the identification of the neural signals produced by each predictor and the estimation of the contribution of each predictor to the total signal during the learning process fail to capture the signal kinetics associated with each predictor. In this work we address this issue. We use optimization tools to decompose the neural signals into the respective time-dependent contributions (kernels) of each predictor under certain modeling assumptions. We used this kernels and their defining properties to examine how the associated signals change on a session-by-session basis and established the extent to which the cerebellum contributes to dopaminergic signaling in the process of conditioned learning.
Keywords: Learning Neural circuits Cerebellum Basal ganglia