Learning exceptions to category rules is supported by distinct white matter networks

Poster Session D - Monday, April 15, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Melisa Gumus¹ (melisa.gumus@mail.utoronto.ca), Nahal Alizadeh Saghati¹, Michael Mack¹; ¹University of Toronto

Flexible concept learning involves a host of core mechanisms including working memory, decision making and attention. Such complexity is reflected in a wide range of brain regions spanning occipital, temporal, and frontal lobes that have been implicated in concept learning. Although a rich literature exists on the functional properties of cortical and subcortical regions in flexible concept formation, the involvement of white matter structures between these regions remains an open question. Participants (N=37) completed a complex visual category learning task in the MRI scanner. Category memberships were based on a multidimensional rule with exceptions to rules, as a measure of flexible learning. We quantified the integrity of white matter tracts in the whole human brain with diffusion weighted imaging. Correlational tractography was then used to identify the white matter structures that were related to different behavioural indices of learning. We observed distinct white matter tracts involved in learning exceptions to categories: forceps minor, bilateral fornix, bilateral superior/anterior thalamic radiation, and left cingulum frontal parietal. The integrity, as indexed with fractional anisotropy, of these tracts predicted participant-specific exception learning ability. In contrast, learning rule following items was supported by a separate white matter network involving middle and inferior longitudinal fasciculus. These findings suggest that learning category regularities and exceptions relies distinctly and primarily on a white matter network that spans thalamus, prefrontal cortex, and hippocampus, and highlight the importance of understanding the role of white matter structures in flexible concept learning.

Topic Area: LONG-TERM MEMORY: Episodic