Extracting Root-Causal Brain Activity Driving Psychopathology from Resting State fMRI
Eric V. Strobl
TL;DR
This work addresses the challenge of mechanistic interpretability in psychiatric rs-fMRI by introducing a bilevel structural causal model that links between-subject symptom structure to within-subject brain activity through independent latent sources. The SOURCE algorithm recovers latent sources, estimates sparse root-proximal voxel maps, and learns interpretable symptom axes driven by a small subset of sources, all while treating voxel propagation as an invariant nuisance. Across psychosis (BSNIP2) and depression (EMBARC) datasets, SOURCE achieves higher predictive accuracy, better localization, and improved interpretability compared with baselines. The approach advances mechanistic understanding of neural drivers of psychopathology and provides a principled framework for root-causal analysis in rs-fMRI.
Abstract
Neuroimaging studies of psychiatric disorders often correlate imaging patterns with diagnostic labels or composite symptom scores, yielding diffuse associations that obscure underlying mechanisms. We instead seek to identify root-causal maps -- localized BOLD disturbances that initiate pathological cascades -- and to link them selectively to symptom dimensions. We introduce a bilevel structural causal model that connects between-subject symptom structure to within-subject resting-state fMRI via independent latent sources with localized direct effects. Based on this model, we develop SOURCE (Symptom-Oriented Uncovering of Root-Causal Elements), a procedure that links interpretable symptom axes to a parsimonious set of localized drivers. Experiments show that SOURCE recovers localized maps consistent with root-causal BOLD drivers and increases interpretability and anatomical specificity relative to existing comparators.