Decoding activity in Broca's area predicts the occurrence of auditory hallucinations across subjects

BACKGROUND: Functional magnetic resonance imaging (fMRI) capture aims at detecting auditory-verbal hallucinations (AVHs) from continuously recorded brain activity. Establishing efficient capture methods with low computational cost that easily generalize between patients remains a key objective in precision psychiatry. To address this issue, we developed a novel automatized fMRI-capture procedure for AVHs in schizophrenia patients. METHODS: We used a previously validated, but labor-intensive, personalized fMRI-capture method to train a linear classifier using machine-learning techniques. We benchmarked the performances of this classifier on 2320 AVH periods vs. resting-state periods obtained from schizophrenia patients with frequent symptoms (n=23). We characterized patterns of BOLD activity that were predictive of AVH both within- and between-subjects. Generalizability was assessed with a second independent sample gathering 2000 AVH labels (n=34 schizophrenia patients), while specificity was tested with a nonclinical control sample performing an auditory imagery task (840 labels, n=20). RESULTS: Our between-subject classifier achieved high decoding accuracy (area-under-the-curve, AUC = 0.85) and discriminated AVH from rest and verbal imagery. Optimizing the parameters on the first schizophrenia dataset and testing its performance on the second dataset led to a 0.85 out-of-sample AUC (0.88 for the converse test). We showed that AVH detection critically depends on local BOLD activity patterns within Broca's area. CONCLUSIONS: Our results demonstrate that it is possible to reliably detect AVH-states from BOLD signals in schizophrenia patients using a multivariate decoder without performing complex regularization procedures. These findings constitute a crucial step toward brain-based treatments for severe drug-resistant hallucinations.