Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Cognitive underpinnings of hallucinations derived from the fMRI literature

Poster Session B - Sunday, March 8, 2026, 8:00 – 10:00 am PDT, Fairview/Kitsilano Ballroom

Sebastian Brstilo¹ (), Abhijit Chinchani¹, Todd S. Woodward¹; ¹University of British Columbia

Background: Auditory verbal hallucinations (AVH) are beliefs about hearing voices without an external source and represent a key positive symptom of schizophrenia and related psychotic disorders. Understanding the physiological mechanisms underlying AVH remains a key challenge in cognitive neuroscience. Comparing fMRI activations collected during hallucination experiences with those collected during a range of other cognitive operations, such as responding and attention to internal representations, provides insights into the cognitive underpinnings of hallucinations, to the extent that the function of the activated brain networks is understood. Methods: This study examined voxel data from a total of 166 voxel locations associated with AVH extracted from 15 fMRI studies. For each voxel, we computed activation strength across 11 task-based functional networks, for which functions are largely known, and published in an edited volume (Percival & Woodward, 2025). Task-based functional networks better capture functionally interpretable cognitive modes than resting-state networks. Results: We found that voxel locations for the 15 studies were primarily located in peaks associated with 6 different anatomical patterns, each associated with distinct cognitive functions, including attending to auditory stimuli, perceiving auditory stimuli, responding, linguistic processing, attending to internal representations, and attending to external representations. Conclusion: An independent fMRI dataset involving AVH capture with motor responses confirmed that consistent activation in the response network reflects an artifact of experimental design rather than intrinsic AVH-related activity. The dominance of confounding activations over theorized auditory perception regions is discussed. In conclusion, capturing reliable BOLD responses to AVH remains challenging.

Topic Area: METHODS: Neuroimaging