EEG-Based Detection of Brain State Dynamics Associated with Auditory Hallucinations

2026 Award: $101,410

For many people with schizophrenia, hearing voices that aren’t there is a daily reality, but existing treatments don’t work for everyone. Our research uses brain activity recordings to map how the brain shifts between different states, and to pinpoint the moments when it becomes most vulnerable to producing hallucinations. This knowledge could transform brain stimulation therapy from a “where to stimulate” approach into a “where and when to stimulate” approach, one that acts at precisely the right moment to weaken the abnormal brain states that give rise to hallucinations.

Need/Problem: Voices heard in the absence of a real sound are known as auditory verbal hallucinations. This symptom affects most people with schizophrenia, and for about 1 in 4 patients, hallucinations persist even with medication, increasing the risk of social isolation and suicide. Current brain stimulation treatments show promise but have produced inconsistent results, likely because most studies focus on which part of the brain to stimulate without considering what the brain is doing at that moment.

Grant Summary: This project will use recordings of the brain’s electrical activity with millisecond precision to identify the specific moments when the brain becomes vulnerable to producing hallucinations. By studying brain activity patterns in individuals with schizophrenia, we aim to build a system that can detect these vulnerable moments in real time.

Goals & Projected Outcomes: This study will create a prototype that recognizes when the brain is shifting toward a state associated with hallucinations. This will lay the groundwork for a future clinical trial testing whether delivering brain stimulation precisely at those vulnerable moments can prevent hallucinations from occurring.

Rosa Vlasova, PhD

Guorong Wu, PhD

Grant Details: Hallucinations in schizophrenia are not caused by any single brain region. The brain constantly transitions between numerous functional states, and some of those states are associated with experiencing hallucinations. Most people have had a glimpse of this: the fleeting voices that occur while falling asleep or waking up. In patients with schizophrenia, these experiences happen more frequently and cannot be controlled. Our team previously used brain imaging (fMRI) to identify which brain networks are involved. This project now addresses when those vulnerable patterns emerge. Using advanced computational methods, we will map the brain’s activity states, identify the warning signs that precede hallucinations, and develop a real-time detector for those warning signs.