Dynamic Network Transitions Preceding Auditory Hallucination Onset: A Systematic Review of Burst-Capture fMRI and EEG-fMRI Integration

Auditory verbal hallucinations (AVH) are transient, internally generated percepts that emerge against spontaneous brain activity, yet most neuroimaging contrasts average hallucination ‘ON’ against ‘OFF’ periods and thereby dissolve the transition that is mechanistically decisive. This systematic review synthesises symptom-capture functional magnetic resonance imaging (fMRI), dynamic functional connectivity, electroencephalography (EEG) microstate, perfusion and simultaneous EEG-fMRI studies of AVH identified in five databases, and critically appraises seven transient-state estimators. We introduce ‘burst-capture’, an onset-anchored strategy in which brief hallucinatory episodes define the temporal frame and the pre-onset interval is the unit of inference. Onset-anchored studies consistently implicate superior temporal and primary auditory cortex, inferior frontal gyrus, hippocampus and thalamus. Dynamic analyses show that hallucination-relevant abnormalities are state-bound and invisible to static connectivity: hallucination proneness tracks reduced dwell time in states of default mode network (DMN)–task-positive anticorrelation; patients with AVH under-occupy, and switch less often into, a DMN–language antagonistic state; hallucination periods show more erratic switching and shortened EEG microstate class D; and stage-resolved effective connectivity implicates the salience network during emergence. Convergent perfusion, gamma-synchrony and prestimulus-state evidence supports a tonic vulnerability that lowers the threshold for a phasic network transition. We propose a transition-competition model in which AVH onset is a threshold crossing within a multistable network system, and specify the requirements of a valid burst-capture EEG-fMRI study: reaction-time-independent anchors, surrogate-data testing of state estimators and sign-aware EEG–BOLD fusion. Transition metrics, not averaged activation, should become the primary outcomes of hallucination neuroimaging.