The Visual Cortex as the Source of Trauma Intrusions: Causal Evidence from TMS and fMRI

Intrusive memories – involuntary, vivid recollections that feel present – are the defining and most distressing symptom of PTSD. Dominant neurobiological models locate the pathology in limbic-prefrontal dysregulation: an overactive amygdala, an underactive ventromedial prefrontal cortex, and a hippocampus that fails to bind the trauma into ordinary autobiographical memory. Within that frame, the sensory richness of an intrusion is downstream noise. The Zhejiang University group inverted the question and asked whether the visual cortex itself is a causal node in producing intrusions, not merely a passive screen onto which limbic activity is projected.

The design combined a trauma-film paradigm with fMRI and a causal perturbation. Healthy participants watched aversive film footage, then underwent fMRI during encoding and a post-encoding rest phase in which spontaneous intrusions were logged. One group received 1-Hz inhibitory rTMS over early visual cortex (V1/V2); a control group received stimulation at the vertex. Intrusions were then recorded across the next week in daily diaries. This is a prevention-window study: the manipulation falls in the consolidation period, when the sensory memory is still labile.

The imaging results are the conceptual core. Intrusive episodes coincided with elevated occipital visual cortex activation and – critically – with stable multivariate neural patterns, meaning the same visual representation was being reinstated each time the memory broke through. Functional and effective connectivity analyses showed that intrusions were predicted by interactions between the middle frontal gyrus and occipital cortex, and dynamic causal modeling confirmed these fronto-occipital interactions were direct and bidirectional. They coexisted with prefrontal-limbic activity but tracked the moment-to-moment intrusion dynamics more tightly. The causal arm closes the inferential gap that correlational neuroimaging cannot: inhibiting visual cortex during consolidation suppressed intrusions for days afterward, while sparing gist recognition – the unwanted re-experiencing weakened, but the participant could still remember what they had seen.

Why the visual system matters clinically

The finding aligns with a long-standing clinical observation that has lacked a tidy mechanism: visuospatial interference tasks performed soon after trauma reduce later intrusions, and reconsolidation-blockade approaches target sensory-perceptual rather than verbal memory. A visual-cortex locus offers a coherent account – intrusions are reinstatements of a sensory representation, and competing for or dampening that representation during its labile window weakens it. It reframes the early hours after trauma as a perceptual, not only an emotional, intervention window.

From probe to target

The authors used TMS as an experimental scalpel to establish causality, not as a ready therapy; one should not read this as a recommendation to stimulate the occipital lobe of trauma survivors. The clinical value is the identification of the visual sensory system as a mechanistically grounded target – one that could be approached pharmacologically, behaviorally (visuospatial loading), or eventually with refined neuromodulation, and that complements rather than replaces the limbic-prefrontal account.