Basic Neurosciences · Paper A
Neuroimaging in schizophrenia
Neuroimaging in schizophreniaA long-term mental illness involving episodes of psychosis such as hallucinations and delusions.Read more → (a psychotic illness involving loss of touch with reality) shows both structural and functional changes, but no single scan finding can diagnose it. For example, imagine a patient with newly diagnosed schizophrenia whose brain scan shows enlarged ventricles even though they have never taken antipsychotic medication. In practice, that finding in a drug-naive patient argues against these changes being purely caused by medication.
Structurally, imaging shows enlarged brain ventricles, modest reduction in grey matterBrain tissue made mostly of nerve cell bodies, where most processing happens.Read more → volume, and reduced hippocampal and amygdalaA brain structure central to processing fear and emotional reactions.Read more → volume. These changes appear even in some drug-naive first-episode patients. Functionally, scans show hypofrontality: reduced activation of the prefrontal cortexThe front part of the brain responsible for planning, judgement and controlling impulses.Read more → (the brain's planning and reasoning region) during tasks needing executive function, and reduced resting metabolism there. Ultra-high-risk patients show accelerated grey matter loss around the transition to first-episode psychosis, supporting a model where the illness develops progressively over time. Take a young person at ultra-high risk of psychosis whose brain scans show accelerating tissue loss just before their first psychotic episode. This pattern supports using neuroimaging as a research tool tracking illness progression, not as a routine diagnostic test.
(a severe mental illness involving , , and disorganised thinking) is a clinical diagnosis, made from a patient's history and mental state examination, not from a scan. Neuroimaging (the use of techniques like CT, MRI, or PET to produce pictures of brain structure or activity) nonetheless plays two useful roles: excluding organic differentials and, in research settings, tracking how the illness develops over time.
For example, imagine a patient with newly diagnosed whose brain scan shows enlarged ventricles even though they have never taken antipsychotic medication. That finding, present before any drug exposure, argues against the change being purely a side effect of treatment, and instead suggests something about the illness itself alters brain structure in some patients, well before treatment begins.
No single neuroimaging finding, structural or functional, is diagnostic of . These are real, reproducible group-level findings, but none is sensitive or specific enough to confirm, or exclude, the diagnosis in an individual person sitting in front of you. Coming from a medical background where imaging often clinches a diagnosis, this is one of the more counterintuitive ideas in psychiatry for a new graduate.
Structural imaging (scans, usually MRI, that show the physical shape and size of brain tissue) in most consistently shows three things: enlarged brain ventricles (the fluid-filled spaces inside the brain), a modest reduction in volume, and reduced hippocampal and volume. Grey matter (the outer, cell-body-rich layer of brain tissue responsible for processing information) loss is diffuse rather than focal, spread broadly rather than concentrated in one lobe. Functional imaging (scans, typically PET or fMRI, that show brain activity or metabolism) reveals a separate pattern called hypofrontality: reduced activation of the (the brain region responsible for planning, reasoning, and executive function) during tasks requiring executive function, alongside reduced resting metabolism in that same region.
These changes are not confined to patients who have been unwell for years: they appear even in some drug-naive first-episode patients, scanned at their very first presentation, before receiving any antipsychotic medication. Longitudinal research, studies that scan the same people repeatedly over time, adds a further layer: patients identified as being at ultra-high risk of psychosis, before any diagnosable illness has emerged, show accelerated loss specifically around the time of transition to a first psychotic episode. This pattern supports a model in which the illness develops progressively, with imaging acting as a research tool for tracking that progression rather than as a routine diagnostic test.
It helps to walk through the logic in order. First, a person at ultra-high risk is scanned and shows a broadly normal, or only mildly reduced, volume. Second, as they approach transition to a first psychotic episode, serial scans show that volume falling faster than it would in someone who never develops psychosis. Third, once the illness is established, cross-sectional studies comparing many patients with against many healthy people find the group-level differences described above: enlarged ventricles, reduced grey matter, hypofrontality. None of these three steps, taken alone, tells you that one specific person in front of you has schizophrenia; taken together across a research cohort, they build a picture of a progressive, biologically real process running alongside the clinical illness.
Because structural changes appear even before any drug exposure, they must be distinguished from antipsychotic-induced brain changes: a drug-naive first-episode finding argues for an illness-related process, not a medication side effect. Imaging must also be distinguished from its actual clinical use: not as a way to positively diagnose , but as a screening tool to exclude organic mimics such as a brain tumour, stroke, or encephalitis that can present with psychotic symptoms but need an entirely different management pathway.
Diagnosis rests on clinical assessment against a recognised classification system, such as ICD-11 or DSM-5, which define through characteristic symptoms, duration, and functional impact rather than any imaging appearance. Where imaging earns its place clinically is as a baseline investigation: a first-episode psychosis assessment will often include a scan specifically to rule out a brain tumour, stroke, or other organic cause before psychotic symptoms are attributed to schizophrenia, catching a treatable organic cause before it is missed.
Because imaging does not diagnose or stage , it also does not directly guide first-line management. Treatment decisions rest on clinical guidelines, such as NICE guidance on psychosis and schizophrenia, which centre on antipsychotic medication alongside psychological therapy (such as cognitive behavioural therapy for psychosis) and social support, chosen and adjusted according to symptoms and response rather than any scan appearance. A patient with a strikingly abnormal scan and a patient with an entirely normal one can both meet full diagnostic criteria for schizophrenia and can both be treated identically.
The trap is assuming a finding like enlarged ventricles or hypofrontality confirms the diagnosis, or that a normal scan rules it out. Neither is true: imaging supports the story, it does not tell it. The clinical interview, mental state examination, and application of diagnostic criteria remain the tools that actually make the diagnosis, with neuroimaging kept in reserve to exclude the organic mimics that would otherwise be missed.
Neuroimaging in shows real, reproducible structural (enlarged ventricles, reduced ) and functional (hypofrontality) changes, but it is used clinically only to exclude organic mimics before a scan-independent ICD-11/DSM-5 clinical diagnosis and NICE-guided antipsychotic-led treatment plan.