For much of modern medicine, diagnosis has progressively moved from describing what is happening to determining why it is happening. Oncology no longer stops at identifying the location of a tumor. Molecular testing can characterize mutations, receptors and other biological features that increasingly determine how a cancer is treated, while neurology is undergoing a similar transformation with PET tracers capable of identifying amyloid and tau pathology in the living brain.
Psychiatry remains different. Major depressive disorder, schizophrenia, post-traumatic stress disorder and many other psychiatric diseases continue to be diagnosed primarily through symptoms, behavior, clinical history and standardized diagnostic criteria. Those methods are essential and often highly effective, but they generally cannot tell us exactly which molecular processes inside a patient's brain are producing those symptoms. PET may eventually help change that.
PET is not simply an anatomical imaging technology. Depending on the radiotracer, it can measure regional glucose metabolism, neurotransmitter activity, receptor availability, inflammation and other biological processes inside the living human brain.
A 2025 review, PET Imaging in Psychiatric Disorders, examined the use of PET across schizophrenia, mood disorders, autism spectrum disorder, attention-deficit/hyperactivity disorder and addiction. The researchers described PET's ability to evaluate regional cerebral metabolism, neurotransmitter dynamics, receptor binding, synaptic density and neuroinflammation, while noting its potential role in future clinical decision-making. PET Imaging in Psychiatric Disorders
That distinction matters because a clinician can determine that a patient meets the diagnostic criteria for major depressive disorder, while PET research is beginning to ask a different question: what biological processes are occurring inside that patient's brain? Two people can meet the clinical definition of depression while potentially having different combinations of metabolic, neurotransmitter, inflammatory, synaptic and network abnormalities. If those differences eventually prove reproducible and clinically meaningful, the future of psychiatric diagnosis may involve characterizing biological phenotypes within diseases rather than assuming every patient carrying the same diagnostic label has the same underlying biology.
One of the most intriguing developments is PET imaging of synaptic vesicle glycoprotein 2A, or SV2A. SV2A is present in synaptic vesicles throughout the brain, making it possible to use radiotracers such as carbon-11 UCB-J to estimate synaptic density in vivo. That means scientists can begin examining something that historically could not readily be measured in a living patient's brain: whether psychiatric symptoms are associated with differences in the density of synaptic connections.
A major review, Synaptic Changes in Psychiatric and Neurological Disorders: State-of-the-Art of In Vivo Imaging, examined more than 50 clinical studies involving more than 1,700 participants. Researchers evaluated SV2A PET findings across depression, schizophrenia, PTSD, addiction and several neurological diseases. Read the SV2A PET review
Depression provides an especially interesting example. In one PET study using carbon-11 UCB-J, researchers examined 26 unmedicated people with major depressive disorder, PTSD or both and compared them with healthy controls. Greater depressive symptom severity was associated with lower SV2A density, while lower SV2A measurements were also associated with altered functional connectivity on MRI. Lower Synaptic Density Is Associated With Depression Severity and Network Alterations
A subsequent review of SV2A imaging described the technology as providing a method for in-vivo quantification of synaptic density and discussed its application to depression, including research investigating whether treatments such as ketamine can produce measurable synaptic changes. Challenges and Rewards of In Vivo Synaptic Density Imaging, and Its Application to the Study of Depression
None of this means that reduced SV2A signal is a diagnostic test for depression. It does show something arguably more important for the long-term development of the field: molecular imaging can measure biological characteristics associated with psychiatric symptoms in living patients.
Inflammation represents another potential biological layer. Researchers have extensively investigated the 18-kDa translocator protein, or TSPO, as a PET target related to activation of microglia and astrocytes. A 2025 review, PET Imaging Unveils Neuroinflammatory Mechanisms in Psychiatric Disorders, examined TSPO PET findings across major depressive disorder, obsessive-compulsive disorder, PTSD, schizophrenia and psychosis. Read the neuroinflammation PET review
The science here requires caution because TSPO is not a simple, disease-specific marker of “brain inflammation,” and findings have varied between diseases, patient populations, tracers and methodologies. Earlier reviews have similarly emphasized that inflammatory biomarkers need to be validated separately for individual disorders before they could realistically be used for patient stratification or treatment monitoring. Neuroinflammation in Psychiatric Disorders: PET Imaging and Promising New Targets
That uncertainty should not obscure the larger development. Psychiatric research is increasingly capable of asking whether an individual patient's illness involves measurable neuroimmune biology rather than inferring those processes exclusively from symptoms or peripheral biomarkers.
PTSD offers another example of why molecular phenotyping may ultimately matter. Two people can experience similar traumatic events while only one develops PTSD, raising a difficult clinical question: what biological differences distinguish trauma exposure from the disease that sometimes follows it?
A 2026 systematic review, Neuromarkers of Post-Traumatic Stress Disorder: A Systematic Review of Positron Emission Tomography Studies, examined 20 PET studies involving 483 patients with PTSD and 522 controls, including both healthy individuals and trauma-exposed people who had not developed PTSD. Read the PTSD PET systematic review
The review found evidence of altered glucose metabolism involving several brain regions as well as abnormalities involving neurotransmitter, stress and inflammatory systems. The authors concluded that PET shows promise for improving understanding of PTSD biology and potentially informing diagnosis, treatment monitoring and personalized interventions, although the evidence is not yet sufficient to establish a routine diagnostic biomarker.
That distinction is critical because the future may not be a PET scan that simply says PTSD: positive or PTSD: negative. The more consequential possibility is identifying biological patterns that help explain why different patients develop different manifestations of what we currently classify as the same disease.
There is another area where the clinical relevance is becoming more immediate: determining whether symptoms attributed to psychiatric disease actually reflect underlying neurodegeneration.
A study evaluating FDG-PET in 98 patients with depression and cognitive impairment found that disclosure of PET results changed the broad clinical diagnosis in 23% of patients. Diagnostic accuracy for differentiating neurodegenerative from non-neurodegenerative disease increased from 72% before PET to 92% after PET information was incorporated. FDG-PET for Differential Diagnosis of Depressive Cognitive Impairment
PET was not diagnosing depression. It was helping determine whether symptoms presenting alongside depression represented something biologically different, which may provide an early glimpse of how molecular imaging eventually enters psychiatry—not by replacing psychiatrists or clinical assessment, but by adding biological information to difficult diagnostic decisions.
Treatment Response Could Be Just as Important as Diagnosis
PET's opportunity may extend beyond determining what disease a patient has. It may eventually help determine what treatment is working, how it is working and perhaps which patients are most likely to respond.
A 2026 systematic review of PET imaging in psychiatric patients receiving deep-brain stimulation identified 27 studies involving major depression, OCD, substance-use disorder, anorexia nervosa and schizophrenia. Researchers used tracers including FDG to evaluate glucose metabolism, oxygen-15 water to measure cerebral blood flow and carbon-11 raclopride to investigate dopamine transmission. Deep Brain Stimulation for Psychiatric Disorders: A Systematic Review of Molecular Imaging With PET
Changes measured with PET generally correlated with improvements in psychiatric symptom scores, although the researchers cautioned that the clinical significance of some observed patterns remains uncertain. They nevertheless concluded that PET could potentially contribute to patient selection, prognosis and long-term monitoring of deep-brain stimulation.
That may ultimately be one of molecular imaging's most important contributions to psychiatry. Instead of waiting weeks or months and asking only whether a patient reports feeling better, future researchers may be able to observe whether a treatment is producing the biological change it was intended to produce.
There are substantial obstacles before any of this becomes routine clinical medicine. Psychiatric illnesses are extraordinarily heterogeneous. PET studies are often relatively small, radiotracers differ between research centers, medication can affect neurochemistry, biomarkers can overlap across diseases, and an abnormal molecular signal does not automatically establish causation.
PET is also expensive, requires specialized infrastructure and, depending on the tracer, can depend on nearby radiopharmaceutical production. Any psychiatric application would therefore need to provide meaningful clinical information that could not be obtained more easily or less expensively through other approaches. For those reasons, I do not expect psychiatry to suddenly acquire a collection of PET scans labeled depression, schizophrenia, OCD and PTSD. I think the opportunity is considerably more sophisticated than that.
Imagine two patients who both meet the diagnostic criteria for major depressive disorder. Today they may receive similar diagnoses despite having very different underlying biology. In a future molecular psychiatry model, one patient's disease could be associated predominantly with abnormalities in neurotransmitter signaling, another with altered synaptic density, another with a particular neuroimmune phenotype and another with metabolic or network abnormalities.
Those differences could eventually matter more than the diagnostic label itself if they predict which therapy is most likely to work. Oncology has already moved substantially in this direction. The question is increasingly not simply where is the cancer? It is what biological characteristics does this cancer possess?
Neurology is moving in the same direction. Molecular imaging can increasingly help answer not simply whether someone is experiencing cognitive decline, but whether disease-associated pathology such as amyloid or tau is present. Psychiatry may eventually face the same transition, moving from asking only what symptoms does this patient have? toward asking what biology is producing them?
That is not where psychiatric PET is today. Much of the evidence remains investigational, and no PET scan should currently be interpreted as a standalone molecular diagnosis of depression, schizophrenia, PTSD or most other psychiatric disorders. But the tools required to begin answering those biological questions are emerging.
PET can measure metabolism, interrogate neurotransmitter systems and receptor binding, image synaptic density and investigate neuroimmune activity. Increasingly, researchers are attempting to connect those measurements with symptoms, disease subtypes and treatment response. If those relationships become reproducible enough to influence patient care, nuclear medicine will have entered one of the largest areas of medicine it has yet to meaningfully penetrate.
Psychiatry may ultimately become less about categorizing symptoms and more about identifying the biology behind them. If that happens, PET could become one of the technologies that helps medicine make that transition.
