The research, led by Yinan Li and colleagues at Kyoto University together with collaborators in Japan and China, tackled the problem by mapping the functional wiring between the dopaminergic midbrain and the striatum in a large sample of 116 patients with schizophrenia and 224 age- and sex-matched healthy controls. All participants underwent resting-state functional magnetic resonance imaging, a technique that measures spontaneous fluctuations in blood-oxygen-level-dependent signal while people simply lie still in the scanner. The team then computed functional connectivity between six subdivisions of the striatum—limbic, executive, and sensorimotor regions in each hemisphere—and individual voxels within the ventral tegmental area and substantia nigra, the two midbrain structures that house the brain’s dopamine-producing neurons.

The technical rigor behind this mapping deserves attention, because the midbrain is one of the hardest structures in the brain to image reliably. The ventral tegmental area and substantia nigra are small, densely packed nuclei, and their signals are easily contaminated by head motion, physiological noise, and susceptibility distortions. The researchers deployed an unusually comprehensive denoising pipeline combining FSL, FreeSurfer, and independent component analysis-based cleanup, used midbrain masks anatomically optimized for East Asian brains, harmonized data from two different 3-Tesla scanners with the ComBat statistical procedure, and deliberately avoided global signal regression to prevent artificial negative correlations from contaminating their connectivity gradients. Sensitivity analyses with different smoothing parameters confirmed that the results were robust.

The first major finding was a confirmation, for the first time in humans, of the canonical ventral-to-dorsal organization of the midbrain-striatal dopamine system. In animal work, tracing studies have shown that the ventral midbrain preferentially projects to the ventral striatum while more dorsal midbrain regions project to dorsal striatal territories. To test whether this holds in living humans, the team split the midbrain mask along two orthogonal axes—inferior versus superior and anterior versus posterior—and counted, for each striatal seed, the proportion of positively connected voxels falling in each half. Using beta regression, a statistical approach well suited to proportional data, they found that these bias measures declined linearly along both axes with extraordinary statistical significance, demonstrating that the ventral limbic striatum couples preferentially to the anterior and inferior midbrain, with progressively more dorsal striatal regions shifting their coupling posteriorly and superiorly.

With the map established, the team turned to the clinical question. When they compared diagnostic groups, they found that patients with schizophrenia showed significantly reduced functional connectivity between the left anterior midbrain and the left ventral limbic striatum—the classic mesolimbic pathway. This hypoconnectivity is consistent with a substantial body of prior resting-state imaging work in chronic schizophrenia, first-episode psychosis, and individuals at clinical high risk, and it persisted even after the researchers statistically controlled for antipsychotic medication doses, illness duration, education, premorbid intelligence, and smoking status. Notably, however, this ventral-limbic reduction did not correlate with the severity of any symptom domain measured by the Positive and Negative Syndrome Scale.

The pivotal discovery emerged when the researchers brought in the Aberrant Salience Inventory, a 29-item questionnaire that captures subjective experiences of aberrant salience, such as whether trivial things suddenly seem especially significant or whether one’s senses seem unusually sharpened. Drawing on a recently validated three-subscale model, they separated motivational salience, cognitive salience, and sensory salience, and tested how each interacted with diagnosis to shape midbrain-striatal connectivity. The striking result was a significant interaction involving the sensory salience subscale—the so-called Sharpening of Senses dimension—specifically for the connection between the left anteroinferior midbrain, corresponding to the lateral ventral tegmental area and ventromedial substantia nigra, and the left dorsal sensorimotor striatum. Not the ventral striatum. The dorsal one.

The nature of this interaction was as revealing as its location. In healthy controls, higher sensory salience scores were associated with stronger positive coupling between the ventral midbrain and the dorsal sensorimotor striatum, suggesting that in the healthy brain this circuit supports the adaptive, coordinated assignment of significance to sensory events. In patients with schizophrenia, the relationship reversed: higher sensory salience scores were associated with weaker, even negative coupling. This crossover pattern indicates not a simple amplification or dampening of the salience circuit, but a pathological desynchronization—a breakdown of the normal coordinated dynamics through which dopaminergic signals and striatal sensorimotor processing work in tandem. The researchers interpret this as the circuit-level signature of ordinary perceptions acquiring abnormal, personally meaningful significance, a process that can ultimately feed hallucinations and delusions.

Several additional analyses strengthened the claim that this finding is genuinely about aberrant salience rather than general psychiatric illness. The sensory salience subscale correlated significantly with positive symptoms, general psychopathology, and total symptom scores in patients. More compellingly, hierarchical regression showed that the subscale remained a robust independent predictor of midbrain-dorsal striatal connectivity even after controlling for these symptom measures, explaining an additional 17 to 24 percent of the variance in connectivity. In other words, the dysconnectivity tracked the subjective experience of aberrant salience more closely than it tracked overt symptom severity. The authors are careful to note, however, that the connectivity measure itself did not correlate directly with positive symptom scores, positioning it as a neural correlate of aberrant salience rather than of psychosis per se—a distinction that reflects the multifactorial nature of psychotic symptoms.

The broader significance of the study lies in how it reconciles a long-standing contradiction. Animal research has historically emphasized the ventral midbrain-to-ventral striatum pathway in dopamine dysregulation, while human positron-emission tomography studies have repeatedly found elevated dopamine synthesis capacity in the dorsal associative and sensorimotor striata of patients, not the ventral limbic striatum. By showing that the relevant circuit in humans runs from the ventral midbrain to the dorsal sensorimotor striatum—a projection pattern that animal tracer studies had actually documented but that human work had not previously linked to salience—the new findings bridge the two literatures. They also highlight an anatomical subtlety: unlike the strictly parallel corticostriatal system, the dopaminergic midbrain-striatal pathway includes overlapping, spiraling connections across the ventral-to-dorsal axis, and the rotation of the brain between quadrupedal animals and bipedal humans complicates any simple translation of ventral and dorsal labels across species.

The study is not without limitations. All patients were taking antipsychotic medication, which cannot be fully ruled out as a contributor despite the statistical controls; the blood-oxygen-level-dependent signal reflects a mixture of dopaminergic and GABAergic activity rather than dopamine alone; and the salience questionnaire was available for a subset of participants, 38 patients and 74 controls, which may have limited statistical power for the interaction analyses. Even so, the combination of a large overall sample, meticulous denoising, scanner harmonization, species-optimized midbrain masks, and converging sensitivity analyses makes this one of the most convincing demonstrations to date of where aberrant salience lives in the human brain. If replicated, the ventral midbrain-dorsal sensorimotor striatal circuit could serve as an imaging biomarker for aberrant salience, potentially enabling earlier identification of people at risk for psychosis and offering a more precise target for interventions aimed at quieting the brain’s tendency to shout meaning where there is none.

Subject of Research: Midbrain-striatal functional connectivity and aberrant salience in schizophrenia

Article Title: Aberrant salience is selectively associated with ventral midbrain connectivity to the dorsal but not ventral striatum

Article References: Li, Y., Oishi, N., Dai, Q., Nakagami, Y., Yao, L., Kawashima, T., Yoshihara, Y., Kubota, M., Nakamura, Y., Koike, S., Murai, T., & Miyata, J. (2026). Aberrant salience is selectively associated with ventral midbrain connectivity to the dorsal but not ventral striatum. Schizophrenia, 12(1), Article 71. https://doi.org/10.1038/s41537-026-00795-2

Image Credits: AI Generated

DOI: 10.1038/s41537-026-00795-2

Keywords: schizophrenia, aberrant salience, dopamine, ventral tegmental area, substantia nigra, striatum, resting-state fMRI, functional connectivity, psychosis, delusions, hallucinations, sensorimotor striatum