Pulvinar-cortical connectomic deviations identify symptom-relevant biotypes in schizophrenia

Schizophrenia exhibits profound clinical heterogeneity, yet biologically interpretable imaging markers capturing symptom-specific variance remain elusive. Although the pulvinar is a critical higher-order thalamic hub for cortico-cortical coordination, whether its individualized abnormalities can define clinically meaningful biotypes remains unexplored. The present study implemented a symptom-guided normative modeling framework using multisite resting-state functional MRI data from 746 healthy controls and 387 patients with schizophrenia.

Individualized pulvinar-cortical deviation profiles were estimated relative to a healthy reference model and integrated with symptom-guided feature selection and unsupervised clustering. Two partially separable biotypes were identified within a clinically enriched low-dimensional feature space. The biotypes were clinically anchored primarily in negative symptom severity, with a significant between-biotype difference that remained after adjustment for site and other available covariates.

Symptom-relevant connections showed a structured cortical distribution concentrated predominantly in somatomotor, visual, and ventral attention networks. Imaging-transcriptomic analyses provided exploratory spatial molecular context, with Biotype 1 showing enrichment profiles involving translation, protein targeting, and synaptic organization, and Biotype 2 showing profiles involving neurodevelopment, synaptic remodeling, and metal-ion homeostasis. Site-exclusion analyses indicated partial stability of the biotype solution, although clinical separation varied across sites.

Projection into an independent repetitive transcranial magnetic stimulation cohort showed nominal, uncorrected differences in 2-week negative symptom improvement that were not sustained at 4 weeks and should therefore be considered hypothesis-generating. These findings support a symptom-guided pulvinar-cortical framework for characterizing schizophrenia heterogeneity, while independent replication using prespecified models is required to establish clinical and biological validity. This work was supported by the National Natural Science Foundation of China (82671963;62476178), Jianghuai Talent Training Program team project (2025), Annual Artificial Intelligence Scenario Innovation Project of Anhui Province (2026), Shanghai Key Laboratory of Forensic Medicine and Key Laboratory of Forensic Science, Ministry of Justice (KF202618), Jiangsu Natural Science Foundation Youth Project (BK2024056), and Wuhu Science and Technology Program (2025kj011).

School of Medicine, Sichuan Science and Technology University, Zigong, China Department of Psychiatry, Wuhu Hospital of Anding Hospital (The Fourth People’s Hospital of Wuhu), Wuhu, China Department of Mental Health, Xi’an Medical College, Xi’an, China Department of Psychiatry, Guangyuan Mental Health Center, Guangyuan, China Department of Psychiatry, Xijing Hospital, Fourth Military Medicine University, Xi’an, China Xizang Autonomous Region Key Laboratory for High Altitude Brain Science and Environmental Acclimatization, Xizang University, Lhasa, China Department of Psychiatry, Zhenjiang Mental Health Center, Zhenjiang, China The authors declare no competing interests. All study protocols and data collection procedures were approved by the local Institutional Review Boards or ethics committees at each respective contributing site. The research was conducted in strict accordance with the principles of the Declaration of Helsinki.

Written informed consent was obtained from all participants or their legal representatives prior to data collection at their respective local sites. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material.