Center for Translational Research in Neuroimaging and Data Science

Artificial intelligence is increasingly advancing both fundamental research and clinical applications in schizophrenia. This review surveys recent literature on artificial intelligence driven approaches for schizophrenia diagnosis, treatment, management, and characterization, using multiple data modalities such as neuroimaging, electrophysiology, electronic health records, and genomic data. Recent work shows substantial progress in leveraging machine learning and deep learning for diagnostic label prediction, treatment response modeling, and brain network characterization. While many studies continue to improve feature extraction and classification methods within single modalities, there is a growing trend to utilize multiple data sources to capture the complexity of schizophrenia from a comprehensive perspective. Emerging themes include multimodal fusion methodologies to identify linked correlates of schizophrenia, as well as data-driven approaches to learn subgroups, brain networks, and psychosis continua. The rise of large-scale multimodal datasets, foundation models, and mechanistic interpretability methods holds promise for scalable symptom assessment and biomarker identification, thereby better supporting early intervention and personalized treatment. Current literature highlights a shift from unimodal prediction to holistic, multimodal characterization of schizophrenia. Transforming these artificial intelligence models into clinical tools, however, requires careful attention to patient privacy and data bias, alongside rigorous validation across diverse populations and settings.

Historically, neural variability observed during task was interpreted as “noise,” assumed to obscure meaningful signal and thus something to be minimized both analytically by researchers and functionally by the brain. Changes to this signal-to-noise ratio have been proposed as a possible neural mechanism behind the increased reaction-time variability (RTV) in attention deficit hyperactivity disorder (ADHD). However, not all variability is the same – in some cases, variability can have some underlying “statistical structure” that can be beneficial to information processing. The challenge lies in distinguishing meaningful variability from random noise. The edge-of-synchrony critical point, which describes a system poised between synchronous and asynchronous regimes, could be a good theoretical framework to study these different types of neural variability. In this study, we investigate whether changes in criticality and oscillatory dynamics preceded slower behavioral responses during a bimodal continuous performance task in ADHD. We find evidence that, prior to slower responses, neural dynamics shift toward criticality in both ADHD and control groups, suggesting that increase variability in ADHD and during attention lapses are related to structured variability and not necessarily random noise. Notably, these findings run counter predictions based on the proposed model and previous literature on neural noise in this population, challenging predictions of edge-of-synchrony criticality as a unifying account of neural variability and behavioral performance. Furthermore, this effect did not emerge at the between-subject level, underscoring the limitations of relying on between-subject correlations to infer neural mechanisms.

Abstract Background Adolescence is a critical development stage, and adverse childhood experiences (ACEs) are important environmental factors influencing behavior. However, the relationship between ACEs, individual behavior, brain mechanisms, and the role of genetic risk remains underexplored. Methods We analyzed data from 9,796 participants aged 9–13 years from the Adolescent Brain Cognitive Development (ABCD) study. Sixty-seven ACE-related variables covering caregiving disruption, family functioning, neighborhood environment, and safety were examined. Sparse canonical correlation analysis (sCCA) and multiple sparse canonical correlation analysis (msCCA) identified multivariate ACE–behavior–brain patterns. Moderated mediation was used to test genetic risk modulation, and a cross-lagged panel model examined longitudinal associations. Results Two robust modes emerged. Mode 1 r eflected parental mental health difficulties linked to child mental health ( r = 0.576, P perm < 0.001). Mode 2 represented socioeconomic and family environment factors associated with cognitive performance ( r = -0.272, P perm < 0.008). Neuroimaging revealed convergent alterations in hippocampal and cerebellar volumes, white-matter integrity, and large-scale functional networks. Polygenic risk for psychiatric disorders significantly moderated ACE–brain–behavior pathways, with effects varying across ancestry groups. Longitudinal analyses showed that parental depression and ADHD strongly predicted youth psychiatric symptoms two years later. Conclusions Two distinct profiles of ACEs influence adolescent behavior through neuroimaging pathways, and these associations are further shaped by genetic risk. Findings advance mechanistic understanding of ACE-related vulnerability and highlight intervention targets for at-risk youth.

ABSTRACT Bullying is an adverse childhood experience affecting up to one-third of the global population and linked to psychosis-like experiences (PLEs), which increase the risk of psychotic disorders. This study aimed to investigate the association between the severity and persistence of bullying and PLEs and the neurobiological pathways from bullying to psychosis-like experiences by assessing multiscale brain functional network connectivity (msFNC). We used data from the ABCD Study, a large, ongoing, multisite, population-based prospective cohort study following U.S. adolescents. We included adolescents with complete bullying and PLEs assessments at the 2- and 3-year follow-ups (T1: n=10,939; T2: n=10,102). We examined the associations between bullying severity and temporal exposure and PLEs using linear mixed-effects models. In a 2-year rsfMRI subsample (n=5,280), we used a Neuromark framework to analyze whether msFNC mediated the pathway from bullying to PLEs. Higher PLEs were associated with the presence and severity of bullying (non-bullied vs. mild bullying: d=-0.19, CI: −0.39 to −0.19, p<0.0001; moderate vs. severe bullying: d=-0.49, CI: −0.69 to −0.56, p<0.0001). When bullying ceased, PLEs returned to non-bullied levels (d=-0.13, CI:-0.20 to –0.05, p=0.16), whereas persistence over two years led to greater elevations (d=-0.36, CI:-0.43 to –0.29, p<0.0001). We observed similar patterns for non-paranoid and hallucination-like experiences and their distress. msFNC in paralimbic, default mode, central executive, somatomotor, temporoparietal, insulotemporal, and frontal networks mediated the association. Bullying is time- and dose-dependently associated with psychosis-like outcomes. msFNC between functional brain networks is a novel neurobiological pathway that mediates the link from bullying to PLEs.

Emerging studies have consistently shown that blood-oxygenation-level-dependent signals within white matter (WM) appear to be related to neural activity. This recognition of WM functional signal has provided a great opportunity to investigate how the brain's WM functionally evolves in youth as well as its relationship to developmental disorders. Using multimodal imaging data from 407 participants aged 8 to 22 years, we constructed the WM functional connectome (WMFC) as well as the WM structural connectome (WMSC). Our findings reveal a global three-stage reorganization process in WMFC characterized by a decreasing-increasing-decreasing pattern. Regional changes are marked by hierarchical development between primary networks and higher-order networks. In conjunction with the WMSC, we measured WM SC‒FC coupling, and found that coupling monotonically decreases with development. Treating these spatial patterns as phenotypes of WM function, we demonstrated significant regulation by neurotransmitter receptors and transcriptomic expression. Notably, WMFC significantly explained the heterogeneous regional vulnerability in neurodevelopmental disorders and outperformed the WMSC in predicting cognitive function. This study underscores the importance of investigating WM function in understanding the neural mechanisms of brain maturation and disorders.

Introduction functional impairments in schizophrenia may arise from disruptions in large-scale brain networks. Emerging evidence highlights the cerebellum’s role in cognitive and affective regulations, yet its directional influence remains poorly understood. This study examines effective connectivity (EC) within cortico-striato-cerebellar networks in schizophrenia and healthy adults. Methods resting-state functional magnetic resonance imaging data from the Centers of Biomedical Research Excellence (COBRE), including people with schizophrenia and healthy controls (n = 134), were used to analyze intrinsic activity and effective connectivity. Cerebellar clusters showing reduced amplitude of low frequency fluctuations (ALFF) and fractional ALFF in schizophrenia were mapped to the Buckner 17-network atlas to define regions of interest (ROIs). Along with prefrontal and striatal ROIs defined a priori, these served as nodes in Group Iterative Multiple Model Estimation (GIMME). In addition to the best-fitting model of the full sample, schizophrenia participants were stratified by symptomology according to the positive and negative syndrome scale. Results four cerebellar voxel clusters in the posterolateral and anteromedial regions showed lower ALFF/fALFF in schizophrenia compared to controls. GIMME revealed distinct intra-cerebellar and cerebello-prefrontal EC patterns in schizophrenia that were absent in controls, including novel directed paths involving the cerebellar control network representation, consistent with its emergence as a central hub. These patterns persisted across symptom-defined subgroups, suggesting core network reorganization. Conclusions findings support the cerebellum’s involvement in disrupted network dynamics in schizophrenia, particularly its directional influence over prefrontal targets. EC analyses uncovered cerebellar reorganization that may underlie affective deficits in schizophrenia, offering novel targets for circuit-level interventions.Supplementary InformationThe online version contains supplementary material available at 10.1007/s12311-026-01983-7.