Multisite magnetic resonance imaging (MRI) is increasingly used in clinical research and development. Measurement biases—caused by site differences in scanner/image‐acquisition protocols—negatively influence the reliability and reproducibility of image‐analysis methods. Harmonization can reduce bias and improve the reproducibility of multisite datasets. Herein, a traveling‐subject (TS) dataset including 56 T1‐weighted MRI scans of 20 healthy participants in three different MRI procedures—20, 19, and 17 subjects in Procedures 1, 2, and 3, respectively—was considered to compare the reproducibility of TS‐GLM, ComBat, and TS‐ComBat harmonization methods. The minimum participant count required for harmonization was determined, and the Cohen's d between different MRI procedures was evaluated as a measurement‐bias indicator. The measurement‐bias reduction realized with different methods was evaluated by comparing test–retest scans for 20 healthy participants. Moreover, the minimum subject count for harmonization was determined by comparing test–retest datasets. The results revealed that TS‐GLM and TS‐ComBat reduced measurement bias by up to 85 and 81.3%, respectively. Meanwhile, ComBat showed a reduction of only 59.0%. At least 6 TSs were required to harmonize data obtained from different MRI scanners, complying with the imaging protocol predetermined for multisite investigations and operated with similar scan parameters. The results indicate that TS‐based harmonization outperforms ComBat for measurement‐bias reduction and is optimal for MRI data in well‐prepared multisite investigations. One drawback is the small sample size used, potentially limiting the applicability of ComBat. Investigation on the number of subjects needed for a large‐scale study is an interesting future problem.

Machine learning approaches using structural magnetic resonance imaging (sMRI) can be informative for disease classification, although their ability to predict psychosis is largely unknown. We created a model with individuals at CHR who developed psychosis later (CHR-PS+) from healthy controls (HCs) that can differentiate each other. We also evaluated whether we could distinguish CHR-PS+ individuals from those who did not develop psychosis later (CHR-PS-) and those with uncertain follow-up status (CHR-UNK). T1-weighted structural brain MRI scans from 1165 individuals at CHR (CHR-PS+, n = 144; CHR-PS-, n = 793; and CHR-UNK, n = 228), and 1029 HCs, were obtained from 21 sites. We used ComBat to harmonize measures of subcortical volume, cortical thickness and surface area data and corrected for non-linear effects of age and sex using a general additive model. CHR-PS+ (n = 120) and HC (n = 799) data from 20 sites served as a training dataset, which we used to build a classifier. The remaining samples were used external validation datasets to evaluate classifier performance (test, independent confirmatory, and independent group [CHR-PS- and CHR-UNK] datasets). The accuracy of the classifier on the training and independent confirmatory datasets was 85% and 73% respectively. Regional cortical surface area measures-including those from the right superior frontal, right superior temporal, and bilateral insular cortices strongly contributed to classifying CHR-PS+ from HC. CHR-PS- and CHR-UNK individuals were more likely to be classified as HC compared to CHR-PS+ (classification rate to HC: CHR-PS+, 30%; CHR-PS-, 73%; CHR-UNK, 80%). We used multisite sMRI to train a classifier to predict psychosis onset in CHR individuals, and it showed promise predicting CHR-PS+ in an independent sample. The results suggest that when considering adolescent brain development, baseline MRI scans for CHR individuals may be helpful to identify their prognosis. Future prospective studies are required about whether the classifier could be actually helpful in the clinical settings.

Although brain morphological abnormalities have been reported in anorexia nervosa (AN), the reliability and reproducibility of previous studies were limited due to insufficient sample sizes, which prevented exploratory analysis of the whole brain as opposed to regions of interest (ROIs). Objective was to identify brain morphological abnormalities in AN and the association with severity of AN by brain structural magnetic resonance imaging (MRI) in a multicenter study, and to conduct exploratory analysis of the whole brain. Here, we conducted a cross-sectional multicenter study using T1-weighted imaging (T1WI) data collected between May 2014 and February 2019 in Japan. We analyzed MRI data from 103 female AN patients (58 anorexia nervosa restricting type [ANR] and 45 anorexia nervosa binge-purging type [ANBP]) and 102 age-matched female healthy controls (HC). MRI data from five centers were preprocessed using the latest harmonization method to correct for intercenter differences. Gray matter volume (GMV) was calculated from T1WI data of all participants. Of the 205 participants, we obtained severity of eating disorder symptom scores from 179 participants, including 87 in the AN group (51 ANR, 36 ANBP) and 92 HC using the Eating Disorder Examination Questionnaire (EDE-Q) 6.0. GMV reduction were observed in the AN brain, including the bilateral cerebellum, middle and posterior cingulate gyrus, supplementary motor cortex, precentral gyrus medial segment, and thalamus. In addition, the orbitofrontal cortex (OFC), ventromedial prefrontal cortex (vmPFC), rostral anterior cingulate cortex (ACC), and posterior insula volumes showed positive correlations with severity of symptoms. This multicenter study was conducted with a large sample size to identify brain morphological abnormalities in AN. The findings provide a better understanding of the pathogenesis of AN and have potential for the development of brain imaging biomarkers of AN. Trial Registration: UMIN000017456. https://center6.umin.ac.jp/cgi-open-bin/icdr/ctr_view.cgi?recptno=R000019303.

Although widespread cortical thinning centered on the fronto-temporal regions in schizophrenia has been reported, the findings in at-risk mental state (ARMS) patients have been inconsistent. In addition, it remains unclear whether abnormalities of cortical thickness (CT) in ARMS individuals, if present, are related to their functional decline irrespective of future psychosis onset. In this multicenter study in Japan, T1-weighted magnetic resonance imaging was performed at baseline in 107 individuals with ARMS, who were subdivided into resilient (77, good functional outcome) and non-resilient (13, poor functional outcome) groups based on the change in Global Assessment of Functioning scores during 1-year follow-up, and 104 age- and sex-matched healthy controls recruited at four scanning sites. We measured the CT of the entire cortex and performed group comparisons using FreeSurfer software. The relationship between the CT and cognitive functioning was examined in an ARMS subsample (n = 70). ARMS individuals as a whole relative to healthy controls exhibited a significantly reduced CT, predominantly in the fronto-temporal regions, which was partly associated with cognitive impairments, and an increased CT in the left parietal and right occipital regions. Compared with resilient ARMS individuals, non-resilient ARMS individuals exhibited a significantly reduced CT of the right paracentral lobule. These findings suggest that ARMS individuals partly share CT abnormalities with patients with overt schizophrenia, potentially representing general vulnerability to psychopathology, and also support the role of cortical thinning in the paracentral lobule as a predictive biomarker for short-term functional decline in the ARMS population.

Although many studies have demonstrated structural brain abnormalities associated with auditory verbal hallucinations (AVH) in schizophrenia, the results remain inconsistent because of the small sample sizes and the reliability of clinical interviews. We compared brain morphometries in 204 participants, including 58 schizophrenia patients with a history of AVH (AVH + ), 29 without a history of AVH (AVH−), and 117 healthy controls (HCs) based on a detailed inspection of medical records. We further divided the AVH+ group into 37 patients with and 21 patients without hallucinations at the time of the MRI scans (AVH++ and AVH+−, respectively) via clinical interviews to explore the morphological differences according to the persistence of AVH. The AVH + group had a smaller surface area in the left caudal middle frontal gyrus (F = 7.28, FDR-corrected p = 0.0008) and precentral gyrus (F = 7.68, FDR-corrected p = 0.0006) compared to the AVH− group. The AVH+ patients had a smaller surface area in the left insula (F = 7.06, FDR-corrected p = 0.001) and a smaller subcortical volume in the bilateral hippocampus (right: F = 13.34, FDR-corrected p = 0.00003; left: F = 6.80, FDR-corrected p = 0.001) compared to the HC group. Of these significantly altered areas, the AVH++ group showed significantly smaller bilateral hippocampal volumes compared to the AVH+− group, and a smaller surface area in the left precentral gyrus and caudal middle frontal gyrus compared to the AVH- group. Our findings highlighted the distinct pattern of structural alteration between the history and presence of AVH in schizophrenia, and the importance of integrating multiple criteria to elucidate the neuroanatomical mechanisms.