EPA-1579 - Classifying schizophrenia using joint multivariate pattern recognition analysis of brain function and structure

Previous studies have shown that structural brain changes are among the best-studied candidate markers for schizophrenia (SZ) along with functional connectivity (FC) alterations of resting-state (RS) patterns. This study aimed to investigate effects of clinical and sociodemographic variables on the classification by applying multivariate pattern analysis (MVPA) to both gray matter (GM) volume and FC measures in patients with SZ and healthy controls (HC). RS and structural magnetic resonance imaging data (sMRI) from 74 HC and 71 SZ patients were obtained from a Mind Research Network COBRE dataset available via COINS (http://coins.mrn.org/dx). We used a MVPA framework using support-vector machines embedded in a repeated, nested cross-validation to generate a multi-modal diagnostic system and evaluate its generalizability. The dependence of neurodiagnostic performance on clinical and sociodemographic variables was evaluated. The RS classifier showed a slightly higher accuracy (70.5%) compared to the structural classifier (69.7%). The combination of sMRI and RS outperformed single MRI modalities classification by reaching 75% accuracy. The RS based moderator analysis revealed that the neurodiagnostic performance was driven by older SZ patients with an earlier illness onset and more pronounced negative symptoms. In contrast, there was no linear relationship between the clinical variables and neuroanatomically derived group membership measures. This study achieved higher accuracy distinguishing HC from SZ patients by fusing 2 imaging modalities. In addition the results of RS based moderator analysis showed that age of patients, as well as their age at the illness onset were the most important clinical features.

Hallucinations are considered one of the primary symptoms of psychosis; however, their neural basis remains unclear. Voxel-based morphometry (VBM)1 studies have associated hallucinations with gray matter (GM) volume reduction in the superior temporal gyrus in schizophrenia.2 Although hallucinations are associated with brain connectivity,3 few studies have investigated their relationship with white matter (WM) volume. Moreover, these studies have shown inconsistent results.4, 5 The chronicity of illness and long-term medications for schizophrenia may limit the interpretation of previous findings.2 Thus, hallucination studies in clinical high risk for psychosis (CHR) and first-episode psychosis (FEP) are required, with minimal effects of chronicity of illness and medication. However, no VBM study has examined the relationship between hallucinations and WM volume in CHR and FEP. Therefore, we aimed to investigate the regional GM and WM volumes associated with hallucinations in CHR and FEP. We enrolled 57 individuals with CHR, 50 individuals with FEP, and 33 healthy controls (HCs). The Comprehensive Assessment of At-Risk Mental States-Japanese version6 was used to confirm whether ultra-high risk criteria7 were met. Table S1 summarizes the demographic and clinical data of the patients. Hallucination severity was assessed using the Positive and Negative Syndrome Scale (PANSS)8 P3 score. Fig. S1 shows the distribution of the severity of hallucinations. Some of those with a PANSS P3 score of 3 or higher were classified as having CHR because of the low frequency and duration of hallucinations. Magnetic resonance imaging (MRI) data were acquired with a 1.5-T Philips scanner. We examined the association of GM and WM volume with hallucination severity in the group that combined patients with CHR and FEP (CHR/FEP combined group) and performed group comparisons of GM and WM volume between the CHR/FEP combined and HC groups. Age, sex, and intracranial volume were entered as nuisance covariates for each analysis. VBM analysis was conducted using DARTEL9 in SPM12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) with standard smoothing of 8-mm full width at half maximum (FWHM). The initial statistical threshold was set at P < 0.001, uncorrected at the voxel level. The extent threshold for cluster size was set based on the expected number of voxels per cluster provided by SPM. Familywise error (FWE) correction at the cluster level (P < 0.05) was applied to identify significant clusters. Appendix S1 provides detailed information on these methods. This study was performed in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Tohoku University Graduate School of Medicine and Tohoku University Hospital. Written informed consent was obtained from participants aged 18 years or older and from the parents of those under 18 years of age, with written assent from the participants. The anonymity of participants was preserved. WM volume was not significantly correlated with hallucination severity in the CHR/FEP combined group after FWE correction at the cluster level. Reanalysis with 15-mm or 20-mm FWHM filtering did not show significant associations. Additionally, there was no significant difference in the GM and WM volumes between the CHR/FEP combined and HC groups. No regional GM volume abnormalities were significantly correlated with hallucination severity in the CHR/FEP combined group. Appendix S1, Figs. S2–S5, and Tables S2–S5 provide uncorrected results and detailed information regarding these results. To the best of our knowledge, this is the first VBM study to investigate WM volume related to hallucinations in CHR and FEP. We found no significant correlation between WM volume and hallucination severity. In contrast to previous studies in schizophrenia,2, 4, 5 this study examined the brain volume association of hallucinations in CHR and FEP. CHR and FEP are heterogeneous groups in the early stages of mental illness and do not necessarily convert to schizophrenia. Our results indicate that brain disturbances involved in hallucinations in CHR and FEP may be more subtle and/or inconsistent with those in schizophrenia. Diffusion tensor imaging, a neuroimaging method applied to CHR,10 may be suitable for examining structural abnormalities in the development of hallucinations. This study has some limitations. First, information regarding hallucinations is not detailed. Second, the cross-sectional design limited the interpretation of our findings. Third, we used an MRI scanner with a lower field strength than that of recent studies.4 Fourth, there were significant age differences between the groups. Fifth, automatic quality assurance for MRI scans was not performed. In conclusion, advanced neuroimaging studies are needed to reveal the mechanism of hallucinations. This work was supported by JSPS KAKENHI (grant number 18K07586). We thank the participants, staff of the SAFE clinic, and radiological technologists. The authors declare no conflicts of interest. Appendix S1. Supporting information Figure S1. Distribution of hallucination severity. CHR, clinical high risk for psychosis; FEP, first-episode psychosis; PANSS, Positive and Negative Syndrome Scale. Figure S2. Brain regions with lower gray matter volume in the CHR/FEP group compared to the HC group. Statistical threshold was set at P < 0.001, uncorrected at the voxel level. The threshold extent for cluster size was set to k = 117 based on the expected number of voxels per cluster provided by SPM. The full width at half maximum was set to 8-mm. CHR, clinical high risk for psychosis; FEP, first-episode psychosis; HC, healthy controls; SPM, Statistical Parametric Mapping Figure S3. Brain regions with lower white matter volume in the CHR/FEP group compared to the HC group. Statistical threshold was set at P < 0.001, uncorrected at the voxel level. The threshold extent for cluster size was set to k = 129 based on the expected number of voxels per cluster provided by SPM. The full width at half maximum was set to 8-mm. CHR, clinical high risk for psychosis; FEP, first-episode psychosis; HC, healthy controls; SPM, Statistical Parametric Mapping Figure S4. Positive correlation between gray matter volume and hallucination severity in the patients at clinical high risk for psychosis. Statistical threshold was set at P < 0.001, uncorrected at the voxel level. The threshold extent for cluster size was set to k = 101 based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. SPM, Statistical Parametric Mapping Figure S5. Negative correlation between white matter volume and hallucination severity in the patients at clinical high risk for psychosis. Statistical threshold was set at P < 0.001, uncorrected at the voxel level. The threshold extent for cluster size was set to k = 121 based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. SPM, Statistical Parametric Mapping Table S1. Demographic and clinical characteristics of the participants. †Educational level, CP-equivalent dose on MRI scan, PANSS positive, PANSS negative, PANSS general, PANSS total, PANSS P3: hallucinatory behavior, and days from enrollment to MRI scan were analyzed using the t-test. Age and intracranial volume were analyzed using one-way analysis of variance. Sex and antipsychotic use were analyzed using the chi-squared test. Significant results are shown in bold. CHR, clinical high risk for psychosis; CP-equivalent, chlorpromazine equivalent; FEP, first-episode psychosis; HC, healthy controls; PANSS, Positive and Negative Syndrome Scale. Table S2. Between-group comparison of regional gray matter volume. CHR/FEP < HC (extent threshold: 117 voxels). † Uncorrected p-value at voxel level. ‡ familywise error-corrected p-value at the cluster level. k, cluster size at P < 0.001 (uncorrected); CHR, clinical high risk for psychosis; FEP, first-episode psychosis; HC, healthy controls; L/R, left/right hemisphere; MNI, Montreal Neurological Institute. The extent threshold for cluster size was set based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. Table S3. Between group comparison of regional white matter volume. CHR/FEP < HC (extent threshold: 129 voxels). † Uncorrected P-value at voxel level. ‡ familywise error-corrected p-value at the cluster level. k, cluster size at P < 0.001 (uncorrected); CHR, clinical high risk for psychosis; FEP, first-episode psychosis; HC, healthy controls; L/R, left/right hemisphere; MNI, Montreal Neurological Institute. The extent threshold for cluster size was set based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. Table S4. Correlation between gray matter volume and hallucination severity in the patients with CHR. Positive correlation (extent threshold: 101 voxels). † Uncorrected p-value at voxel level. ‡ familywise error-corrected p-value at the cluster level. k, cluster size at P < 0.001 (uncorrected); CHR, clinical high risk for psychosis; L/R, left/right hemisphere; MNI, Montreal Neurological Institute. The extent threshold for cluster size was set based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. Table S5. Correlation between white matter volume and hallucination severity in the patients with CHR. Negative correlation (extent threshold: 121 voxels). † Uncorrected p-value at voxel level. ‡ familywise error-corrected p-value at the cluster level. k, cluster size at P < 0.001 (uncorrected); CHR, clinical high risk for psychosis; L/R, left/right hemisphere; MNI, Montreal Neurological Institute. The extent threshold for cluster size was set based on the expected number of voxels per cluster provided by the SPM. The full width at half maximum was set to 8-mm. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Gray matters: ViT-GAN framework for identifying schizophrenia biomarkers linking structural MRI and functional network connectivity

IntroductionAlthough schizophrenia (SCZ) and bipolar disorder (BD) share elements of pathology, their neural underpinnings are still under investigation. Here, structural Magnetic Resonance Imaging (MRI) data collected from a large sample of BD and SCZ patients and healthy controls (HC) were analyzed in terms of gray matter volume (GMV) using both voxel based morphometry (VBM) and a region of interest (ROI) approach.MethodsThe analysis was conducted on two datasets, Dataset1 (802 subjects: 243 SCZ, 176 BD, 383 HC) and Dataset2, a homogeneous subset of Dataset1 (301 subjects: 107 HC, 85 BD and 109 SCZ). General Linear Model analyses were performed 1) at the voxel-level in the whole brain (VBM study), 2) at the regional level in the anatomical regions emerged from the VBM study (ROI study). The GMV comparison across groups was integrated with the analysis of GMV correlates of different clinical dimensions.ResultsThe VBM results of Dataset1 showed 1) in BD compared to HC, GMV deficits in right cingulate, superior temporal and calcarine cortices, 2) in SCZ compared to HC, GMV deficits in widespread cortical and subcortical areas, 3) in SCZ compared to BD, GMV deficits in insula and thalamus (p<0.05, cluster family wise error corrected). The regions showing GMV deficits in the BD group were mostly included in the SCZ ones. The ROI analyses confirmed the VBM results at the regional level in most of the clusters from the SCZ vs. HC comparison (p<0.05, Bonferroni corrected). The VBM and ROI analyses of Dataset2 provided further evidence for the enhanced GMV deficits characterizing SCZ. Based on the clinical-neuroanatomical analyses, we cannot exclude possible confounding effects due to 1) age of onset and medication in BD patients, 2) symptoms severity in SCZ patients.ConclusionOur study reported both shared and specific neuroanatomical characteristics between the two disorders, suggesting more severe and generalized GMV deficits in SCZ, with a specific role for insula and thalamus.

Direct and indirect effects of fetal irradiation on cortical gray and white matter volume in the macaque

Increased grey matter volume and associated resting-state functional connectivity in chronic spontaneous urticaria: A structural and functional MRI study

Background: Although some Magnetic Resonance Imaging (MRI) studies have investigated the relationship between clinical severity and neuroanatomical alterations in patients with schizophrenia (SCZ), the biological signature associated with illness severity in schizophrenia is still uncertain. Therefore, this study aims to investigate structural brain abnormalities in SCZ, with particular regards to the identification of potential deficits associated with the severity of illness. Methods: In total, 1.5T MRI data were acquired for 61 subjects with SCZ and 59 matched healthy controls (HC). The patient group was divided in two sub-groups based on clinical severity, one composed of 34 mild-to-moderately ill patients, and the other of 27 severely ill patients, and compared with matched HC. Results: The whole group of patients with SCZ had significantly reduced grey matter (GM) volumes in the left inferior and middle temporal gyrus compared to HC (p < 0.05, pFWE corrected). Furthermore, compared to HC, patients with mild-to-moderate illness showed decreased GM volumes in the inferior and middle temporal gyrus, whereas those with severe illness had reduced GM volumes in the middle temporal gyrus and cerebellum bilaterally (all p < 0.001 uncorrected). No differences were observed between the two sub-groups of patients. Conclusion: The results showed significant GM volume reductions in temporal regions in patients with SCZ compared to matched HC, confirming the role of these regions in the pathophysiology of SCZ. Furthermore, specific cerebellar grey matter volume reductions were identified in patients with severe illness, which may contribute to stratifying patients with SCZ according to their clinical phenotype expression, ultimately helping in guiding targeted therapeutic/rehabilitation interventions.

Prefrontal–thalamic–cerebellar gray matter networks and executive functioning in schizophrenia

Psychoses in Alzheimer’s disease (AD) are associated with worse prognosis. Genetic vulnerability for schizophrenia (SCZ) may drive AD-related psychoses, yet its impact on brain constituents is still unknown. This study aimed to investigate the association between polygenic risk scores (PRSs) for SCZ and psychotic experiences (PE) and grey matter (GM) volume in patients with AD with (AD-PS) and without (AD-NP) psychosis. Clinical, genetic and T1-weighted MRI data for 800 participants were extracted from the ADNI database: 203 healthy controls, 121 AD-PS and 476 AD-NP. PRSs were calculated using a Bayesian approach and analysed at ten p-value thresholds. Standard voxel-based morphometry was used to process MRI data. Logistic regression models including both PRSs for SCZ and PE, and an AD-PRS were used to predict psychosis in AD. Associations between PRSs and GM volume were investigated in the whole sample and the three groups independently. Only the AD-PRS predicted psychosis in AD. Inconsistent associations between the SCZ-PRS and PE-PRS and GM volumes were found across groups. The SCZ-PRS was negatively associated with medio-temporal/subcortical volumes and positively with medial/orbitofrontal volumes in the AD-PS group. Only medio-temporal areas were more atrophic in the AD-PS group, while there was no significant correlation between psychosis severity and GM volume. Although not associated with psychoses, the SCZ-PRS was correlated with smaller medio-temporal and larger orbitofrontal volumes in AD-PS. Similar alterations have also been observed in SCZ patients. This finding suggest a possible disconnection between these regions associated with psychoses in more advanced AD.

This study evaluated childhood brain volumes using voxel-based morphometry and examined their relationships with grey matter, white matter and total brain volume in children diagnosed with intellectual disability (ID) and healthy controls. Sixty children (age range 6-12 years) diagnosed with mild to moderate ID according to DSM-5 and 60 age- and sex-matched healthy controls (age range 6-12 years) were included in the study. Total brain volume (TBV), white matter volume (WMV) and grey matter volume (GMV) were automatically calculated from magnetic resonance images of all participants. Normality was assessed using the Shapiro-Wilk test, group differences were assessed using Welch t-tests and associations were assessed using multiple regression and logistic regression analyses. The Shapiro-Wilk test results indicated that many variables were not normally distributed (TBV: p < 0.001; GMV: p < 0.001). In group comparisons, no significant effect of gender on TBV, WMV and GMV was found (t-values ranged from -0.757 to -0.276, all p > 0.45). Age was a strong predictor of TBV (β = 7.25, 95% CI [3.07, 11.43], p < 0.001), and the regression model had high explanatory power (R2 = 0.977). Logistic regression analyses showed that WMV (OR = 0.82, 95% CI [0.73, 0.92], p = 0.001) and GMV (OR = 1.16, 95% CI [1.06, 1.26], p = 0.0012) significantly predicted diagnostic status. However, the very high odds ratios calculated for the WMVM variable (OR > 1063, p < 0.001) indicate a multicollinearity problem. The findings indicate that white and grey matter volumes may differ in children with MR compared to healthy controls, and these structures have potential as diagnostic biomarkers. The study demonstrates the developmental heterogeneity of the brain during childhood. However, the limited sample size, the presence of outliers and multicollinearity should be considered as limitations.

Volumetric MRI correlates of persistent auditory verbal hallucinations and olfactory identification impairment in chronic schizophrenia: A cross-sectional study

Spatial working memory (SWM) deficit is a common problem in attention deficit disorder with hyperactivity (ADHD), often correlating with the severity of ADHD symptoms and academic difficulties. Although previous studies have broadly described abnormal brain structural changes in ADHD, the potential changes in brain morphology in children with ADHD with SWM dysfunction are still uncharacterized. This cross-sectional study was used to elucidate the brain morphological alterations associated with SWM performance in boys with ADHD. Data for this investigation were retrieved from one public dataset. A cohort of 23 boys diagnosed with ADHD and an age-matched group of 23 healthy male controls were selected for the study. Participants were administered n-back SWM tasks, with task accuracy and response times recorded. Voxel-based morphometry (VBM) was used to quantify gray matter (GM) volume, thereby characterizing the brain morphological features in both the ADHD and healthy control groups. Linear or rank correlation analyses were conducted to examine the relationship between GM volume and SWM task performance. VBM analysis revealed significantly lower GM volume in the right middle cingulate cortex (MCC), left precuneus, and right retrosubicular region among boys with ADHD. Moreover, a significant positive correlation was observed between the GM volume of the right MCC and the accuracy of the 2-back SWM task under conditions of small reward and immediate feedback. The reduced GM volume in the right MCC, left precuneus, and right retrosubicular might have a potential impact on SWM performance in children with ADHD.

Objective To investigate the changes of brain gray matter volume (GMV) and brain activity in patients with long-term methamphetamine (MA) abstinence compared with healthy controls. Methods This study included 44 abstinent MA patients and 40 demographically equivalent healthy controls. Structural magnetic resonance imaging (MRI) and resting-state functional MRI (rs-fMRI) were obtained on Siemens 3.0 T MR. Differences in GMV between abstinent MA group and control group were assessed using voxel-based morphometry (VBM) method. Within each region showed significant differences between two groups, the value of the fractional amplitude of low frequency fluctuation (fALFF) was calculated using the rs-fMRI data. Differences of fALFF between groups were also evaluated. Furthermore, partial correlation coefficients adjusted for age, years of education, smoking and drinking were calculated in the abstinent MA group to assess associations between the mean GMV and fALFF values in significant clusters and variables of MA use and abstinence. Results Compared with controls, abstinent MA group showed increased GMV in the right cerebellum crus and decreased GMV in the right calcarine. While fALFF values within bilateral cerebellum crus in abstinent MA patients were significantly increased. Moreover, GMV of the left cerebellar crus was positively correlated with the withdrawal time among abstinent MA patients. Conclusions Both structural and functional abnormalities were found in abstinent MA patients even after long-term abstinence. The positive correlation between GMV of left cerebellum crus and abstinent duration suggested that prolonged abstinence may be beneficial to brain recovery. The fALFF of the right calcarine was increased even with decreased GVM in the abstinent MA group. It indicates that the brain activity of this area is increased, which may imply that drug cue-induced craving is significant among methamphetamine patients even after long-term abstinence. Key words: Methamphetamine; Magnetic resonance imaging; Voxel-based morphometry; Fractional amplitude of low frequency fluctuation

Brain-derived Neurotrophic Factor (BDNF) and gray matter volume in bipolar disorder

Schizophrenia (SZ) is a complex psychiatric disorder that is currently defined by symptomatic and behavioral, rather than biological, criteria. Neuroimaging is an appealing avenue for SZ biomarker development, as several neuroimaging-based studies have shown measurable group differences in brain structure, as well as functional brain alterations in both static and dynamic functional network connectivity (sFNC and dFNC, respectively), between SZ and controls. The recently proposed filter-banked connectivity (FBC) method extends the standard dFNC sliding-window approach to estimate FNC within an arbitrary number of distinct frequency bands. Initial FBC results found that individuals with SZ spend more time in a less structured, more disconnected low-frequency (i.e., static) FNC state than HC, as well as preferential SZ occupancy in high-frequency connectivity states, suggesting a frequency-specific component underpinning the functional dysconnectivity observed in SZ. Building on these findings, we sought to link such frequency-specific patterns of FNC to covarying data-driven structural brain networks in the context of SZ. Specifically, we employ a multi-set canonical correlation analysis + joint independent components analysis (mCCA + jICA) data fusion framework to study the connection between gray matter volume (GMV) maps and FBC states across the full connectivity frequency spectrum. Our multimodal analysis identified two joint sources that captured co-varying patterns of frequency-specific functional connectivity and alterations in GMV with significant group differences in loading parameters between the SZ group and HC. The first joint source linked frequency-modulated connections between the subcortical and sensorimotor networks and GMV alterations in the frontal and temporal lobes, while the second joint source identified a relationship between low-frequency cerebellar-sensorimotor connectivity and structural changes in both the cerebellum and motor cortex. Together, these results show a strong connection between cortico-subcortical functional connectivity at both high and low frequencies and alterations in cortical GMV that may be relevant to the pathogenesis and pathophysiology of SZ.