Using brain structural neuroimaging measures to predict psychosis onset for individuals at clinical high-risk

Abstract

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.

Fig. 1. Diagram employed for the processing and analysis.

HC healthy control, CHR clinical high risk for psychosis, CHR-PS+ individuals at CHR who developed psychosis later, CHR-PS- individuals at CHR who did not develop psychosis later, CHR-UNK individuals at CHR who could not follow up, SD standard deviation.

Fig. 2. Non-linear age associations of the surface area in healthy controls.

Each graph shows a partial effect of the best fit in GAMs. Shading around the line indicates the standard error. The bar underneath the age plots reflects the derivative of the slope.

Fig. 3. Surface area feature contributions and predictive performance comparisons of the XGBoost classifier.

A Weighted surface area features of XGBoost classification in Desikan-Killiany atlas. B Predictive performance of HC and CHR-PS+ groups was evaluated using the independent confirmatory dataset, and CHR-PS- and CHR-UNK groups using the independent group dataset. C Box and scatter plot of predict probabilities of XGBoost. P-values of post hoc comparisons were corrected using a Bonferroni method (***p < 0.001, **p < 0.01, *p < 0.05). D Best fit for the association of age with the predict probability in a GAM. Shading around the line indicates the standard error. E Decision curve analysis showed the benefits of XGBoost predicting the risk of psychosis conversion according to MRI scan.

Fig. 4. Age association of the surface area in the right superior temporal gyrus.

Each graph shows a GAM fit of age, group, and age by group interaction. Shading around the line indicates the standard error.