Background Conventional MRI cannot be used to identify H3 K27M mutation status. This study aimed to investigate the feasibility of predicting H3 K27M mutation status by applying an automated machine learning (autoML) approach to the MR radiomics features of patients with midline gliomas. Methods This single-institution retrospective study included 100 patients with midline gliomas, including 40 patients with H3 K27M mutations and 60 wild-type patients. Radiomics features were extracted from fluid-attenuated inversion recovery images. Prior to autoML analysis, the dataset was randomly stratified into separate 75% training and 25% testing cohorts. The Tree-based Pipeline Optimization Tool (TPOT) was applied to optimize the machine learning pipeline and select important radiomics features. We compared the performance of 10 independent TPOT-generated models based on training and testing cohorts using the area under the curve (AUC) and average precision to obtain the final model. An independent cohort of 22 patients was used to validate the best model. Results Ten prediction models were generated by TPOT, and the accuracy obtained with the best pipeline ranged from 0.788 to 0.867 for the training cohort and from 0.60 to 0.84 for the testing cohort. After comparison, the AUC value and average precision of the final model were 0.903 and 0.911 in the testing cohort, respectively. In the validation set, the AUC was 0.85, and the average precision was 0.855 for the best model. Conclusions The autoML classifier using radiomics features of conventional MR images provides high discriminatory accuracy in predicting the H3 K27M mutation status of midline glioma.

Anorexia nervosa (AN) is a severe psychiatric disorder with high mortality. The underlying neurobiological mechanisms are not well understood, and high-resolution structural magnetic resonance brain imaging studies have given inconsistent results. Here we aimed to psychoradiologically define the most prominent and replicable abnormalities of gray matter volume (GMV) in AN patients, and to examine their relationship to demographics and clinical characteristics, by means of a new coordinate-based meta-analytic technique called seed-based d mapping (SDM). In a pooled analysis of all AN patients we identified decreased GMV in the bilateral median cingulate cortices and posterior cingulate cortices extending to the bilateral precuneus, and the supplementary motor area. In subgroup analysis we found an additional decreased GMV in the right fusiform in adult AN, and a decreased GMV in the left amygdala and left anterior cingulate cortex in AN patients without comorbidity (pure AN). Thus, the most consistent GMV alterations in AN patients are in the default mode network and the sensorimotor network. These psychoradiological findings of the brain abnormalities might underpin the neuropathophysiology in AN.

No abstract

Background: Anorexia nervosa (AN) is a debilitating illness whose neural basis remains unclear. Studies using tract-based spatial statistics (TBSS) with diffusion tensor imaging (DTI) have demonstrated differences in white matter (WM) microarchitecture in AN, but the findings are inconclusive and controversial. Objectives: To identify the most consistent WM abnormalities among previous TBSS studies of differences in WM microarchitecture in AN. Methods: By systematically searching online databases, a total of 11 datasets were identified, including 245 patients with AN and 246 healthy controls (HC). We used Seed-based d Mapping to analyze fractional anisotropy (FA) differences between AN patients and HC, and performed meta-regression analysis to explore the effects of clinical characteristics on WM abnormalities in AN. Results: The pooled results of all AN patients showed robustly lower FA in the corpus callosum (CC) and the cingulum compared to HC. These two regions preserved significance in the sensitivity analysis as well as in all subgroup analyses. Fiber tracking showed that the WM tracts primarily involved were the body of the CC and the cingulum bundle. Meta-regression analysis revealed that the body mass index and mean age were not linearly correlated with the lower FA. Conclusions: The most consistent WM microstructural differences in AN were in the interhemispheric connections and limbic association fibers. These common "targets" advance our understanding of the complex neural mechanisms underlying the puzzling symptoms of AN, and may help in developing early treatment approaches.

• Magnetic resonance spectroscopy probes the brain metabolism noninvasively. • Dorsolateral prefrontal reductions in NAA (a neuronal marker) and Ins are observed in TLE. • Temporal lobe epilepsy can result in metabolic changes remote from the seizure focus.