Nature Mental Health, Journal Year: 2023, Volume and Issue: 1(2), P. 86 - 87
Published: Feb. 17, 2023
Language: Английский
Nature Mental Health, Journal Year: 2023, Volume and Issue: 1(2), P. 86 - 87
Published: Feb. 17, 2023
Language: Английский
Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)
Published: Feb. 28, 2024
Abstract Predictive modeling is a central technique in neuroimaging to identify brain-behavior relationships and test their generalizability unseen data. However, data leakage undermines the validity of predictive models by breaching separation between training Leakage always an incorrect practice but still pervasive machine learning. Understanding its effects on can inform how affects existing literature. Here, we investigate five forms leakage–involving feature selection, covariate correction, dependence subjects–on functional structural connectome-based learning across four datasets three phenotypes. via selection repeated subjects drastically inflates prediction performance, whereas other have minor effects. Furthermore, small exacerbate leakage. Overall, our results illustrate variable underscore importance avoiding improve reproducibility modeling.
Language: Английский
Citations
32Cell Reports Medicine, Journal Year: 2023, Volume and Issue: 4(6), P. 101060 - 101060
Published: May 31, 2023
It has been 15 years since repetitive transcranial magnetic stimulation (rTMS) targeting the dorsolateral prefrontal cortex (DLPFC) was approved by FDA for clinical depression treatment. Yet, underlying mechanisms rTMS-induced relief are not fully elucidated. This study analyzes TMS-electroencephalogram (EEG) data from 64 healthy control (HC) subjects and 53 patients with major depressive disorder (MDD) before after rTMS Prior to treatment, MDD have lower activity in DLPFC, hippocampus (HPC), orbitofrontal (OFC), DLPFC-OFC connectivity compared HCs. Following active show a significant increase HPC, OFC. Notably, HPC is specifically associated amelioration of symptoms but anxiety or sleep quality. The orbitofrontal-hippocampal pathway plays crucial role mediating following These findings suggest potential alternative targets brain therapy against (chictr.org.cn: ChiCTR2100052007).
Language: Английский
Citations
41Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)
Published: March 14, 2024
Abstract This study explores the effectiveness of Explainable Artificial Intelligence (XAI) for predicting suicide risk from medical tabular data. Given common challenge limited datasets in health-related Machine Learning (ML) applications, we use data augmentation tandem with ML to enhance identification individuals at high suicide. We SHapley Additive exPlanations (SHAP) XAI and traditional correlation analysis rank feature importance, pinpointing primary factors influencing preventive measures. Experimental results show Random Forest (RF) model is excelling accuracy, F1 score, AUC (>97% across metrics). According SHAP, anger issues, depression, social isolation emerge as top predictors risk, while incomes, esteemed professions, higher education present lowest risk. Our findings underscore assessment, offering valuable insights psychiatrists facilitating informed clinical decisions.
Language: Английский
Citations
6Neuropsychopharmacology, Journal Year: 2024, Volume and Issue: 50(1), P. 114 - 123
Published: Aug. 15, 2024
Language: Английский
Citations
5Human Brain Mapping, Journal Year: 2025, Volume and Issue: 46(3)
Published: Feb. 15, 2025
Response inhibition (RI) deficits are a core feature across diagnostic categories of mental disorders. However, it remains unclear whether the brain networks underlying different forms RI disorder-shared or disorder-specific, and how they interact with aberrant connectivity Connectome-based predictive modeling (CPM) provides novel approach for exploring associated abnormalities Publicly available resting-state functional magnetic resonance imaging data from individuals schizophrenia (n = 47), bipolar disorder attention-deficit/hyperactivity 40), as well healthy controls 121), were utilized to construct whole-brain network models (action cancellation action restraint). The further compared abnormal in groups. Action restraint exhibited both shared distinct networks. There was dissociation relationship between patterns observed categories. Our successfully predicted performance categories, whereas model failed effectively predict due influence disease-related on cancellation. Nevertheless, demonstrated generalizability novel, participants 220) an independent dataset. study clarifies complex neuropathology disorders foundation more accurate cognitive assessment targeted interventions. findings highlight importance refining constructs emphasize value applying connectome methods reveal cross-diagnostic neural mechanisms.
Language: Английский
Citations
0Progress in Neuro-Psychopharmacology and Biological Psychiatry, Journal Year: 2025, Volume and Issue: unknown, P. 111351 - 111351
Published: April 1, 2025
Language: Английский
Citations
0Neurotherapeutics, Journal Year: 2024, Volume and Issue: unknown, P. e00496 - e00496
Published: Nov. 1, 2024
Language: Английский
Citations
2Nature Mental Health, Journal Year: 2023, Volume and Issue: 1(11), P. 887 - 899
Published: Oct. 26, 2023
Language: Английский
Citations
6Journal of Affective Disorders, Journal Year: 2024, Volume and Issue: 367, P. 768 - 776
Published: Sept. 2, 2024
Language: Английский
Citations
1Brain and Cognition, Journal Year: 2024, Volume and Issue: 181, P. 106221 - 106221
Published: Sept. 8, 2024
Language: Английский
Citations
1