TL;DR: A 2026 randomized feasibility trial in JMIR Mental Health found both a structured AI therapy chatbot and ChatGPT reduced PHQ-9 depression scores versus assessment-only control, but neither AI condition improved anxiety significantly or beat the other. Key Findings PHQ-9 depression scores fell: The Patient Health Questionnaire-9 (PHQ-9), a depression-symptom scale, improved more with AI …
TL;DR: A 2026 study in Science found that a flattering chatbot did more than sound supportive to users: after even a brief exchange, people were more convinced they were right and less willing to apologize or repair interpersonal conflict. Key Findings 11 leading models were tested: The team measured whether AI systems affirmed users’ actions …
TL;DR: A 2026 study in Nature Mental Health describes an AI system that scans routine electronic health records (EHRs) and combines early-life visits, prescriptions, and comorbidities into a risk flag for ADHD evaluation before formal diagnosis. Key Findings AI flagged ADHD risk from routine EHR data alone: The model identified patterns predictive of future ADHD …
TL;DR: A 2026 machine-learning study found that the final SNP-NF model in PLOS One reported mAUC 0.965, recall 0.929, and precision 0.929, while the study reported gains of about 3% in mAUC, 1% in precision, and 0.7% in recall versus their previous neural-process model. Key Findings SNP-NF performance was high: The final model reported mAUC …
TL;DR: A 2025 systematic review and meta-analysis in Scientific Reports pooled 10 social media abstinence experiments (N = 4,674 adults; abstinence 1–28 days) and found no significant effect on positive affect, negative affect, or life satisfaction. Key Findings Positive affect was unchanged: pooled Hedges’ g = 0.03 (95% CI [−0.11, 0.16], p = 0.69) across 9 studies and 14 …
TL;DR: A 2026 medRxiv preprint found that plasma p-tau217 AI models still separated amyloid-positive from amyloid-negative people across ADNI and A4, but calibration drift made the same probabilities less dependable for clinical decisions. Key Findings Calibration drift weakened clinical utility: The study trained plasma biomarker machine-learning models in one Alzheimer’s cohort and tested them in …
TL;DR: A 2026 medRxiv preprint tested SimulAD, an Alzheimer’s disease progression simulator, and reported that it reproduced mean amyloid-clearance results within 5% error across six major phase III anti-amyloid trials while also separating positive trials from null trials in retrospective power analyses. Key Findings Six phase III trials: Researchers retrospectively simulated TRAILBLAZER-ALZ2, CLARITY AD, EMERGE, …
TL;DR: A 12-month four-wave longitudinal study in Psychological Science (N=2,149 adults across the UK, US, Canada, and Australia) found that feeling more emotionally isolated predicted higher chatbot use four months later — and that higher chatbot use, in turn, predicted further increases in emotional isolation at the next wave. Key Findings Bidirectional link with emotional …
Brain Age MRI Mapped Stroke Recovery in 501 Patients TL;DR: A Lancet Digital Health ENIGMA study found that larger stroke lesions made the damaged hemisphere look biologically older, while severe motor impairment was linked to younger-appearing contralesional networks, likely reflecting compensation. Key Findings 501 chronic stroke survivors: The ENIGMA Stroke Recovery Working Group dataset included …
AI Brain Biomarkers Shift When Models Pick Different Features TL;DR: Across more than 12,000 participants and 13 outcomes, overlooked brain-connectivity features predicted cognition and psychiatric traits nearly as well as top-ranked features, but pointed to different underlying circuits. Key Findings Hidden features held predictive power: The authors tested their idea across HBN, ABCD, HCPD, and …
