Mental Health via Writing

Gamified human-AI collaborative system integrating painting therapy reduces adolescent depression and increases motivation in randomized controlled trial.

Study reveals how financial insecurity and legal precarity increase susceptibility to online scams, and documents help needs across different stages of victimization.

Linguistic analysis of 2 million forum comments reveals distinct psychological trajectories in users who leave, stay in, or return to an incel community, with re-engagement marked by reduced.

Deep learning framework classifies undergraduate GPA by integrating family background and psychological factors, achieving 0.798 accuracy in identifying academic risk among Chinese college students.

Intersectional analysis of depression disparities in All of Us data reveals that risk varies markedly by overlapping demographic identities and area-level socioeconomic factors.

Framework formalizes how therapists rapidly assess patient capacity and calibrate intervention intensity using four observable signals and safety-gated decision rules in psychotherapy.

Cross-sectional study of 112 adults finds online social support functions as adjunct to traditional support rather than substitute, supporting rich-get-richer hypothesis.

Study examines how Japanese Twitter users' attitudes toward price increases shifted after 2021, revealing erosion of decades-old deflationary norms and growing acceptance of inflation.

Comparative analysis of writing assessment and linguistic patterns in narratives from public and private Mexican universities using automated analysis and SISAT scoring.

LLM-based coaching system for training emotionally supportive communication through AI-generated scenarios, exemplary responses, and assessment feedback.

Longitudinal predictive modeling identifies psychosocial and demographic risk factors for emergent major depressive disorder using machine learning explainability methods.

Structured model for hospital suicide-risk detection integrating EHR screening, clinical follow-up, and staff training to address pandemic-era mental health vulnerabilities.

Study validates cognitive distortion architecture in Russian discourse using AI analysis of 249,414 social media texts, identifying 443,447 distortion instances across 18 categories with stable.

Graph neural networks with augmented brain signals improve MDD diagnosis through gender-specific and stage-wise analysis, enabling personalized therapeutic strategies.

Study evaluates reasoning-based large language models for predicting 12-week remission in depressive disorder patients undergoing antidepressant monotherapy.