A 2026 medRxiv preprint used 68 NeuroMark SPECT components to analyze 2,746 psychiatric patient scans and 76 controls, linking depression-related worry/rumination and social disinterest to brain-network patterns.1 This preprint is useful as a large biomarker-generating map, but the clinical-diagnostic claim needs much more caution than the sample size alone suggests. Research Highlights Large clinical SPECT …
A 2026 PLOS One study fine-tuned Qwen-2.5 and GLM-4 for psychology questionnaire generation and reported the largest gain in readability and comprehension difficulty: 3.41 to 4.66 on a 0–5 scale, a 36.7% relative improvement.1 The result supports item drafting inside a psychometric workflow, while clinical or research scale use still requires ordinary validation. Research Highlights …
A 2026 medRxiv preprint found that REDDI, a resting-state magnetoencephalography machine-learning pipeline, separated mild cognitive impairment, multiple sclerosis, Parkinson’s disease, and amyotrophic lateral sclerosis with 0.81 ± 0.04 balanced accuracy across 5 folds.1 That is a meaningful jump over the prior 67.1% MEG benchmark, but it is still research-stage decision support rather than a clinical …
A 2026 Parkinsonism study involving 63 male rats found that Antrodia cinnamomea plus citrate-stabilized silver nanoparticles reduced 6-OHDA motor, oxidative-stress, inflammatory, α-synuclein, and apoptosis markers, while dopamine and acetylcholine recovery remained nonsignificant. The useful claim is a multitarget preclinical signal, not a Parkinson’s disease treatment recommendation. Research Highlights Combination treatment looked strongest: Tekiner et al. …
A 2026 nationally-representative study by Johnson, Parker, and Wade in the Canadian Journal of Psychiatry tested ACE-style screening in 7,608 Canadian adults and replicated what US, UK, and New Zealand cohorts already found: predictive accuracy is poor, with AUCs of 0.62 to 0.67 across anxiety, mood, substance use, and suicidality — all below the conventional …
A 2026 preprint found that an independent supervisory safety agent detected suicide-risk signals in 205 of 224 synthetic evaluations, compared with 41 of 224 for native ChatGPT Health crisis safeguards.1 The result is a strong benchmark signal for external monitoring, not proof that AI systems can safely manage suicidal patients without human oversight. Research Highlights …
A 2026 preprint combining 2,189 university participants with 477 clinical patients with problematic sexual behavior found a gender-specific self-harm signal: women with problematic sexual behavior had higher non-suicidal self-injury risk, but clinical severity of problematic sexual behavior did not predict NSSI inside the patient cohort.1 Research Highlights Total sample was 2,666 people: Jiang et al. …