A 2026 ADNI study reported 85.23% accuracy for ANA-GNN, a graph neural network that combined structural MRI regional features with clinical variables to classify cognitively normal controls, mild cognitive impairment, and Alzheimer’s disease.1 The result is useful, but the clinical-feature ablation dropped accuracy to 68.35%, so the model should be read as multimodal decision-support research, …
A 2026 Molecular Psychiatry study found that RVI-AD, a structural-MRI score measuring how closely a person’s brain anatomy matches an Alzheimer’s disease pattern, predicted mild-cognitive-impairment conversion to dementia most strongly over the next 3 years: OR 2.16, 95% CI 1.81-2.57, p < 2e-16, AUC 0.70.1 Research Highlights Short-term conversion signal: in 965 ADNI participants with …
A 2026 UK Biobank MRI preprint involving 71,214 adults found that chronic pain plus current depression had a structural brain profile that was more than chronic pain plus depression pasted together. The 1,377-person comorbid group showed widespread lower cortical volume, lower bilateral thalamic and hippocampal volume, lower left accumbens volume, and broad white-matter microstructure differences …