Routine primary care conversations may contain measurable acoustic signals of cognitive impairment. In this diagnostic study, machine learning models using speech features from visit recordings showed moderate screening…

A large cohort study suggests that blood proteins can capture aging signals at cellular resolution, linking specific cell-type aging patterns to human disease, mortality, and resilience.

A multicentre Nature Communications study presents MM-DLS, a multimodal multitask deep learning system for non-invasive assessment of non-small cell lung cancer from pretreatment PET/CT and clinical data.

Transcript INTRO Welcome to Medical AI Digest Podcast. Today, a paper from Nature communications. MAIN Today we’re looking at a study called “Speech as a biomarker for supported…

This week’s medical AI papers pushed beyond leaderboard gains: ICU forecasting, pathology robustness, multimodal IBD atlases, ophthalmic ultrasound triage, and renal CT foundation models all asked whether AI…

Recent Digital Health & Informatics coverage shows AI shifting from benchmark promise toward clinical proximity: faster signals, workflow assistance, short-horizon risk, and reusable imaging infrastructure.

FairGen proposes a fairness-aware route for medical image synthesis: generate demographically balanced images while preserving disease-relevant visual features.

Recent Emergency Medicine AI coverage shows the field moving upstream: dispatch, triage, acute diagnosis, resuscitation guidance, and post-ED risk prediction are all becoming targets for prospective workflow impact.

Ionizable drugs may do more than act as payloads: some can help package siRNA into nanoparticles. Slaughter, Dang, Donders et al. built predictive tools to identify such drugs…

In Oncology, AI is automating routine clinical tasks—delineation, staining, reporting—while expanding diagnostic reach through breath biomarkers and smartphone imaging. Foundation models reshape both specialized cancer analysis and general…