Podcast MedAI Digest

A new study demonstrates that machine learning-based prognostic models can meaningfully improve accuracy and decision-making efficiency for surgeons managing colorectal liver metastases, particularly benefiting less experienced clinicians. Original paper: Impact of an AI prognostic tool on clinician performance in colorectal…

Researchers developed VesselNet, a deep learning model that predicts hemoglobin and red blood cell counts from brief magnified videos of eye blood vessels, offering noninvasive point-of-care blood testing. Original paper: Towards noninvasive blood count using a deep learning pipeline from…

Large language models are now competitive—and often superior—to traditional machine learning for clinical prediction tasks, according to a comprehensive benchmark study. Original paper: ClinicRealm: Re-evaluating large language models with conventional machine learning for non-generative clinical prediction tasks. — NPJ digital…

A new transfer-learning framework combines machine learning with physics-based design rules to discover high-performance conjugated polymers for low-voltage organic electrochemical transistors, achieving unprecedented prediction accuracy. Original paper: Transfer-learning guided design of high-performance conjugated polymers for low-voltage electrochemical transistors. — Nature…

A new study shows that radiology AI models fine-tuned on institutional data outperform general-purpose language models like GPT-4.1 in clinical acceptance, challenging assumptions about off-the-shelf AI solutions in healthcare. Original paper: Comparison of AI-generated radiology impressions: a multi-stakeholder evaluation. —…

A new study shows that radiology AI models fine-tuned on institutional data outperform general-purpose language models like GPT-4.1 in clinical acceptance, challenging assumptions about off-the-shelf AI solutions in healthcare. Original paper: Comparison of AI-generated radiology impressions: a multi-stakeholder evaluation. —…

Novel machine learning models predict 1-year stroke risk in newly diagnosed atrial fibrillation with dramatic improvements over the standard CHA₂DS₂-VASc score, offering clinicians better tools for personalized anticoagulation decisions. Original paper: Interpretable machine learning models for stroke risk prediction in…

ChatGPT-5 Pro can generate clinically relevant psychiatric vignettes for education, but significant safety gaps require expert oversight before classroom use. Original paper: Evaluation of artificial intelligence-generated vignettes depicting patient chatbot use in psychiatric contexts. — NPJ digital medicine. 10.1038/s41746-026-02605-6 📄…

Researchers developed a deep learning model that predicts bone mineral density from pediatric chest radiographs with high accuracy, potentially offering a more accessible screening tool for pediatric bone health assessment. Original paper: Deep Learning-based Bone Mineral Density Prediction Using Pediatric…

Researchers developed an interpretable machine learning framework that separates disease-specific abnormalities from normal physiologic uptake in Alzheimer’s PET imaging, generating a clinically relevant biomarker more strongly associated with cognitive outcomes than standard metrics. Original paper: Decoupling Alzheimer Disease Pathologic Abnormalities…