Section
A deep learning-derived sarcopenia marker, the muscle-bone ratio (MBR), has been shown to predict the efficacy of anti-EGFR maintenance therapy in patients with RAS wild-type metastatic colorectal cancer, potentially identifying those who t
The PECTUS-AI study demonstrates that AI-based identification of thin-cap fibroatheromas using OCT provides superior prognostic value for cardiovascular events compared to manual analysis, particularly when assessing the entire imaged coron
The novel PILOT machine learning architecture integrates time-phased perioperative data and regeneration-associated biomarkers to predict post-hepatectomy liver failure within six hours of surgery, significantly outperforming traditional cl
A multicenter study demonstrates that deep transfer learning (DTL) algorithms using preimplant MRI can predict spoken language development in children with cochlear implants with over 92% accuracy, significantly outperforming traditional ma
A secondary analysis of the TEAM trial using machine learning demonstrates that enhanced early mobilization in mechanically ventilated patients has highly individualized effects, ranging from a 34% mortality reduction to a 39% increase in r
AI-driven analysis of routine chest CTs reveals that coronary artery calcium is highly prevalent and strongly predictive of MACE and mortality in patients with IMIDs, identifying a critical treatment gap in this high-risk population.
An mHealth text messaging intervention significantly improves smoking cessation rates and reduces mortality among TB patients compared to usual care, supporting its implementation in TB programs.
A comprehensive meta-analysis indicates that rule-based chatbots provide a modest but statistically significant reduction in depressive symptoms within a 4-8 week window, whereas evidence for the clinical efficacy of Large Language Model (L
This article explores a landmark mixed-methods study establishing a digital biomarker ontology for the early detection of Cytokine Release Syndrome (CRS). By identifying core physiological markers, researchers aim to transform immunotherapy
Recent robust randomized trials demonstrate that smartphone-delivered digital cognitive behavioral therapy provides significant, sustained improvements in generalized anxiety disorder, overcoming key access barriers and highlighting its pot
A large-scale prospective study across England’s NHS reveals that AI imaging software significantly boosts endovascular thrombectomy rates for acute stroke patients. Implementation was associated with a 100% relative increase in treat
A 102-participant pilot RCT in Germany found that a self-guided digital psychological intervention produced clinically meaningful reductions in psychological distress and modest improvements in quality of life at 3 months in people with inf
A systematic review and meta-analysis finds that real‑time automated clinical deterioration alert systems reduce in‑hospital cardiac arrests and may shorten ICU stay, but mortality benefits are uncertain and higher‑quality trials are needed
This review evaluates AI algorithms predicting heart failure risk from noisy single-lead ECGs, synthesizing evidence from large multinational cohorts, highlighting improved risk stratification with AI-ECG models beyond traditional clinical
An international study developed AI-ECG models that diagnose and predict future moderate–severe mitral, tricuspid, and aortic regurgitation, validated across distinct populations and linked to subclinical chamber remodelling.
A multicentre deep‑learning model reliably classified biopsy focus score and ACR‑EULAR–defined Sjögren’s disease (AUROC ≈0.88–0.89) and identified a novel CD8+ T‑cell peri‑acinar pattern associated with disease; prospective validation is ne
Contemporary foundation models match experts on text-only ophthalmology exam questions but underperform on image-based items; targeted multimodal training and prospective validation are needed before clinical deployment.
A Danish RCT of TG‑iConquerFear, a 10‑week therapist‑guided online program for colorectal cancer survivors with clinical fear of cancer recurrence (FCR), reduced FCR scores substantially compared with augmented self‑help at 3 months; the be
This critical synthesis of a 2020–2025 narrative review (327 DL studies) evaluates deep learning (DL) applications across diagnosis, prognosis, segmentation, and emerging intraoperative uses in otolaryngology and outlines practical steps fo
An AI-driven screening program (ATTRACTnet) applied to ECG, echocardiography, demographics and orthopedic history substantially increased case detection of transthyretin cardiac amyloidosis (ATTR‑CM) and led to rapid treatment initiation in
