2MM: AI Roundup –Dermatology accuracy, faster drug discovery, heart disease detection, and global medical training [September 11th, 2025]

Google’s AI matches dermatologists in melanoma diagnosis
England’s health service is piloting an artificial intelligence tool called Deep Ensemble for Recognition of Malignancy, abbreviated as DERM, across multiple hospitals as part of a three-year trial program, reported on August 27 in Nature. The system analyzes skin lesion images and triages which patients should be prioritized for biopsy or dermoscopy. In early testing, sensitivity for melanoma exceeded 90 percent, a rate higher than most individual dermatologists achieve. Equally important, the model is being evaluated across a wide range of skin tones and lesion types, areas where earlier tools struggled. Referrals for suspected skin cancer in the National Health Service have more than doubled over the past decade, and clinics often face months-long waiting lists. By providing consistent triage, artificial intelligence could help shorten these delays and make sure high-risk patients are seen first. Clinicians stress that the technology is a support rather than a replacement, with final decisions still resting on expert review. This represents one of the clearest examples of artificial intelligence moving from research into structured health-system testing.

Artificial intelligence gains momentum in drug discovery
On September 2, Reuters highlighted that the U.S. Food and Drug Administration is encouraging use of artificial intelligence to model drug safety and efficacy before human studies. Traditional preclinical programs can take more than three years, yet some artificial intelligence platforms have delivered viable candidates in as little as 18 months. Case studies suggest timelines and costs can be cut by 50 percent, a significant shift when developing a single new medicine often exceeds two billion dollars. One contract research organization reported about 200 million dollars in annual revenue from these non-animal methods, underscoring that adoption is already underway. Regulators continue to require human data but are signaling openness to computational submissions if validated carefully. If successful, this hybrid approach may allow cancer, rare-disease, and cardiometabolic programs to progress more quickly without sacrificing safety. For clinicians, this trend could mean a broader set of therapies entering trials sooner. Faster pipelines often translate into more therapeutic options for patients in the years ahead.

AI stethoscope detects heart disease in seconds
Results released on August 30 at the European Society of Cardiology congress showed that an artificial intelligence enabled stethoscope could identify major cardiac conditions in about 15 seconds, as reported by The Guardian. The trial, run by Imperial College London in partnership with the National Health Service, included nearly twelve thousand patients. Compared with routine exams, the tool doubled detection of heart failure, tripled identification of atrial fibrillation, and nearly doubled recognition of significant valve disease. A British Heart Foundation summary confirmed these results and emphasized the device’s ease of use. The stethoscope records both heart sounds and an electrocardiogram, then transmits the data to the cloud for algorithmic analysis. Clinicians noted that it added little time to the exam while providing immediate triage guidance. Heart failure alone affects more than 64 million people worldwide, making earlier diagnosis a major global health need. While false positives remain a consideration, the ability to flag disease rapidly at the point of care could reshape frontline cardiology.

India launches first Department of Artificial Intelligence in Healthcare
Kasturba Medical College in Manipal inaugurated a Department of Artificial Intelligence in Healthcare on August 29, confirmed by the university’s official release. The program brings medical students, engineers, and data scientists together with the goal of training hundreds of learners in the coming years. According to The Times of India, projects will range from imaging algorithms that prioritize urgent scans to outbreak prediction models for regional hospitals. Leaders stressed that artificial intelligence should be as familiar to new physicians as electronic health records. India faces the dual challenge of scaling its healthcare workforce while caring for a large and diverse population, and this initiative is meant to address both. Global observers note that this is not a short-term pilot but a permanent academic department with a defined curriculum. For physicians abroad, the signal is that formal training in artificial intelligence is becoming part of mainstream medical education. Graduates who are comfortable using machine learning in daily care will likely accelerate adoption and raise expectations for safety and transparency worldwide.

Image: PD

©2025 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.