A recent patent application from Laleh Rad, associate professor of Biomedical Engineering and Radiology at Northwestern University, describes the use of machine learning for real-time risk assessment of magnetic resonance imaging in patients with conductive implants for whom tissue heating from radiofrequency excitation fields remains a major concern.

A 1,000-person study using Heartflow Inc. technology demonstrated that findings of coronary plaque on heart CT scans could indicate which patients were at risk of cardiovascular events up to seven years in the future – potentially providing a road map toward screening for heart disease, the leading cause of death for both men and women in the U.S. and worldwide.

Skin Analytics Ltd. received CE mark for its AI-based tool Deep Ensemble for the Recognition of Malignancy (DERM), which can assess images of lesions and detect skin cancer autonomously. The technology has an accuracy rate of 99.8% compared to 98.9% for dermatologists.

The U.K. Medicines and Health Care Products Regulatory Agency has provided guidance on what does and does not constitute a regulated digital mental health technology, but developers should be aware that promotional claims could push a product from the unregulated to the regulated category.

U.S. federal law is still lacking where AI is concerned, but California’s attorney general issued a bulletin that elevates the legal hazards of AI in medical use. The bulletin says that AI tools may not supplant a physician’s decision-making, a provision that would seem to render some FDA-cleared products illegal there.

C the Signs Ltd. recently received a capital injection of $8 million from Khosla Ventures to expand its AI-powered cancer prediction platform to the U.S. market.

The EU has moved aggressively on legislation in recent years, with the AI Liability Directive serving as the latest example of legislation that sparked widespread opposition.

Device makers know all too well the hazards of liability where their products are concerned, but those liabilities may soon balloon dramatically in the EU. Two pieces of EU legislation are in development, including the AI Liability Directive, which may have the effect of forcing the defendant in a liability case to prove that its product was not the cause of the alleged harm.

Researchers from Cornell University are seeking protection for their invention of a low-cost artificial intelligence (AI)-based platform that analyses low resolution electrocardiogram data and/or photoplethysmography data to accurately and objectively assess pain.

The discussion about taxpayer appropriations for CMS has been ongoing for decades, leading to futile speculation regarding user fees for the agency. Louis Jacques, who formerly worked at CMS, told BioWorld that Congress tends to be somewhat reactive when it comes to appropriations for CMS – a dynamic which suggests that appropriations for CMS are not likely to improve significantly in the near term.