AI-Powered Atlas Reveals Hidden Links Between Body Fat and Disease Risk

A new AI-driven study has created the first detailed atlas of human body composition. Researchers used advanced imaging and machine learning to analyse whole-body MRI scans from more than 66,000 people. The findings reveal strong links between specific fat and muscle measurements and serious health risks. The study developed an open-source, fully automated AI framework to process MRI data. This system extracts precise body composition metrics with little need for human input. It can also work with routine chest or abdominal CTs and MRIs, making it widely adaptable.

Researchers identified that high levels of intramuscular fat raise the risk of major cardiovascular events by 1.54 times. Visceral fat, meanwhile, was linked to a 2.26-fold greater chance of developing diabetes. The study further showed that low skeletal muscle mass independently increases all-cause mortality risk by 1.44 times.

The team also generated reference curves that track body composition changes as people age. These findings challenge the reliance on BMI as the main health risk indicator, suggesting more precise measurements are needed. The AI-powered atlas provides a new way to assess health risks by analysing fat and muscle distribution. Its open-source framework allows other researchers to apply the same methods to existing imaging data. The results could lead to more accurate health predictions and personalised medical advice.