In Japan, AI was taught to detect liver diseases from chest X-rays

Japanese scientists have created an AI that detects fatty liver disease from a regular chest X-ray. The breakthrough research has been published in the prestigious journal Radiology Cardiothoracic Imaging (RCI).

Traditional diagnosis of fatty liver disease—a dangerous condition that can progress to cirrhosis and cancer—requires expensive ultrasound, CT, or MRI. However, a team from Osaka University found a revolutionary solution. It turns out that even on a standard chest X-ray, which millions of patients undergo, a portion of the liver is visible that is sufficient for analysis using artificial intelligence.

The researchers conducted extensive neural network training on an impressive database of 6,599 X-ray images obtained from 4,414 patients. The results exceeded expectations—the system demonstrated the highest accuracy in detecting hepatic steatosis with an AUC score of 0.82–0.83, indicating the reliability of the method.

"Using regular X-rays to diagnose this disease can significantly simplify and reduce the cost of early detection of fatty liver disease," emphasized Professor Sawako Uchida-Kobayashi.

The scientific team expects the rapid implementation of their development into everyday medical practice, which could be a real breakthrough in the mass diagnosis of this common disease.