AI Tool Improves Early Detection of Silicosis and Black Lung on X-Rays

Black lung and silicosis are forms of lasting lung damage that can develop after years of inhaling coal or rock dust. More than 2.2 million U.S. workers inhale these particles on the job, and damage may develop over 10 to 20 years before severe breathing problems appear. Routine chest X-rays require review by specially trained physicians called B readers, but only about 200 are available in the United States. To address screening delays, researchers have developed an artificial intelligence tool to assist with these reviews.

The program was developed by researchers at Michigan State University (MSU; East Lansing, MI, USA) using a specialized dataset of U.S. worker scans. It is described as the first tool of its kind built specifically with images from U.S. workers. The study, titled “Pneumoconiosis screening and classification using deep learning models,” was published in Occupational and Environmental Medicine.

The AI system is intended to function as a high-speed assistant for four distinct screening tasks. It reviews chest X-rays and helps identify normal studies so that B readers can focus on images showing possible early disease. The software also overlays a color map on the X-ray to show where early lung damage was detected.

The program identified and cleared about half of normal X-rays from the review queue. It also achieved 91% accuracy in detecting the earliest dots of lung scarring. Human readers averaged 77% accuracy for the same task, according to the reported findings.

This workflow addresses several operational barriers in occupational lung screening. Dust-related lung damage can develop over 10 to 20 years, and the resulting scarring cannot be cured or reversed. Routine screening is required for workers such as coal miners every five years, while about 95% of routine workplace scans show healthy lungs.

The Michigan State University team is partnering with the National Institute for Occupational Safety and Health to package the technology into an application for wider use. The approach is designed to provide B readers with an objective second opinion and to help document findings earlier in the screening process.

“For workers in dusty environments, time is everything. If a worker's lung disease goes undetected because of screening delays, they may remain in a high-dust environment, and their lungs will continue to scar. By giving doctors an objective second opinion, this tool helps us catch disease at an early stage, allowing employers to remove workers from dangerous dust and reduce the likelihood that their lung disease will progress,” said Kenneth Rosenman, chief of the Division of Occupational and Environmental Medicine within the MSU College of Human Medicine and a certified B reader.

Related Links
MSU College of Human Medicine