Single Chest X-Ray Predicts Mortality Risk

A deep learning CXR-risk score (very low, low, moderate, high, and very high) is generated based on CNN analysis of a submitted radiograph.

To develop the CNN, the researchers used 41,856 x-rays from the screening radiography arm of the Prostate, Lung, Colorectal, and Ovarian (PLCO) cancer screening trial, a community cohort of asymptomatic nonsmokers and smokers enrolled at 10 U.S. sites from November 8, 1993, through July 2, 2001. The results of the CNN were tested in a further 10,464 cases from the screening radiography arm of the National Lung Screening Trial (NLST), a community cohort of heavy smokers enrolled at 21 U.S. sites from August 2002 through April 2004.

The results revealed a graded association between CXR-risk score and mortality. The very high-risk group had an all-cause mortality of 53% (PLCO) and 33.9% (NLST), compared with the very low-risk group. The association was robust to adjustment for radiologists’ findings and risk factors. Comparable results were seen for lung cancer death, non-cancer cardiovascular death, and respiratory death. The study was published on July 19, 2019, in JAMA Network Open.

“We get chest x-rays to make a diagnosis like pneumonia, but our study shows that there is also free prognostic information about health and longevity on the images. Based on the chest x-ray image alone, AI identified people at up to a 53% risk of death over 12 years,” said lead author Michael Lu, MD, MPH, of MGH and HMS. “Scores calculated using AI may incentivize high-risk individuals to lower their chance of dying with prevention, regular screening, and lifestyle modification.”

Deep learning is part of a broader family of AI machine learning methods based on learning data representations, as opposed to task specific algorithms. It involves CNN algorithms that use a cascade of many layers of nonlinear processing units for feature extraction, conversion and transformation, with each successive layer using the output from the previous layer as input to form a hierarchical representation.

Related Links:
Massachusetts General Hospital
Harvard Medical School