Chest Radiographs Can Tell How Old You Really Are

Background: Biologic age is based on the idea that one may undergo functional or physiologic decline that age an individual greater than their chronologic age. A potential source of estimating this has been developed using chest radiographs and a deep learning model (CXR-Age). This was shown in a prior study to better predict all-cause and cardiovascular mortality than does simple chronologic age. Objective: To test the CXR-Age model using a large Asian population (original study done on a largely white population in the United States) and assess its ability to predict additional outcomes, such as lung cancer and respiratory disease mortality, in comparison to chronologic age. Design: Retrospective study. Methods: The records of 36,924 asymptomatic patients enrolled in a medical checkup program at a single center in Seoul, South Korea, were included in the study. Patients were aged 50 to 80 years between 2004 and 2018. Patients who underwent anything other than a posteroanterior ches more...