Osaka Public University in Japan recently announced that its researchers have developed an advanced artificial intelligence model that can accurately estimate the age of a person taking a chest X-ray. When the estimated age differs from the actual age, it may indicate a link to a chronic disease, and this research could help improve the early detection and intervention of diseases.
The research team first built a deep learning-based artificial intelligence model to estimate the age of a person taking a chest X-ray from a healthy individual. They then used this model to analyze chest X-rays of patients with known diseases to investigate the relationship between AI-estimated age and various diseases.
Between 2008 and 2021, researchers obtained 67,099 chest X-rays from 36,051 healthy individuals who underwent health checkups at three institutions. The analysis showed that the correlation coefficient between the age estimated by the artificial intelligence model and the actual age was 0.95.
To verify the effectiveness of artificial intelligence in estimating age using chest X-rays, researchers collected 34,197 chest X-rays from patients with known diseases at two other institutions. Analysis showed a positive correlation between the AI-estimated age and the patient's actual age and several chronic diseases, including hypertension, hyperuricemia, and chronic obstructive pulmonary disease (COPD). The higher the AI-estimated age compared to the actual age, the greater the individual's likelihood of having these diseases.
The research paper has been published in the recent issue of The Lancet Health on Aging.
Researchers say that the apparent age estimated by artificial intelligence based on chest X-rays can accurately reflect health status based on actual age. The next goal of the research is to apply it to assessing the severity of chronic diseases, predicting life expectancy, and predicting potential surgical complications.
