A breakthrough in the realm of medical imaging has emerged through a novel artificial intelligence (AI) model, as outlined in a recent publication in The Lancet Healthy Longevity journal. The study, originating from Osaka Metropolitan University in Japan, introduces an AI model capable of deducing an individual’s age and identifying chronic ailments by analysing their chest X-ray images.
This revolutionary AI system not only estimates a person’s age but also has the capability to signal the presence of chronic conditions, such as hypertension and chronic obstructive pulmonary disease, by discerning the variance between the predicted age and the chronological age of the patient. The implications of this research are substantial, indicating a significant stride forward in medical imaging technology that could pave the way for more effective early detection of diseases and targeted intervention strategies.
With the world’s population increasingly ageing, the focus on ageing and longevity research has intensified. The intricate nature of ageing, entwined with numerous health disorders, leads to diverse effects across individuals. The lead researcher, Yasuhito Mitsuyama, emphasised, “Chronological age plays a pivotal role in the realm of medicine. Our findings propose that the apparent age calculated from chest radiography holds the potential to provide insights into health conditions that surpass chronological age.”
The foundation of the AI model’s age estimation was built upon a dataset encompassing approximately 67,100 chest X-rays from 36,051 healthy individuals who had undergone health assessments between 2008 and 2021. This initial training phase demonstrated a robust correlation between the AI-derived age and the participants’ actual chronological age.
Subsequently, the AI model was further developed to explore the connection between its predicted age and various diseases. An additional 34,197 chest X-rays, sourced from individuals with documented diseases, were employed to establish these associations. Cumulatively, the model was honed using an extensive dataset of around 101,300 chest X-rays obtained from 70,248 participants spanning five Japanese institutions.
The research illuminated a significant link between the disparity in AI-estimated age and the presence of chronic ailments like hypertension, hyperuricemia (elevated uric acid levels in blood), and chronic obstructive pulmonary disease. Consequently, a higher AI-predicted age corresponded to an increased probability of these individuals harbouring the mentioned health conditions.
Beyond its age estimation prowess, chest X-rays exhibited potential as markers of ageing and longevity due to their ability to visualise internal organs, skeletal structures, and bodily features. Mitsuyama conveyed the team’s aspirations, noting, “Our objective is to not only advance this research but also apply it to gauge the severity of chronic illnesses, predict life expectancy, and foresee potential surgical complications.”
In essence, this groundbreaking AI model signifies a significant advancement in medical diagnostics, offering a novel perspective on age estimation and disease identification through chest X-ray analysis. Its potential impact on the realm of healthcare holds promise for refining disease management and enhancing patient outcomes.
