AI Can Now Tell Your Age By Using A Chest X-Ray

Have you ever pondered the phenomenon of certain individuals appearing older than their chronological age suggests? A recent study conducted by researchers at Osaka Metropolitan University (OMU) in Japan presents a compelling notion: such premature aging might be indicative of latent health conditions yet to be comprehended.

The OMU researchers have devised an ingenious computational program that engages with chest X-ray images to deduce an individual’s age. Distinct from other computational systems focused on identifying pulmonary abnormalities in X-rays, this program endeavors to ascertain the perceived age of the subject. Subsequently, this estimated age is employed to unveil potential underlying health issues.

To elucidate, if the computational program indicates an older age than the individual’s actual years, it could signify the presence of ailments such as Chronic Obstructive Pulmonary Disease (COPD), elevated blood pressure, or hyperuricemia – characterized by excessive uric acid levels in the bloodstream.

This novel computational program stands as an innovative paradigm, correlating the appearance of age in X-ray imagery with plausible health concerns. The researchers postulate its potential to facilitate early detection of health disorders, a pivotal facet in rendering therapeutic interventions more efficacious.

https://media.springernature.com/full/springer-static/image/art%3A10.1038%2Fs43856-022-00220-6/MediaObjects/43856_2022_220_Fig1_HTML.png

The training process of the computational program entailed feeding it extensive datasets comprising X-ray images and corresponding ages. This compilation included over 67,000 X-rays from a cohort exceeding 36,000 asymptomatic individuals, spanning the temporal range of 2008 to 2021.

Evaluation of the program’s accuracy entailed juxtaposing its age estimates against actual ages, culminating in remarkably congruent results. The program demonstrated notable adeptness in age estimation.

Furthermore, the program underwent validation using an additional dataset of 34,197 X-rays, this time encompassing individuals afflicted with diagnosed medical conditions. Intriguingly, a noteworthy positive correlation surfaced between the estimated age provided by the program and the likelihood of harboring chronic medical ailments.

Specifically, instances where the program inferred an older age than the chronological age often corresponded with the presence of ongoing health afflictions, such as elevated blood pressure or chronic bronchitis.

The researchers harbor sanguine prospects for this computational program’s future development. Further refinement and research endeavors are anticipated, with the ultimate objective of expanding its capabilities beyond age estimation – encompassing prognostication of lifespan, determination of disease survival probabilities, and formulation of tailored treatment protocols for diverse medical conditions.

Leave a Reply

Your email address will not be published. Required fields are marked *