Chest X-rays using artificial intelligence can predict the risk of osteoporosis
Sep 19, 2025
A chest X-ray using deep learning technology-based artificial intelligence (AI) can predict the risk of osteoporosis, a study has been published.
A research team led by Professor Kim Kwan-chang of Cardiovascular Thoracic Surgery at Ewha Womans University Medical Research Institute and Professor Ahn So-hyun of Ewha Womans University analyzed the risk of osteoporosis through a deep learning model (PROS® CXR: OSTEO, osteoporosis diagnosis assistance software) for 80 adults who underwent chest X-ray and dual energy X-ray absorption measurement (DXA) bone density tests at the Ewha Womans University Medical Center from 2021 to 2024.
AUC is an indicator that the closer to 1, the better the performance, and if it is 0.8 or higher, it is classified as a high-performance model. As a result of learning the bone density diagnostic test data to AI, the research team showed excellent performance, recording the AI's diagnostic accuracy (AUC) level of 0.93.
The 'Dual-energy X-ray Absorption Metrology (DXA)' used in the existing osteoporosis diagnosis had limitations in using it as a standard test for osteoporosis due to the high cost of testing and concerns of exposure to radiation.
The present study demonstrated that chest X-rays are a stable treatment option for osteoporosis patients in the future in that they can screen for osteoporosis patients early with efficient cost and high accessibility and provide opportunities for proper treatment and prevention.
Professor Kim Kwan-chang, who participated as the corresponding author of this study, said, "This study is of great clinical significance in that it can screen the risk of osteoporosis with chest X-rays. Beyond osteoporosis diagnosis research, we are also currently working on developing an AI diagnostic model for lung diseases. We plan to continue our research with the aim of establishing an integrated AI platform that can screen multiple diseases with a single imaging test by linking the two studies."
On the other hand, the study was published in the latest issue of the international journal Journal of Thoracic Disease with a paper titled 「Performance Evaluation of Deep Learning-Based Osteoporosis Diagnostic Models Using Traditional Chest X-rays」 in the clinical cohort.
A research team led by Professor Kim Kwan-chang of Cardiovascular Thoracic Surgery at Ewha Womans University Medical Research Institute and Professor Ahn So-hyun of Ewha Womans University analyzed the risk of osteoporosis through a deep learning model (PROS® CXR: OSTEO, osteoporosis diagnosis assistance software) for 80 adults who underwent chest X-ray and dual energy X-ray absorption measurement (DXA) bone density tests at the Ewha Womans University Medical Center from 2021 to 2024.
AUC is an indicator that the closer to 1, the better the performance, and if it is 0.8 or higher, it is classified as a high-performance model. As a result of learning the bone density diagnostic test data to AI, the research team showed excellent performance, recording the AI's diagnostic accuracy (AUC) level of 0.93.
The 'Dual-energy X-ray Absorption Metrology (DXA)' used in the existing osteoporosis diagnosis had limitations in using it as a standard test for osteoporosis due to the high cost of testing and concerns of exposure to radiation.
The present study demonstrated that chest X-rays are a stable treatment option for osteoporosis patients in the future in that they can screen for osteoporosis patients early with efficient cost and high accessibility and provide opportunities for proper treatment and prevention.
Professor Kim Kwan-chang, who participated as the corresponding author of this study, said, "This study is of great clinical significance in that it can screen the risk of osteoporosis with chest X-rays. Beyond osteoporosis diagnosis research, we are also currently working on developing an AI diagnostic model for lung diseases. We plan to continue our research with the aim of establishing an integrated AI platform that can screen multiple diseases with a single imaging test by linking the two studies."
On the other hand, the study was published in the latest issue of the international journal Journal of Thoracic Disease with a paper titled 「Performance Evaluation of Deep Learning-Based Osteoporosis Diagnostic Models Using Traditional Chest X-rays」 in the clinical cohort.
|
This article was translated by Naver AI translator.
