(IMAGE)
Caption
Overview of the proposed method. The training dataset consists of real X-ray images, quantitative CT (QCT) images, and the BMD measurements of 315 patients. The bone region in QCT was extracted and aligned with the X-ray image through the 2D-3D registration method. We trained an AI model that generates a projection image of the bone-segmented QCT from an X-ray image. The bone mineral density (BMD) was calculated as the average intensity of the generated projection image. In the clinical application, the system requires only an X-ray image to predict the BMD.
Credit
Yoshito Otake
Usage Restrictions
Credit required
License
Original content