Device and trained neural network for determining fracture probability using a soft part image

Abstract

A processor is configured to acquire a first and a second radiation image which are acquired by imaging a subject including a bone part and a soft part with radiation having different energy distributions, perform weighting subtraction on the first radiation image and the second radiation image to derive a bone part image representing a bone tissue of the subject and a soft part image representing a soft tissue of the subject, derive a bone mineral density for each pixel in a target bone region of the subject from the bone part image, acquire a pixel value for each pixel of a corresponding region, which corresponds to the target bone region, in the soft part image, and derive a fracture probability of a target bone from the bone mineral density for each pixel in the target bone region and the pixel value for each pixel in the corresponding region.