Scoliosis diagnosis support device, scoliosis diagnosis support system, machine learning device, scoliosis diagnosis support method, machine learning method and program

摘要

PROBLEM TO BE SOLVED: To provide a scoliosis diagnosis support device capable of accurately and easily estimating the degree of scoliosis with a minimally invasive and simple configuration. SOLUTION: A scoliosis diagnosis support device 102 has a first shape acquisition unit 103 for acquiring interest region information indicating a three-dimensional shape of an interest region on the back surface of a subject, and a subject in a three-dimensional shape indicated by the interest region information. The second shape acquisition unit 104 that acquires the mirror image information indicating the mirror image of the three-dimensional shape indicated by the region of interest information and the third shape that the region of interest indicates 3 The deviation distribution acquisition unit 105 that acquires the deviation distribution information showing the distribution of the deviation between the three-dimensional shape and the mirror image indicated by the mirror image information, and the estimated curvature angle that is the estimated value of the curvature angle by inputting the deviation distribution information. Based on the trained learning model that has been machine-learned to output, a scramble angle estimation unit 108 that outputs the estimated scramble angle of the subject by inputting the deviation distribution information is provided. [Selection diagram] Fig. 1