Diagnosis support method, diagnosis support apparatus, and non-transitory computer-readable storage medium

A diagnosis support apparatus comprising:

• a memory;
• and
• a processor coupled to the memory and configured to
• execute a process that includes
• obtaining a first case image group from among plural case images about each of plural patients, the first case image group including one or more case images, each of the one or more case images included in the first case image group having a same abnormality as a first abnormality detected from an image of a subject, each of the plural case images indicating an image in which a progression stage of a disease is different;
• execute a first selection process that includes
• calculating a first similarity about a first site where the first abnormality appears between each of the one or more case images included in the first case image group and the image of the subject, and
• obtaining a second case image group from the first case image group in accordance with the first similarity with respect to each of the one or more case images included in the first case image group;
• execute a process that includes
• obtaining a patient of the case image with respect to each of the case images included in the second case image group,
• obtaining another case image which is obtained by scanning later than the case image with respect to the patient, and
• for each patient in the second case image group, obtaining a second site associated with information indicating a second abnormality different from the first abnormality, the obtaining of the second site being performed by using any of the case image and the other case image with respect to each patient in the second case image group;
• execute an area limitation process that includes obtaining a region in the image of the subject, the obtained region being a region corresponding to an area in which a prescribed number or more of second sites each associated with the information indicating the second abnormality are overlapped with each other by the obtaining of the second site for each patient;
• execute a process that includes detecting a third site where the second abnormality appears in the obtained region in the image of the subject;
• and
• execute a second selection process that includes
• calculating, for each of the one or more case images included in the second case image group, a second similarity between the second site and the third site by using the second case image group and the image of the subject, and
• selecting the case image from the second case image group in accordance with the second similarity with respect to each of the one or more case images included in the second case image group.

A non-transitory computer-readable storage medium for storing a diagnosis program that causes a processor to execute processing for route search, the processing comprising:

• executing a process that includes
• obtaining a first case image group from among plural case images about each of plural patients, the first case image group including one or more case images, each of the one or more case images included in the first case image group having a same abnormality as a first abnormality detected from an image of a subject, each of the plural case images indicating an image in which a progression stage of a disease is different;
• executing a first selection process that includes
• calculating a first similarity about a first site where the first abnormality appears between each of the one or more case images included in the first case image group and the image of the subject, and
• obtaining a second case image group from the first case image group in accordance with the first similarity with respect to each of the one or more case images included in the first case image group;
• executing a process that includes
• obtaining a patient of the case image with respect to each of the case images included in the second case image group,
• obtaining another case image which is obtained by scanning later than the case image with respect to the patient, and
• for each patient in the second case image group, obtaining a second site associated with information indicating a second abnormality different from the first abnormality, the obtaining of the second site being performed by using any of the case image and the other case image with respect to each patient in the second case image group;
• executing an area limitation process that includes obtaining a region in the image of the subject, the obtained region being a region corresponding to an area in which a prescribed number or more of second sites each associated with the information indicating the second abnormality are overlapped with each other by the obtaining of the second site for each patient;
• executing a process that includes detecting a third site where the second abnormality appears in the obtained region in the image of the subject;
• and
• executing a second selection process that includes
• calculating, for each of the one or more case images included in the second case image group, a second similarity between the second site and the third site by using the second case image group and the image of the subject, and
• selecting the case image from the second case image group in accordance with the second similarity with respect to each of the one or more case images included in the second case image group.