The present invention provides a case search device which is capable of obtaining a similar case which a physician desires, as appropriate to an application, and also provides a method for operation and program for operation thereof, and a similar case search system. Using a first weighting coefficient (CF1) corresponding to a degree of contribution of each type of lesion to the identification of a name of a disease, a weighting processing unit (65) carries out a weighting process upon a lesion-specific degree of similarity (LS) which is a degree of similarity between a subject lesion and a case lesion which are of the same type, and computes a post-processing lesion-specific similarity (LSA-1) corresponding to a first application of identifying the name of the disease. Using a second weighting coefficient (CF2) corresponding to a degree of contribution of each type of lesion to the identification of the degree of severity of the disease, a weighting process is carried out upon the lesion-specific degree of similarity (LS), and a post-processing lesion-specific similarity (LSA-2) corresponding to a second application of identifying the degree of severity of the disease is computed. On the basis of the post-processing lesion-specific similarities (LSA-1), (LSA-2), a total similarity (TS), which is the total similarity between a subject image (20T) and a case image (20C), is computed for each application.La présente invention concerne un dispositif de recherche de cas qui est capable d'obtenir un cas similaire qu'un médecin souhaite, selon le cas d'une application, et fournit également un procédé de fonctionnement et un programme de fonctionnement associé, et un système de recherche de cas similaire. En utilisant un premier coefficient de pondération (CF1) correspondant à un degré de contribution de chaque type de lésion à l'identification d'un nom d'une maladie, une unité de traitement de pondération (65) effectue un processus de pondération sur un degré de similarit
