Method and system for predicting shape of human body after treatment

A method for predicting a breast shape after a treatment comprising:

• a step of extracting multi-dimensional pre-operative feature vectors Fpre(i) which have, as elements, a plurality of feature variables that are selected in advance based on pre-operative three-dimensional breast shape data of a plurality of patients who have received an operational treatment;
• a step of calculating case classes classified into a plurality of classes by carrying out clustering processing of a group of pre-operative feature vectors Fpre(i) of the plurality of patients, and cluster centers Gpre(1) of the each case class;
• a step of extracting a multi-dimensional subject patient feature vector Fnew having the feature variables as elements based on three-dimensional breast shape data of a patient contemplating the treatment;
• a step of selecting, from a plurality of the cluster centers Gpre(1) after the clustering, a similar case class having a cluster center that has the shortest distance from the subject patient feature vector Fnew;
• a step of calculating pre-operative breast shape models Hpre (i=i1, i2, . . . , ik) as a breast shape model of the each case patient in which the arrangements of feature points obtained from pre-operative three-dimensional breast shape data of the each selected case patient belonging to the similar case class are normalized;
• a step of calculating post-operative breast shape models Hpost (i=i1, i2, . . . , ik) of the each case patient in which the arrangements of feature points obtained from post-operative three-dimensional breast shape data of the each selected case patient belonging to the similar case class are normalized;
• a step of calculating a vector average AVEpre of pre-operative breast shape models Hpre (i=i1, i2, . . . , ik) of each selected case patient belonging to the similar case class;
• a step of calculating a vector average AVEpost of post-operative breast shape models Hpost (i=i1, i2, . . . , ik) of each selected case patient belonging to the similar case class;
• a step of calculating a breast shape vector average difference AVEpost−AVEpre that is obtained by subtracting the vector average AVEpre of pre-operative breast shape models from the vector average AVEpost of post-operative breast shape models;
• a step of calculating a subject patient breast shape model Hnew as a breast shape model of the patient in which the arrangement of feature points obtained from the three-dimensional breast shape data of the patient as an evaluation subject is normalized;
• and
• a step of calculating a three-dimensional predicted breast shape model Hprd as predicted after operation by modifying the breast shape model Hnew of the patient as an evaluation subject with the breast shape vector average difference AVEpost−AVEpre of the each selected case patient belonging to the similar case class.