Automatic detection of a pectoral muscle

摘要

The method involves selecting a limited number of selection candidate locations (AKO) (15) within breast exposure data based on image features in the breast exposure data. A limited number of selection candidate directions (AKR) at the selection candidate locations are selected (17) for formation of candidate tuples (KT) with one of the selection candidate locations and one of the selection candidate directions. A representation line (RL) and/or a representation surface (RE) are dissipated (25) by aggregation of the candidate tuples. Independent claims are also included for the following: (1) a detection system for automatic detection of a human pectoral muscle in medical breast exposure data of a patient (2) a medical retaining system (3) a computer program product having a set of instructions for executing a method for automatic detection of a human pectoral muscle in medical breast exposure data of a patient.