Automatic detection of osteolytic lesions in rat femur with bone metastases

Abstract

A method for the automatic detection of osteolytic lesions caused by bone metastases is presented. Osteolytic means that bone mass is lost and these lesions are visible as holes in the bone structure. For the analysis of the process of metastases and their response to therapies the measurement of these lesions is neccessary. As manual segmentation of all lesions is too complex for a larger study, automatic tools are needed. The challenging task of measuring missing structures is solved here by comparison of a modified bone with a healthy model. The algorithm is tested on rat femur bones. First tests have shown that the presented algorithm can be used for the global identification of osteolytic regions.