Ischemic region segmentation in rat heart photos using DRLSE algorithm

Abstract

Heart attack preceded by ischemia is responsible for many deaths worldwide. Thus, the detection of ischemic cardiac areas is very important not only to help the prevention of that mortal disease but also for teaching/learning purposes. This work presents the results of a new approach for ischemic region detection in rat heart photo. Such an approach is based on segmentation using “Distance Regularized Level Set Evolution” method (DRLSE). The DRLSE method is an improvement on “Level Set method”. Evolving Interfaces in geometry, fluid mechanics, computer vision and materials sciences, 1999). The advantage of DRLSE is that the restart of level set function is not necessary. It was verified that the best identification of the ischemic region was obtained by using the yellow channel image in the processing, instead of the other color channels. Results show that the present approach is able to fairly segment ischemic regions in heart photos, being suitable for teaching/learning purposes.