Tissue classification on histopathological imaging of a porcine model of myocardial infarction.
Myocardial infarction, commonly known as heart attack, is a condition that affects worldwide to millions of people per year. Furthermore, this condition gives rise to multiple complications particularly to arrhythmias and cardiac failure which can lead to sudden death. This Bachelor Thesis aims to build a classifier that enables the segmentation of the different component tissues on histopathological samples of Myocardial Infarction in order to provide a tool that facilitates finding the relationships between the tissue composition of the heart affected by Myocardial Infarction and the associate complications. For the evaluation and creation of that classifier, images from a porcine model of myocardial infarction were donated by Danial Herzka from John Hopkings University. The methodology followed basically consist on the use of manual segmentation and state of the art segmentation algorithms for both data reduction and image segmentation. The results of this thesis could be used in several research applications, such as testing of drugs or regenerative therapy approaches that influence the healing and remodeling of the heart on a preclinical stage.