Automatic myocardial infarction size extraction in an experimental murine model using an anatomical model

摘要

Experimental rodent models of induced ischemic injury have been extensively used in biomedical research to study molecular, cellular and histological alterations following myocar-dial infarction. These models are increasingly employed to assess the potential of newly developed therapies for functional restoration of the damaged heart. Such studies are based on myocardial infarction induction followed by different therapeutic interventions and subsequent analysis of the infarct size. This analysis is used to evaluate the extent to which such interventions meet recovery of the lost myocardial tissue. Infarct size is defined as the percentage of the left ventricle affected by coronary artery occlusion. The infarct size is traditionally estimated manually delineating the infarcted and normal tissue areas in the left ventricle of the excised heart. However, this is a time-consuming, arduous and prone to bias process. Herein, we developed an anatomic model, adapted through expectation maximization, which allows for fully automatic analysis of the data. Experimental validation is performed comparing the proposed approach with manual annotation. The results obtained through anatomical model adaptation were coherent with those manually obtained and the differences where never higher than 10%.