Segmentation and Registration Methods in Short Axis Cardiac MRI and SPECT Images in Chagas Disease

Abstract

The relationship between the myocardial tissue damage, myocardial perfusion defects and gaps in the autonomic innervations is still poorly understood. This paper proposes methods capable of providing an integrated visualization and analysis of tissue injuries through MRI images, autonomic innervations and myocardial perfusion through SPECT images. The proposed method is based on segmentation of MRI and registration between MRI images and SPECT images using 123I-MIBG and 99mTc-MIBI as tracers. Fibrosis segmentation in MRI images was performed based on the algorithm of maximum Tsallis entropy. Nonrigid registrations method based on B-Spline were used to align image volumes. Results show correlation between fibrosis and denervation areas in the middle short axis was obtained by Spearman correlation (r=0.571, p-valueale;). The development tool provides a better understanding of te relatons injuries amd ischemia caused by Chagas disease.