Exploring the Feasibility of Investigating Myocarditis Progression Using Histopathological Features from Whole-Slide Images of H&E Tissue

Myocarditis is an inflammatory condition affecting the heart muscle, if left unaddressed, may lead to severe complications such as heart failure, arrhythmias, and sudden cardiac death. Current subtyping methods lack sufficient pathological analysis to guide healthcare. Here, we present a clustering analysis using pathological features of H&E stained whole-slide images of from 39 patients with myocarditis who have gone through heart transplantation. Totally, 3,791 pathological features were extracted using each nucleus segmented from StarDist segmentation network. After feature dimensionality reduction, 181 features were selected for clustering with K-Modes method. Two groups of patients with significant differences in time from heart failure to transplantation and time from onset to transplantation were identified. This study demonstrates the feasibility of using pathohistological images to enhance the progression assessment of patients with myocarditis, which may help improve diagnostic accuracy and facilitating targeted therapeutic interventions for cardiac-related diseases.