Exploring the Feasibility of Investigating Myocarditis Progression Using Histopathological Features from Whole-Slide...

Area of Science:

• Cardiovascular Pathology
• Computational Pathology
• Digital Health

Background:

• Myocarditis, an inflammatory heart muscle condition, can lead to severe outcomes like heart failure and sudden cardiac death.
• Current subtyping methods for myocarditis lack detailed pathological analysis, hindering effective healthcare guidance.
• Accurate patient stratification is crucial for timely and targeted therapeutic interventions in cardiac diseases.

Purpose of the Study:

• To explore the potential of clustering pathological features from H&E stained whole-slide images for myocarditis subtyping.
• To investigate if pathological image analysis can enhance the assessment of myocarditis progression.
• To identify distinct patient groups within myocarditis based on histopathological characteristics.

Main Methods:

• Utilized H&E stained whole-slide images from 39 heart transplant patients with myocarditis.
• Extracted 3,791 pathological features using nucleus segmentation via the StarDist network.
• Applied feature dimensionality reduction and K-Modes clustering on 181 selected features.

Main Results:

• Clustering analysis successfully identified two distinct patient groups.
• These groups showed significant differences in time from heart failure to transplantation and time from onset to transplantation.
• Demonstrated the feasibility of using histopathological image data for patient stratification.

Conclusions:

• Pathohistological image analysis is a feasible method to enhance progression assessment in myocarditis patients.
• This approach may improve diagnostic accuracy for myocarditis and related cardiac conditions.
• Findings suggest potential for facilitating targeted therapeutic interventions based on pathological subtypes.