Novel MRI-derived quantitative biomarker for cardiac function applied to classifying ischemic cardiomyopathy within a...

Area of Science:

• Cardiology
• Medical Imaging
• Biomarker Discovery

Background:

• Regional left ventricular (LV) function characterization is crucial for prognosticating treatment response in ischemic cardiomyopathy.
• Objective metrics are needed to assess pathological remodeling and declining cardiac performance.

Purpose of the Study:

• To characterize LV function using 4D endocardial motion analysis.
• To define objective metrics for pathological remodeling and declining cardiac performance.
• To develop and validate a novel biomarker (RMS-P2PD) for patient classification.

Main Methods:

• Analysis of 4D cardiac MRI data from asymptomatic (MESA) and symptomatic (DETERMINE) cohorts.
• Segmentation of LV endocardium and computation of phase-to-phase Hausdorff distance (HD).
• Development of RMS-P2PD marker and its application in a Bayesian Rule Learning (BRL) framework for patient classification.

Main Results:

• The novel RMS-P2PD biomarker showed significant differences between symptomatic and asymptomatic cohorts (p<0.001).
• BRL classification accuracy improved from 83.8% (using LV-EF, LV-ESVI) to 91.9% with RMS-P2PD inclusion.
• Sensitivity and specificity for patient/control identification increased significantly with RMS-P2PD.

Conclusions:

• RMS-P2PD is a promising biomarker for identifying quantitative signs of pathological endocardial function.
• This marker can enhance standard imaging techniques for early detection of declining cardiac performance.
• RMS-P2PD may aid in prognosticating response and guiding therapy in ischemic cardiomyopathy.