Deep Learning Analysis of Cardiac MRI in Legacy Datasets: Multi-Ethnic Study of Atherosclerosis

TL;DR

This paper presents a deep learning pipeline for automated analysis of legacy cardiac MRI data from the MESA study, enabling precise heart structure quantification and risk factor analysis across a large, multi-ethnic cohort.

Contribution

It introduces a novel deep learning-based atlas construction method tailored for legacy MRI datasets, overcoming challenges of differing pulse sequences and enabling automated cardiac analysis.

Findings

High accuracy in landmark detection and segmentation comparable to manual annotations

Automated atlases show similar risk factor relationships as manual ones

Pipeline enables large-scale, automated cardiac morphology studies in legacy datasets

Abstract

The shape and motion of the heart provide essential clues to understanding the mechanisms of cardiovascular disease. With the advent of large-scale cardiac imaging data, statistical atlases become a powerful tool to provide automated and precise quantification of the status of patient-specific heart geometry with respect to reference populations. The Multi-Ethnic Study of Atherosclerosis (MESA), begun in 2000, was the first large cohort study to incorporate cardiovascular MRI in over 5000 participants, and there is now a wealth of follow-up data over 20 years. Building a machine learning based automated analysis is necessary to extract the additional imaging information necessary for expanding original manual analyses. However, machine learning tools trained on MRI datasets with different pulse sequences fail on such legacy datasets. Here, we describe an automated atlas construction pipeline using deep learning methods applied to the legacy cardiac MRI data in MESA. For detection of anatomical cardiac landmark points, a modified VGGNet convolutional neural network architecture was used in conjunction with a transfer learning sequence between two-chamber, four-chamber, and short-axis MRI views. A U-Net architecture was used for detection of the endocardial and epicardial boundaries in short axis images. Both network architectures resulted in good segmentation and landmark detection accuracies compared with inter-observer variations. Statistical relationships with common risk factors were similar between atlases derived from automated vs manual annotations. The automated atlas can be employed in future studies to examine the relationships between cardiac morphology and future events.