What is the clinical evidence from this study?

Study design: Other. Population: Ischemic Heart Disease (animal model). Intervention: Semi-supervised learning-based regularization (MLP) vs. Unregularized tracking methods. Primary outcome: Pearson correlation between crystal and image-derived peak strains.

April 20, 2021Open Access

Learning-Based Regularization for Cardiac Strain Analysis via Domain Adaptation

ALAllen LuYale University SAShawn S. AhnCardiac Imaging KTKevinminh TaYale University

Key Result

A semi-supervised learning-based regularization framework with biomechanical constraints improved the correlation between echocardiography-derived and sonomicrometer-derived peak cardiac strains compared to unregularized methods.

Structured PICO

Does a semi-supervised learning regularization approach improve motion estimation and infarct region identification in 3D+ time echocardiography?

Population

In vivo data with implanted sonomicrometers and postmortem excised hearts

Intervention

Semi-supervised Multi-Layered Perceptron (MLP) network with biomechanical constraints for domain adaptation

Outcome

Identification of infarct regions using estimated regional strain mapssurrogate

A novel semi-supervised neural network regularization framework improves motion estimation and strain analysis in 3D echocardiography, enabling accurate identification of infarct regions.

Limitations

Performance of feature tracking-based methods depends heavily on segmentation accuracy
Errors propagate temporally in segmentation

Abstract

Reliable motion estimation and strain analysis using 3D+ time echocardiography (4DE) for localization and characterization of myocardial injury is valuable for early detection and targeted interventions. However, motion estimation is difficult due to the low-SNR that stems from the inherent image properties of 4DE, and intelligent regularization is critical for producing reliable motion estimates. In this work, we incorporated the notion of domain adaptation into a supervised neural network regularization framework. We first propose a semi-supervised Multi-Layered Perceptron (MLP) network with biomechanical constraints for learning a latent representation that is shown to have more physiologically plausible displacements. We extended this framework to include a supervised loss term on synthetic data and showed the effects of biomechanical constraints on the network's ability for domain adaptation. We validated the semi-supervised regularization method on in vivo data with implanted sonomicrometers. Finally, we showed the ability of our semi-supervised learning regularization approach to identify infarct regions using estimated regional strain maps with good agreement to manually traced infarct regions from postmortem excised hearts.

Ask AI

Helpful

Bookmark

View Full Paper