LaMoD: Latent Motion Diffusion Model For Myocardial Strain Generation

TL;DR

LaMoD addresses the challenge of obtaining accurate myocardial motion and strain from standard CMR by leveraging a pre-trained registration encoder to learn latent motion features and a latent diffusion model supervised by DENSE ground-truth. The method reconstructs realistic, time-series deformation fields in a latent space and decodes them into high-quality motion maps, enabling DENSE-like accuracy without extra acquisitions at test time. Key contributions include integrating latent diffusion in motion spaces, and showing improved motion and segmental strain accuracy over state-of-the-art baselines on both DENSE and cine CMR data. The approach has potential to enhance cardiac disease assessment and planning in busy clinical workflows by maximizing information from routinely acquired CMR scans. $v_t$ and $\phi_t$ notation from LDDMM and DDPM-style diffusion underpin the mathematical framework.

Abstract

Motion and deformation analysis of cardiac magnetic resonance (CMR) imaging videos is crucial for assessing myocardial strain of patients with abnormal heart functions. Recent advances in deep learning-based image registration algorithms have shown promising results in predicting motion fields from routinely acquired CMR sequences. However, their accuracy often diminishes in regions with subtle appearance changes, with errors propagating over time. Advanced imaging techniques, such as displacement encoding with stimulated echoes (DENSE) CMR, offer highly accurate and reproducible motion data but require additional image acquisition, which poses challenges in busy clinical flows. In this paper, we introduce a novel Latent Motion Diffusion model (LaMoD) to predict highly accurate DENSE motions from standard CMR videos. More specifically, our method first employs an encoder from a pre-trained registration network that learns latent motion features (also considered as deformation-based shape features) from image sequences. Supervised by the ground-truth motion provided by DENSE, LaMoD then leverages a probabilistic latent diffusion model to reconstruct accurate motion from these extracted features. Experimental results demonstrate that our proposed method, LaMoD, significantly improves the accuracy of motion analysis in standard CMR images; hence improving myocardial strain analysis in clinical settings for cardiac patients. Our code is publicly available at https://github.com/jr-xing/LaMoD.

Figures (5)

• Figure 1: An overview of our proposed network framework. (A) Registration-based network to learn latent motion features represented by initial velocity fields. (B) Diffusion model in latent motion spaces.
• Figure 2: Panel A (healthy volunteer) & B (heart failure patient with left bundle branch block): exemplary comparison of end-systolic displacement and circumferential strain maps (Ecc) derived from DENSE input across all methods. Left to right: DENSE ground truth and predictions from our model vs. baselines. Top to bottom: enlarged view of selected displacement region; full displacements; circumferential strain maps (contraction in blue vs. stretch in red).
• Figure 3: Panel A (healthy volunteer) & B (heart failure patient with left bundle branch block): exemplary comparison of end-systolic displacement and circumferential strain maps (Ecc) derived from standard cine MRI videos.
• Figure 4: Left to right: a comparison of displacement field EPE error on DENSE; predicted segmental circumferential strain error on DENSE; and predicted segmental circumferential strain error on standard cine CMRs from our model vs. all baselines.
• Figure 5: A comparison of time-sequential myocardial strain generated by DENSE vs. our model throughout the cardiac cycle.