Deep learning motion correction of quantitative stress perfusion cardiovascular magnetic resonance

2025-10-01

AI Summary

Quantitative first-pass perfusion cardiovascular magnetic resonance (CMR) is essential for myocardial ischemia assessment, yet conventional registration-based motion correction is slow, non-robust, and sensitive to acquisition variability. This paper proposes an unsupervised deep learning framework for one-click, rapid motion correction of dynamic perfusion sequences. It integrates a 3D deformable registration network with robust principal component analysis (RPCA) to suppress contrast-agent–induced artifacts and enhance cross-vendor generalizability; concurrently, it aligns arterial input functions with proton density images to ensure quantitative accuracy. Experiments demonstrate a 15× speedup in processing time, a myocardial segmentation Dice score of 0.92, significantly improved temporal signal curve smoothness, reduced standard deviation in myocardial blood flow estimates, and marked reduction in motion artifacts.

Technology Category

• Computer Vision: Motion & Tracking
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: Motion and Path Planning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Background: Quantitative stress perfusion cardiovascular magnetic resonance (CMR) is a powerful tool for assessing myocardial ischemia. Motion correction is essential for accurate pixel-wise mapping but traditional registration-based methods are slow and sensitive to acquisition variability, limiting robustness and scalability. Methods: We developed an unsupervised deep learning-based motion correction pipeline that replaces iterative registration with efficient one-shot estimation. The method corrects motion in three steps and uses robust principal component analysis to reduce contrast-related effects. It aligns the perfusion series and auxiliary images (arterial input function and proton density-weighted series). Models were trained and validated on multivendor data from 201 patients, with 38 held out for testing. Performance was assessed via temporal alignment and quantitative perfusion values, compared to a previously published registration-based method. Results: The deep learning approach significantly improved temporal smoothness of time-intensity curves (p<0.001). Myocardial alignment (Dice = 0.92 (0.04) and 0.91 (0.05)) was comparable to the baseline and superior to before registration (Dice = 0.80 (0.09), p<0.001). Perfusion maps showed reduced motion, with lower standard deviation in the myocardium (0.52 (0.39) ml/min/g) compared to baseline (0.55 (0.44) ml/min/g). Processing time was reduced 15-fold. Conclusion: This deep learning pipeline enables fast, robust motion correction for stress perfusion CMR, improving accuracy across dynamic and auxiliary images. Trained on multivendor data, it generalizes across sequences and may facilitate broader clinical adoption of quantitative perfusion imaging.

Problem

• Correcting motion artifacts in quantitative stress perfusion cardiac MRI
• Replacing slow traditional registration with efficient deep learning
• Improving robustness and scalability of myocardial perfusion mapping

Innovation

• Unsupervised deep learning replaces iterative registration
• Robust principal component analysis reduces contrast effects
• Efficient one-shot estimation aligns perfusion series and auxiliary images