Learning Correspondences of Cardiac Motion from Images Using Biomechanics-Informed Modeling

Zhang, Xiaoran; You, Chenyu; Ahn, Shawn S.; Zhuang, Jian; Staib, Lawrence H.; Duncan, James S.

doi:10.1007/978-3-031-23443-9_2

Cited by 4 publications

(2 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many other studies on cardiac MRI motion estimation have also utilized the registration approach with different temporal smoothness strategies [13,21]. Recently, researchers have also incorporated biomechanical modeling knowledge into deep learning networks with the aim of improving the generalizability of motion estimation performance [22,26].…”

Section: Introductionmentioning

confidence: 99%

“…The motion of the myocardium is not arbitrary, and various spatial regularization techniques have been explored in the literature, such as the divergence penalty [15] for enforcing incompressibility, the rigidity penalty for smoothness [24], and the elastic strain energy for mechanical correctness [19], among others. Many deep learning models have incorporated these regularization techniques into their work [1,8,26]. However, these regularization techniques are usually based on first-or second-order derivatives of the displacement vector field, which are applied at the pixel level in discrete implementation and are insufficient for studying the myocardium in echocardiography.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Polyaffine Transformation Learning for Echocardiography Motion Estimation

Yang

Sermesant

2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Echocardiography plays an important role in the diagnosis of cardiac dysfunction. In particular, motion estimation in echocardiography is a challenging task since ultrasound images suffer largely from low signal-to-noise ratio and out-of-view problems. Current deep learningbased models for cardiac motion estimation in the literature estimate the dense motion field with spatial regularization. However, the underlying spatial regularization can only cover a very small region in the neighborhood, which is not enough for a smooth and realistic motion field for the myocardium in echocardiography. In order to improve the performance and quality with deep learning networks, we propose applying polyaffine transformation for motion estimation, which intrinsically regularizes the myocardium motion to be polyaffine. Our thorough experiments demonstrate that the proposed method not only presents better evaluation metrics on the registration of cardiac structures but also shows great potential in abnormal wall motion detection.

show abstract