eduzhai > Applied Sciences > Engineering >

Anatomy-Aware Cardiac Motion Estimation

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 10 pages

Abstract: Cardiac motion estimation is critical to the assessment of cardiac function.Myocardium feature tracking (FT) can directly estimate cardiac motion from cineMRI, which requires no special scanning procedure. However, current deeplearning-based FT methods may result in unrealistic myocardium shapes since thelearning is solely guided by image intensities without considering anatomy. Onthe other hand, motion estimation through learning is challenging becauseground-truth motion fields are almost impossible to obtain. In this study, wepropose a novel Anatomy-Aware Tracker (AATracker) for cardiac motion estimationthat preserves anatomy by weak supervision. A convolutional variationalautoencoder (VAE) is trained to encapsulate realistic myocardium shapes. Abaseline dense motion tracker is trained to approximate the motion fields andthen refined to estimate anatomy-aware motion fields under the weak supervisionfrom the VAE. We evaluate the proposed method on long-axis cardiac cine MRI,which has more complex myocardium appearances and motions than short-axis.Compared with other methods, AATracker significantly improves the trackingperformance and provides visually more realistic tracking results,demonstrating the effectiveness of the proposed weakly-supervision scheme incardiac motion estimation.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×