Improved strain analysis of cine images by deep learning from DENSE: Comparison of a 3D Unet and an optical-flow net

Yu Wang, Changyu Sun, Sona Ghadimi, Auger Daniel, Pierre Croisille, Magalie Viallon, Jie Cao, Cheng Yang, Andrew D. Scott, Pedro Ferreira, John N. Oshinski, Daniel B. Ennis, Kenneth C. Bilchick, Frederick H. EpsteinProceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition 2023 · DOI

Abstract

Cine DENSE provides both myocardial contours and intramyocardial displacements. We propose to use DENSE to train deep networks to predict intramyocardial motion from contour motion. Two workflows were implemented: a two-step FlowNet2-based framework with a through-time correction network and a 3D (2D+t) Unet framework. Both networks depicted cardiac contraction and abnormal motion patterns. The 3D Unet showed excellent reliability for global circumferential strain (E cc ) and good reliability for segmental E cc , and it outperformed commercial FT for both global and segmental E cc .
BibTeX
@inproceedings{wang2023improved,
  author = {Yu Wang and Changyu Sun and Sona Ghadimi and Auger Daniel and Pierre Croisille and Magalie Viallon and Jie Cao and Cheng Yang and Andrew D. Scott and Pedro Ferreira and John N. Oshinski and Daniel B. Ennis and Kenneth C. Bilchick and Frederick H. Epstein},
  title = {{Improved strain analysis of cine images by deep learning from DENSE: Comparison of a 3D Unet and an optical-flow net}},
  booktitle = {Proceedings on CD-ROM - International Society for Magnetic Resonance in Medicine. Scientific Meeting and Exhibition/Proceedings of the International Society for Magnetic Resonance in Medicine, Scientific Meeting and Exhibition},
  year = {2023},
  doi = {10.58530/2022/4928},
}

Loading related content...

Figure