Imperial College London

ProfessorDavidFirmin

Faculty of MedicineNational Heart & Lung Institute

Emeritus Professor of Biomedical Imaging
 
 
 
//

Contact

 

+44 (0)20 7351 8801d.firmin

 
 
//

Location

 

Cardiovascular MR UnitRoyal Brompton Campus

//

Summary

 

Publications

Citation

BibTex format

@unpublished{Xing:2022,
author = {Xing, X and Wu, Y and Firmin, D and Gatehouse, P and Yang, G},
publisher = {ArXiv},
title = {Synthetic velocity mapping cardiac MRI coupled with automated left ventricle segmentation},
url = {http://arxiv.org/abs/2110.01304v2},
year = {2022}
}

RIS format (EndNote, RefMan)

TY  - UNPB
AB - Temporal patterns of cardiac motion provide important information for cardiacdisease diagnosis. This pattern could be obtained by three-directional CINEmulti-slice left ventricular myocardial velocity mapping (3Dir MVM), which is acardiac MR technique providing magnitude and phase information of themyocardial motion simultaneously. However, long acquisition time limits theusage of this technique by causing breathing artifacts, while shortening thetime causes low temporal resolution and may provide an inaccurate assessment ofcardiac motion. In this study, we proposed a frame synthesis algorithm toincrease the temporal resolution of 3Dir MVM data. Our algorithm is featured by1) three attention-based encoders which accept magnitude images, phase images,and myocardium segmentation masks respectively as inputs; 2) three decodersthat output the interpolated frames and corresponding myocardium segmentationresults; and 3) loss functions highlighting myocardium pixels. Our algorithmcan not only increase the temporal resolution 3Dir MVMs, but can also generatesthe myocardium segmentation results at the same time.
AU - Xing,X
AU - Wu,Y
AU - Firmin,D
AU - Gatehouse,P
AU - Yang,G
PB - ArXiv
PY - 2022///
TI - Synthetic velocity mapping cardiac MRI coupled with automated left ventricle segmentation
UR - http://arxiv.org/abs/2110.01304v2
UR - http://hdl.handle.net/10044/1/95829
ER -