Whole-heart single breath-hold cardiac cine: A robust motion-compensated compressed sensing reconstruction method

Javier Royuela-del-Val, Muhammad Usman, Lucilio Cordero Grande, Marcos Martín-Fernández, Federico Simmross-Wattenberg, Claudia Prieto Vasquez, Carlos Alberola-López

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

1 Citation (Scopus)
154 Downloads (Pure)

Abstract

In this paper we propose a methodology to achieve single breath-hold whole-heart cine MRI with a temporal resolution of ∼46 ms and a spatial resolution of 2 × 2 mm 2 out of a previously described method (JW-tTV) for single slice reconstruction. Its feasibility is tested by itself and in comparison with another state-of-the-art reconstruction method (MASTeR); both methods are adapted to Golden Radial k-space trajectories. From a formal viewpoint, we make use of a realistic numerical phantom to have a ground truth of deformation fields so that the methods performances against noise can be quantified and the sparsity regularization parameter involved in the reconstructions can be fixed according to the signal to noise ratio. Phantom results show that the adapted JW-tTV method is more robust against noise and provides more precise motion estimations and better reconstructions than MASTeR. Both methods are then applied to the reconstruction of 12–14 short axis slices covering the whole heart on eight volunteers. Finer details are better preserved with JW-tTV. Ventricle volumes and ejection fractions were computed from the volunteers data and preliminary results show agreement with conventional multiple breath-hold Cartesian acquisitions.
Original languageEnglish
Title of host publicationRAMBO 2016/HVSMR 2016, Lecture Notes in Computer Science
Pages58-69
Number of pages12
Volume10129
ISBN (Electronic)978-3-319-52280-7
Publication statusPublished - 17 Jan 2017

Fingerprint

Dive into the research topics of 'Whole-heart single breath-hold cardiac cine: A robust motion-compensated compressed sensing reconstruction method'. Together they form a unique fingerprint.

Cite this