Literature DB >> 33720465

Highly accelerated free-breathing real-time phase contrast cardiovascular MRI via complex-difference deep learning.

Hassan Haji-Valizadeh1, Rui Guo1, Selcuk Kucukseymen1, Amanda Paskavitz1, Xiaoying Cai1,2, Jennifer Rodriguez1, Patrick Pierce1, Beth Goddu1, Daniel Kim3, Warren Manning1,4, Reza Nezafat1.   

Abstract

PURPOSE: To develop and evaluate a real-time phase contrast (PC) MRI protocol via complex-difference deep learning (DL) framework.
METHODS: DL used two 3D U-nets to separately filter aliasing artifact from radial real-time velocity-compensated and complex-difference images. U-nets were trained with synthetic real-time PC generated from electrocardiograph (ECG) -gated, breath-hold, segmented PC (ECG-gated segmented PC) acquired at the ascending aorta of 510 patients. In 21 patients, free-breathing, ungated real-time (acceleration rate = 28.8) and ECG-gated segmented (acceleration rate = 2) PC were prospectively acquired at the ascending aorta. Hemodynamic parameters (cardiac output [CO], stroke volume [SV], and mean velocity at peak systole [peak mean velocity]) were measured for ECG-gated segmented and DL-filtered synthetic real-time PC and compared using Bland-Altman and linear regression analyses. Additionally, hemodynamic parameters were quantified from DL-filtered, compressed-sensing (CS) -reconstructed, and gridding reconstructed prospective real-time PC and compared to ECG-gated segmented PC.
RESULTS: Synthetic real-time PC with DL showed strong correlation (R > 0.98) and good agreement with ECG-gated segmented PC for quantified hemodynamic parameters (mean-difference: CO = -0.3 L/min, SV = -4.3 mL, peak mean velocity = -2.3 cm/s). On average, DL required 0.39 s/frame to filter prospective real-time PC, which was 4.6-fold faster than CS. Compared to CS, DL showed superior correlation, tighter limits of agreement (LOAs), better bias for peak mean velocity, and worse bias for CO and SV. Compared to gridding, DL showed similar correlation, tighter LOAs for CO and SV, similar bias for CO, and worse bias for SV and peak mean velocity.
CONCLUSION: The complex-difference DL framework accelerated real-time PC-MRI by nearly 28-fold, enabling rapid free-running real-time assessment of flow hemodynamics.
© 2021 International Society for Magnetic Resonance in Medicine.

Entities:  

Keywords:  GROG-GRASP; compressed sensing; deep learning; radial MRI; real-time phase contrast

Mesh:

Year:  2021        PMID: 33720465      PMCID: PMC8145775          DOI: 10.1002/mrm.28750

Source DB:  PubMed          Journal:  Magn Reson Med        ISSN: 0740-3194            Impact factor:   3.737


  34 in total

1.  Adaptive reconstruction of phased array MR imagery.

Authors:  D O Walsh; A F Gmitro; M W Marcellin
Journal:  Magn Reson Med       Date:  2000-05       Impact factor: 4.668

2.  An optimal radial profile order based on the Golden Ratio for time-resolved MRI.

Authors:  Stefanie Winkelmann; Tobias Schaeffter; Thomas Koehler; Holger Eggers; Olaf Doessel
Journal:  IEEE Trans Med Imaging       Date:  2007-01       Impact factor: 10.048

3.  A Deep Cascade of Convolutional Neural Networks for Dynamic MR Image Reconstruction.

Authors:  Jo Schlemper; Jose Caballero; Joseph V Hajnal; Anthony N Price; Daniel Rueckert
Journal:  IEEE Trans Med Imaging       Date:  2017-10-13       Impact factor: 10.048

4.  Deep complex convolutional network for fast reconstruction of 3D late gadolinium enhancement cardiac MRI.

Authors:  Hossam El-Rewaidy; Ulf Neisius; Jennifer Mancio; Selcuk Kucukseymen; Jennifer Rodriguez; Amanda Paskavitz; Bjoern Menze; Reza Nezafat
Journal:  NMR Biomed       Date:  2020-04-30       Impact factor: 4.044

5.  Shared velocity encoding: a method to improve the temporal resolution of phase-contrast velocity measurements.

Authors:  Hung-Yu Lin; Jacob A Bender; Yu Ding; Yiu-Cho Chung; Alice M Hinton; Michael L Pennell; Kevin K Whitehead; Subha V Raman; Orlando P Simonetti
Journal:  Magn Reson Med       Date:  2011-12-02       Impact factor: 4.668

6.  Rapid Reconstruction of Four-dimensional MR Angiography of the Thoracic Aorta Using a Convolutional Neural Network.

Authors:  Hassan Haji-Valizadeh; Daming Shen; Ryan J Avery; Ali M Serhal; Florian A Schiffers; Aggelos K Katsaggelos; Oliver S Cossairt; Daniel Kim
Journal:  Radiol Cardiothorac Imaging       Date:  2020-06-25

7.  Model-based reconstruction for real-time phase-contrast flow MRI: Improved spatiotemporal accuracy.

Authors:  Zhengguo Tan; Volkert Roeloffs; Dirk Voit; Arun A Joseph; Markus Untenberger; K Dietmar Merboldt; Jens Frahm
Journal:  Magn Reson Med       Date:  2016-03-07       Impact factor: 4.668

8.  Optimization and validation of accelerated golden-angle radial sparse MRI reconstruction with self-calibrating GRAPPA operator gridding.

Authors:  Thomas Benkert; Ye Tian; Chenchan Huang; Edward V R DiBella; Hersh Chandarana; Li Feng
Journal:  Magn Reson Med       Date:  2017-11-28       Impact factor: 4.668

9.  Rapid dealiasing of undersampled, non-Cartesian cardiac perfusion images using U-net.

Authors:  Lexiaozi Fan; Daming Shen; Hassan Haji-Valizadeh; Nivedita K Naresh; James C Carr; Benjamin H Freed; Daniel C Lee; Daniel Kim
Journal:  NMR Biomed       Date:  2020-01-14       Impact factor: 4.044

10.  Trajectory optimized NUFFT: Faster non-Cartesian MRI reconstruction through prior knowledge and parallel architectures.

Authors:  David S Smith; Saikat Sengupta; Seth A Smith; E Brian Welch
Journal:  Magn Reson Med       Date:  2018-10-17       Impact factor: 4.668

View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.