Dataset Information

Super-resolution 4D flow MRI to quantify aortic regurgitation using computational fluid dynamics and deep learning.

ABSTRACT: Changes in cardiovascular hemodynamics are closely related to the development of aortic regurgitation (AR), a type of valvular heart disease. Metrics derived from blood flows are used to indicate AR onset and evaluate its severity. These metrics can be non-invasively obtained using four-dimensional (4D) flow magnetic resonance imaging (MRI), where accuracy is primarily dependent on spatial resolution. However, insufficient resolution often results from limitations in 4D flow MRI and complex aortic regurgitation hemodynamics. To address this, computational fluid dynamics simulations were transformed into synthetic 4D flow MRI data and used to train a variety of neural networks. These networks generated super-resolution, full-field phase images with an upsample factor of 4. Results showed decreased velocity error, high structural similarity scores, and improved learning capabilities from previous work. Further validation was performed on two sets of in vivo 4D flow MRI data and demonstrated success in de-noising flow images. This approach presents an opportunity to comprehensively analyse AR hemodynamics in a non-invasive manner.

SUBMITTER: Long D

PROVIDER: S-EPMC10220149 | biostudies-literature | 2023 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Super-resolution 4D flow MRI to quantify aortic regurgitation using computational fluid dynamics and deep learning.

Long Derek D McMurdo Cameron C Ferdian Edward E Mauger Charlène A CA Marlevi David D Nash Martyn P MP Young Alistair A AA

The international journal of cardiovascular imaging 20230223 6

Changes in cardiovascular hemodynamics are closely related to the development of aortic regurgitation (AR), a type of valvular heart disease. Metrics derived from blood flows are used to indicate AR onset and evaluate its severity. These metrics can be non-invasively obtained using four-dimensional (4D) flow magnetic resonance imaging (MRI), where accuracy is primarily dependent on spatial resolution. However, insufficient resolution often results from limitations in 4D flow MRI and complex aort ...[more]

PMID: 36820960

Similar Datasets

Project description:PurposeTo compare the accuracy of semiautomated flow tracking with that of semiautomated valve tracking in the quantification of mitral valve (MV) regurgitation from clinical four-dimensional (4D) flow MRI data obtained in patients with mild, moderate, or severe MV regurgitation.Materials and methodsThe 4D flow MRI data were retrospectively collected from 30 patients (21 men; mean age, 61 years ± 10 [standard deviation]) who underwent 4D flow MRI from 2006 to 2016. Ten patients had mild MV regurgitation, nine had moderate MV regurgitation, and 11 had severe MV regurgitation, as diagnosed by using semiquantitative echocardiography. The regurgitant volume (Rvol) across the MV was obtained using three methods: indirect quantification of Rvol (RvolINDIRECT), semiautomated quantification of Rvol using valve tracking (RvolVALVE), and semiautomated quantification of Rvol using flow tracking (RvolFLOW). A second observer repeated the measurements. Aortic valve flow was quantified as well to test for intervalve consistency. The Wilcoxon signed rank test, orthogonal regression, Bland-Altman analysis, and coefficients of variation were used to assess agreement among measurements and between observers.ResultsRvolFLOW was higher (median, 24.8 mL; interquartile range [IQR], 14.3-45.7 mL) than RvolVALVE (median, 9.9 mL; IQR, 6.0-16.9 mL; P < .001). Both RvolFLOW and RvolVALVE differed significantly from RvolINDIRECT (median, 19.1 mL; IQR, 4.1-47.5 mL; P = .03). RvolFLOW agreed more with RvolINDIRECT (ŷ = 0.78x + 12, r = 0.88) than with RvolVALVE (ŷ = 0.16x + 8.1, r = 0.53). Bland-Altman analysis revealed underestimation of RvolVALVE in severe MV regurgitation. Interobserver agreement was excellent for RvolFLOW (r = 0.95, coefficient of variation = 27%) and moderate for RvolVALVE (r = 0.72, coefficient of variation = 57%). Orthogonal regression demonstrated better intervalve consistency for flow tracking (ŷ = 1.2x - 13.4, r = 0.82) than for valve tracking (ŷ = 2.7x - 92.4, r = 0.67).ConclusionFlow tracking enables more accurate 4D flow MRI-derived MV regurgitation quantification than valve tracking in terms of agreement with indirect quantification and intervalve consistency, particularly in severe MV regurgitation.Supplemental material is available for this article.© RSNA, 2020.

Dataset Information

Super-resolution 4D flow MRI to quantify aortic regurgitation using computational fluid dynamics and deep learning.

Publications

Super-resolution 4D flow MRI to quantify aortic regurgitation using computational fluid dynamics and deep learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets