Investigation of Blind Deconvolution Method with Total Variation Regularization in Cardiac Cine Magnetic Resonance Imaging
Year: 2025
Type: article
Abstract: Various studies have been conducted to reduce the blurring caused by movement in cine magnetic resonance imaging (MRI) of the heart. This study proposed a blind deconvolution method using a total variation regularization algorithm to remove blurring in cardiac cine magnetic resonance (MR) images. The MR data were acquired using a rat cardiac cine sequence in an open format. We investigated a blind deconvolution method with a total variation regularization, incorporating a 3-dimensional point-spread function on cardiac cine MRI. The gradient of magnitude (GM) and perceptual sharpness index (PSI) were used to evaluate the usefulness of the proposed deblurring method. We confirmed that the proposed method can reduce temporal blur relatively efficiently compared with the generalized variation-based deblurring algorithm. In particular, the GM and PSI values of the cardiac cine MR image corrected using the proposed method were improved by approximately 7.59 and 4.18 times, respectively, compared with the degraded image. We achieved improved image quality by validating a blind deconvolution method using a total variation regularization algorithm on the cardiac cine MR images of small animals.
Source: Electronics
Authors: Kyuseok Kim, Youngjin Lee
Institutions: Eulji University, Gachon University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,896
Award: G5223761292
OpenAlex ID: W4407507538