SuCor: Susceptibility Distortion Correction via Parameter-Free and Self-Regularized Optimal Transport

Sreekar Chigurupati; Eleftherios Garyfallidis

cs.CVMar 17, 2026

SuCor: Susceptibility Distortion Correction via Parameter-Free and Self-Regularized Optimal Transport

Authors: Sreekar Chigurupati, Eleftherios Garyfallidis

Organizations: Intelligent Systems Engineering, Indiana University, Bloomington, Indiana, USA.

Abstract

We present SuCor, a method for correcting susceptibility induced geometric distortions in echo planar imaging (EPI) using optimal transport (OT) along the phase encoding direction. Given a pair of reversed phase encoding EPI volumes, we model each column of the distortion field as a Wasserstein-2 barycentric displacement between the opposing-polarity intensity profiles. Regularization is performed in the spectral domain using a bending-energy penalty whose strength is selected automatically via the Morozov discrepancy principle, requiring no manual tuning. On a human connectome project (HCP) dataset with left-right/right-left b0 EPI pairs and a co-registered T1 structural reference, SuCor achieves a mean volumetric mutual information of 0.341 with the T1 image, compared to 0.317 for FSL TOPUP, while running in approximately 12 seconds on a single CPU core.

Figures & tables

Method	$T_{1}$ -MI	LR–RL NCC	RMSE	Runtime
Uncorrected	0.284	0.670	2215	-
TOPUP	0.317	0.964	706	55 min
SuCor	0.341	0.933	948	12 sec

Table 1: Volumetric (3D) evaluation of susceptibility correction methods.

T_{1}

-MI: mutual information with

T_{1}

structural (higher is better). LR–RL NCC: normalized cross-correlation between corrected LR and RL images (higher is better). RMSE: root mean squared LR–RL difference (lower is better).

Method	$z=27$	$z=55$	$z=83$
Uncorrected	0.296	0.503	0.519
TOPUP	0.342	0.476	0.512
SuCor	0.363	0.523	0.570

Table 2: Per-slice

T_{1}

-MI at three representative axial levels.

Figure 1: Multi-slice correction comparison. Columns show (left to right): uncorrected (LR+RL)/2 average, TOPUP, SuCor (ours), $T_{1}$ reference, $|$ LR $-$ RL $|$ before correction, and $|$ LR $-$ RL $|$ after SuCor correction. Three axial slices are shown at $z=27$ , $55$ , and $83$ .

Figure 2: Estimated displacement fields at $z=55$ . Top row: TOPUP field (Hz) and SuCor field (scaled to TOPUP units). Bottom row: cross-PE gradient magnitude $|\partial u/\partial y|$ , showing that SuCor preserves sharper field transitions at tissue boundaries.

Explore similar work

Unsupervised Susceptibility Distortion Correction of EPI without Calibration Scans via Image Translation-Based Registration
Jun 19, 2026Wooseung Kim, Sung-Hong ParkMedical Image Registration Magnetic Resonance Imaging
Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea
Distortion-Corrected Diffusion MRI Using Rotated-View EPI and Joint Field-Map/Image Estimation with Gaussian Primitives
Jun 30, 2026Wenqi Huang, Zhitao Li, Nan Wang +8Image Reconstruction Magnetic Resonance Imaging
Chair for AI in Healthcare and Medicine, Technical University of Munich (TUM) and TUM University Hospital, and the School of Computation, Information and Technology, TUM, Munich, Germany · Innovation Academy for Precision Measurement Science and Technology, Chinese Academy of Sciences, Wuhan, China · Department of Radiology, Stanford University, Stanford, CA, USA +2
FINDER: Zero-Shot Field-Integrated Network for Distortion-free EPI Reconstruction in Diffusion MRI
Mar 27, 2026Namgyu Han, Seong Dae Yun, Chaeeun Lim +7Image Reconstruction Diffusion MRI
Department of Artificial Intelligence and Robotics, Sejong University, Seoul 05006, South Korea · Institute of Neuroscience and Medicine 4, INM-4, Forschungszentrum Jülich, Jülich, Germany · Athinoula A. Martinos Center for Biomedical Imaging, Charlestown, MA 02129, USA +2

Jun 19, 2026J/K move · Enter open · S save

Unsupervised Susceptibility Distortion Correction of EPI without Calibration Scans via Image Translation-Based Registration

Wooseung Kim, Sung-Hong Park

Department of Bio and Brain Engineering, Korea Advanced Institute of Science and Technology, Daejeon, Republic of Korea

Medical Image Registration Magnetic Resonance Imaging

Open paper AskPDF arXiv

Functional magnetic resonance imaging (fMRI) utilizes echo-planar imaging (EPI) to capture blood-oxygen-level-dependent (BOLD) signals with high temporal resolution. However, EPI is inherently sensitive to magnetic field inhomogeneities, resulting in susceptibility-induced geometric distortions along the phase-encoding (PE) direction. To correct these distortions, conventional approaches rely on additional calibration scans, such as field maps or reverse PE acquisitions, which are not always available in practice. To overcome this limitation, we propose SACRED, a calibration scan-free susceptibility distortion correction framework that corrects geometric distortions via image translation-based registration using only a routinely acquired anatomical T1-weighted (T1w) image and a unidirectional PE BOLD image. SACRED employs an invertible neural network as the image translation backbone to bridge the contrast gap between BOLD and T1w images while enforcing structural consistency through a modality independent neighborhood descriptor. This design enables the use of a mono-contrast similarity objective to train the registration network in an unsupervised manner without requiring distortion-corrected BOLD images. In addition, we incorporate test-time adaptation (TTA) to further enhance performance on out-of-distribution (OOD) data at inference time. SACRED was evaluated on one in-distribution (ID) dataset and two OOD datasets, and was compared with representative fMRI distortion correction methods. The results demonstrate that SACRED significantly outperforms competing methods on both ID and OOD datasets, exhibiting robustness to scanner and population shifts, partly enabled by TTA. The code will be made publicly available upon acceptance.

Abstract

Explore similar work

Unsupervised Susceptibility Distortion Correction of EPI without Calibration Scans via Image Translation-Based Registration

Distortion-Corrected Diffusion MRI Using Rotated-View EPI and Joint Field-Map/Image Estimation with Gaussian Primitives

FINDER: Zero-Shot Field-Integrated Network for Distortion-free EPI Reconstruction in Diffusion MRI