Saved in:
Bibliographic Details
Main Authors: Chigurupati, Sreekar, Garyfallidis, Eleftherios
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.16758
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911523311452160
author Chigurupati, Sreekar
Garyfallidis, Eleftherios
author_facet Chigurupati, Sreekar
Garyfallidis, Eleftherios
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2603_16758
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SuCor: Susceptibility Distortion Correction via Parameter-Free and Self-Regularized Optimal Transport
Chigurupati, Sreekar
Garyfallidis, Eleftherios
Computer Vision and Pattern Recognition
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.
title SuCor: Susceptibility Distortion Correction via Parameter-Free and Self-Regularized Optimal Transport
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.16758