000147075 001__ 147075
000147075 005__ 20241220120719.0
000147075 0247_ $$2doi$$a10.1109/ISBI56570.2024.10635118
000147075 0248_ $$2sideral$$a140849
000147075 037__ $$aART-2024-140849
000147075 041__ $$aeng
000147075 100__ $$aRamón-Júlvez, Ubaldo$$uUniversidad de Zaragoza
000147075 245__ $$aEPDiff-JF-Net: Adjoint Jacobi Fields for Diffeomorphic Registration Networks
000147075 260__ $$c2024
000147075 5203_ $$aLarge Deformation Diffeomorphic Metric Mapping (LDDMM) is a well-established diffeomorphic registration method, a critical step for many medical imaging applications. Of special interest are its geodesic shooting variants, which enable statistical shape analysis from transformations usable in computational anatomy studies. This paper introduces a novel deep learning-based unsupervised approach for diffeomorphic image registration called EPDiff-JF-Net. Our method predicts an initial velocity field, performs geodesic shooting to obtain the corresponding path of diffeomorphisms, and utilizes an adjoint Jacobi fields layer to calculate the relevant gradients for training from parallel transport along the geodesic. The model is trained in a fully unsupervised end-to-end manner, with no requirement of ground-truth in the training loss. Experimental results on 3D brain MRI datasets demonstrate the effectiveness of EPDiff-JF-Net, outperforming EPDiff-based LDDMM and deep learning methods while significantly reducing computation time.
000147075 536__ $$9info:eu-repo/grantAgreement/ES/DGA/T64-20R$$9info:eu-repo/grantAgreement/ES/MICINN/PID2019-104358RB-I00$$9info:eu-repo/grantAgreement/ES/MICINN/PID2022-138703OB-I00
000147075 540__ $$9info:eu-repo/semantics/closedAccess$$aAll rights reserved$$uhttp://www.europeana.eu/rights/rr-f/
000147075 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/publishedVersion
000147075 700__ $$0(orcid)0000-0003-1270-5852$$aHernández, Mónica$$uUniversidad de Zaragoza
000147075 700__ $$0(orcid)0000-0002-9109-5337$$aMayordomo, Elvira$$uUniversidad de Zaragoza
000147075 7102_ $$15007$$2570$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Lenguajes y Sistemas Inf.
000147075 773__ $$g2024 (2024), 1-5$$pProc. (IEEE Int. Symp. Biomed. Imaging)$$tProceedings - International Symposium on Biomedical Imaging$$x1945-7928
000147075 8564_ $$s2168389$$uhttps://zaguan.unizar.es/record/147075/files/texto_completo.pdf$$yVersión publicada
000147075 8564_ $$s3063034$$uhttps://zaguan.unizar.es/record/147075/files/texto_completo.jpg?subformat=icon$$xicon$$yVersión publicada
000147075 909CO $$ooai:zaguan.unizar.es:147075$$particulos$$pdriver
000147075 951__ $$a2024-12-20-12:05:37
000147075 980__ $$aARTICLE