Instance-specific brain image registration: learning deformable fields from a single image pair without external datasets
Physics in Medicine & Biology · 14.11.2026 · 71 (21) · 215035
DOI: 10.1088/1361-6560/aebd37 ↗Kurz gefasst
Fachartikel in der Zeitschrift Physics in Medicine & Biology, online erschienen am 13.11.2026. Verfasst von Hao Lin und 1 weiteren Autor:innen. Beteiligte Einrichtung: School of Software Engineering, Xi'an Jiaotong University, No. 28 Xianning West Road, Xi'an 710049, People's Republic of China. Hauptthemen laut MEDLINE: Brain, Image Processing, Computer-Assisted, Deep Learning. Die Arbeit nennt 45 Literaturangaben. Gefördert von: the Key R&D Plan of Shaanxi Province, Program, the National Natural Science Foundation of China, the Xi’an Jiaotong University Free Exploration and Innovation Program–Student Category.
Abstract (Originalsprache)
Objective. To eliminate the reliance on large-scale external training datasets and achieve high-precision alignment for complex brain structures by overcoming the prevalent challenges of data scarcity and domain shifts in current deep learning-based registration methods. Approach. We propose a novel Instance-Specific Deep Brain Registration Framework that independently trains a dedicated model for each specific image pair. By reconceptualizing a single image pair as an internal dataset generator, we introduce a Multi-Scale Random Patch Learning strategy to construct a diverse local-to-global sample space. Architecturally, we design a Context-Aware Flow Predictor integrated with a Recurrent Context Transfer mechanism within a coarse-to-fine paradigm. Furthermore, a Topology-Aware Adaptive Optimization strategy is incorporated during the inference phase. By dynamically tracking the local spatial folding status, it adaptively adjusts spatial regularization to effectively suppress topological distortions while maximizing voxel-level accuracy. Main results. Extensive experiments on mono-modal, multi-modal, and pathological brain MRI datasets demonstrate that our method can handle large anatomical deformations without relying on external training datasets. It achieves highly competitive registration performance compared to mainstream population-based and implicit neural representation methods. Crucially, it balances this alignment precision with effective numerical topology preservation. Significance. This framework challenges the traditional population-based learning paradigm, offering a highly flexible, robust, and promising solution for practical neuroimage analysis, particularly in clinical scenarios characterized by extreme data scarcity and complex cross-modality variations.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Software Engineering, Xi'an Jiaotong University, No. 28 Xianning West Road, Xi'an 710049, People's Republic of China.
ORCID-Profile: Hao Lin, Yonghong Song
Förderung
- the Key R&D Plan of Shaanxi Province, Program
- the National Natural Science Foundation of China
- the Xi’an Jiaotong University Free Exploration and Innovation Program–Student Category
Zitieren
APA
Lin, H., & Song, Y. (2026). Instance-specific brain image registration: learning deformable fields from a single image pair without external datasets. Physics in Medicine & Biology, 71(21), 215035. https://doi.org/10.1088/1361-6560/aebd37
Vancouver
Lin H, Song Y. Instance-specific brain image registration: learning deformable fields from a single image pair without external datasets. Physics in Medicine & Biology. 2026;71(21):215035. doi:10.1088/1361-6560/aebd37
BibTeX
@article{Lin2026,
author = {Lin, Hao and Song, Yonghong},
title = {Instance-specific brain image registration: learning deformable fields from a single image pair without external datasets},
journal = {Physics in Medicine & Biology},
year = {2026},
volume = {71},
number = {21},
pages = {215035},
doi = {10.1088/1361-6560/aebd37}
}
EUVIMED