Fully Differentiable Correlation-driven 2D/3D Registration for X-ray to CT Image Fusion

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Chen, Minheng, Zhang, Zhirun, Gu, Shuheng, Ge, Zhangyang, Kong, Youyong
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909137612308480
author Chen, Minheng
Zhang, Zhirun
Gu, Shuheng
Ge, Zhangyang
Kong, Youyong
author_facet Chen, Minheng
Zhang, Zhirun
Gu, Shuheng
Ge, Zhangyang
Kong, Youyong
contents Image-based rigid 2D/3D registration is a critical technique for fluoroscopic guided surgical interventions. In recent years, some learning-based fully differentiable methods have produced beneficial outcomes while the process of feature extraction and gradient flow transmission still lack controllability and interpretability. To alleviate these problems, in this work, we propose a novel fully differentiable correlation-driven network using a dual-branch CNN-transformer encoder which enables the network to extract and separate low-frequency global features from high-frequency local features. A correlation-driven loss is further proposed for low-frequency feature and high-frequency feature decomposition based on embedded information. Besides, a training strategy that learns to approximate a convex-shape similarity function is applied in our work. We test our approach on a in-house datasetand show that it outperforms both existing fully differentiable learning-based registration approaches and the conventional optimization-based baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2402_02498
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Fully Differentiable Correlation-driven 2D/3D Registration for X-ray to CT Image Fusion
Chen, Minheng
Zhang, Zhirun
Gu, Shuheng
Ge, Zhangyang
Kong, Youyong
Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
Image-based rigid 2D/3D registration is a critical technique for fluoroscopic guided surgical interventions. In recent years, some learning-based fully differentiable methods have produced beneficial outcomes while the process of feature extraction and gradient flow transmission still lack controllability and interpretability. To alleviate these problems, in this work, we propose a novel fully differentiable correlation-driven network using a dual-branch CNN-transformer encoder which enables the network to extract and separate low-frequency global features from high-frequency local features. A correlation-driven loss is further proposed for low-frequency feature and high-frequency feature decomposition based on embedded information. Besides, a training strategy that learns to approximate a convex-shape similarity function is applied in our work. We test our approach on a in-house datasetand show that it outperforms both existing fully differentiable learning-based registration approaches and the conventional optimization-based baseline.
title Fully Differentiable Correlation-driven 2D/3D Registration for X-ray to CT Image Fusion
topic Image and Video Processing
Artificial Intelligence
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2402.02498