TotalRegistrator: Towards a Lightweight Foundation Model for CT Image Registration

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pham, Xuan Loc, Vuurberg, Gwendolyn, Doppen, Marjan, Roosen, Joey, Stille, Tip, Ha, Thi Quynh, Quach, Thuy Duong, Dang, Quoc Vu, Luu, Manh Ha, Smit, Ewoud J., Mai, Hong Son, Heinrich, Mattias, van Ginneken, Bram, Prokop, Mathias, Hering, Alessa
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909724945940480
author Pham, Xuan Loc
Vuurberg, Gwendolyn
Doppen, Marjan
Roosen, Joey
Stille, Tip
Ha, Thi Quynh
Quach, Thuy Duong
Dang, Quoc Vu
Luu, Manh Ha
Smit, Ewoud J.
Mai, Hong Son
Heinrich, Mattias
van Ginneken, Bram
Prokop, Mathias
Hering, Alessa
author_facet Pham, Xuan Loc
Vuurberg, Gwendolyn
Doppen, Marjan
Roosen, Joey
Stille, Tip
Ha, Thi Quynh
Quach, Thuy Duong
Dang, Quoc Vu
Luu, Manh Ha
Smit, Ewoud J.
Mai, Hong Son
Heinrich, Mattias
van Ginneken, Bram
Prokop, Mathias
Hering, Alessa
contents Image registration is a fundamental technique in the analysis of longitudinal and multi-phase CT images within clinical practice. However, most existing methods are tailored for single-organ applications, limiting their generalizability to other anatomical regions. This work presents TotalRegistrator, an image registration framework capable of aligning multiple anatomical regions simultaneously using a standard UNet architecture and a novel field decomposition strategy. The model is lightweight, requiring only 11GB of GPU memory for training. To train and evaluate our method, we constructed a large-scale longitudinal dataset comprising 695 whole-body (thorax-abdomen-pelvic) paired CT scans from individual patients acquired at different time points. We benchmarked TotalRegistrator against a generic classical iterative algorithm and a recent foundation model for image registration. To further assess robustness and generalizability, we evaluated our model on three external datasets: the public thoracic and abdominal datasets from the Learn2Reg challenge, and a private multiphase abdominal dataset from a collaborating hospital. Experimental results on the in-house dataset show that the proposed approach generally surpasses baseline methods in multi-organ abdominal registration, with a slight drop in lung alignment performance. On out-of-distribution datasets, it achieved competitive results compared to leading single-organ models, despite not being fine-tuned for those tasks, demonstrating strong generalizability. The source code will be publicly available at: https://github.com/DIAGNijmegen/oncology_image_registration.git.
format Preprint
id arxiv_https___arxiv_org_abs_2508_04450
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TotalRegistrator: Towards a Lightweight Foundation Model for CT Image Registration
Pham, Xuan Loc
Vuurberg, Gwendolyn
Doppen, Marjan
Roosen, Joey
Stille, Tip
Ha, Thi Quynh
Quach, Thuy Duong
Dang, Quoc Vu
Luu, Manh Ha
Smit, Ewoud J.
Mai, Hong Son
Heinrich, Mattias
van Ginneken, Bram
Prokop, Mathias
Hering, Alessa
Image and Video Processing
Computer Vision and Pattern Recognition
Image registration is a fundamental technique in the analysis of longitudinal and multi-phase CT images within clinical practice. However, most existing methods are tailored for single-organ applications, limiting their generalizability to other anatomical regions. This work presents TotalRegistrator, an image registration framework capable of aligning multiple anatomical regions simultaneously using a standard UNet architecture and a novel field decomposition strategy. The model is lightweight, requiring only 11GB of GPU memory for training. To train and evaluate our method, we constructed a large-scale longitudinal dataset comprising 695 whole-body (thorax-abdomen-pelvic) paired CT scans from individual patients acquired at different time points. We benchmarked TotalRegistrator against a generic classical iterative algorithm and a recent foundation model for image registration. To further assess robustness and generalizability, we evaluated our model on three external datasets: the public thoracic and abdominal datasets from the Learn2Reg challenge, and a private multiphase abdominal dataset from a collaborating hospital. Experimental results on the in-house dataset show that the proposed approach generally surpasses baseline methods in multi-organ abdominal registration, with a slight drop in lung alignment performance. On out-of-distribution datasets, it achieved competitive results compared to leading single-organ models, despite not being fine-tuned for those tasks, demonstrating strong generalizability. The source code will be publicly available at: https://github.com/DIAGNijmegen/oncology_image_registration.git.
title TotalRegistrator: Towards a Lightweight Foundation Model for CT Image Registration
topic Image and Video Processing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.04450