Robust liver segmentation method improves MRI accuracy across scanners
CIRSeg: Coarse-to-Fine Intensity-Robust Liver Segmentation with Source-Free Continual Test-Time Adaptation
Computer Vision and Pattern Recognition
Summary
Liver segmentation in MRI scans helps doctors plan treatments and monitor diseases, but differences between scanners can make this task hard. The authors created CIRSeg, a two-step method that first finds the liver roughly and then refines the details, making it more reliable across different MRI machines. It uses clever tricks to handle varying image intensities and adjusts itself while being used to get better results. This approach helps produce more accurate liver boundaries without extra false spots.
What this means in practice
- •For medical imaging teams: Improve liver MRI segmentation accuracy across different scanners and acquisition settings for better clinical workflows.
- •For clinical software developers: Integrate adaptive segmentation techniques to enhance robustness of liver analysis tools without requiring retraining on new data.
Authors
Ruoshi Xu, Mingqi Gao, Shengda Luo, Jingkun Chen
Abstract
Reliable liver segmentation in contrast-enhanced MRI is essential for quantitative hepatic assessment, treatment planning, and longitudinal disease monitoring. However, limited annotated data and scanner- or vendor-dependent intensity variations can cause overfitting and poor generalization to unseen acquisition domains. Moreover, simultaneously achieving robust global localization and precise boundary delineation remains challenging, while predictions may contain isolated false-positive regions outside the main liver component. To address these challenges, we propose CIRSeg, a coarse-to-fine, intensity-robust liver segmentation framework based on nnU-Netv2. CIRSeg combines 3D CutMix with stochastic intensity transfer using either Nyul augmentation or histogram matching to improve robustness to heterogeneous MRI intensities. Its cascaded architecture decouples low-resolution anatomical localization from full-resolution boundary refinement. At inference, source-free test-time adaptation based on confidence-filtered predictions and probability-prior regularization further improves robustness to out-of-distribution inputs. As a final deterministic post-processing step, largest connected component filtering removes isolated false-positive regions. On the CARE 2026 test set, CIRSeg achieves Dice scores of 97.13\% and 97.93\% on the in-domain and unseen-domain subsets, with corresponding HD95 values of 20.18 mm and 11.30 mm, respectively. These results demonstrate consistently accurate segmentation across both in-domain and unseen acquisition settings. The code is available at https://github.com/jingkunchen/MICCAI_CARE_2026