Abstract
Accurate segmentation of the left atrium (LA) structure using cardiac computed tomography (CT) is critical to assessing atrial fibrillation and planning catheter ablation. However, automating this segmentation is difficult due to severe class imbalance, significant anatomical variance, and the limited availability of manually annotated data. In this work, we propose a quality-controlled semi-supervised framework based on UU-Mamba for multi-structure cardiac CT segmentation. To improve supervised learning, we use a logit-adjusted cross-entropy loss to address class imbalance among the target structures. A semi-automatic quality-control framework evaluates pseudo labels based on their morphological characteristics, anatomical consistency, and prediction stability, followed by a manual review to select reliable pseudo labels to refine the model. Experiments on the CARE 2026 dataset demonstrate that the proposed framework consistently improves segmentation performance over the baseline UU-Mamba model. These results show that quality-controlled pseudo-label fine-tuning successfully enhances multi-structure cardiac CT segmentation while reducing the need for extensive manual annotation.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/CARE_028.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=ZbDEXgxrZi
BibTex
@InProceedings{CheYin_Progressive_MICCAISAT2026,
author = { Chen, Yineng AND Gui, Liangqiao AND Liu, Shucheng AND Hu, Mingcheng AND Huang, Peng AND Guo, Hui AND Hu, Shu AND Li, Tengfei AND Lin, Shao AND Zhu, Hongtu AND Wang, Xin},
title = { { Progressive Optimization of UU-Mamba for Left Atrial Multi-Structure Segmentation in CT } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17273},
month = {pending},
page = {pending}
}