Progressive Reliability-Guided Refinement with Cross-Modal Consistency and Iterative Correction for Weakly Supervised PET/CT Tumor Segmentation
- 1School of Software, Shenyang University of Technology, Shenyang 110870, China
- 2Software College, Northeastern University, Shenyang 110819, China
Abstract
Accurate tumor segmentation in Positron Emission Tomography/Computed Tomography (PET/CT) is essential for clinical diagnosis, staging, and treatment planning. However, the reliance on time-consuming pixel-level annotations limits the practical deployment of deep learning-based segmentation methods. Weakly supervised segmentation using image-level classification labels provides a promising alternative, yet generating accurate supervision remains highly challenging. Specifically, PET provides critical metabolic information but is susceptible to physiological hypermetabolism and background noise, while CT offers anatomical structure but suffers from low soft-tissue contrast and boundary uncertainty. Consequently, static or single-modality pseudo-labels derived from Class Activation Maps (CAMs) often suffer from boundary ambiguity and false-positive activations, failing to reliably guide segmentation. To address the core problem of dynamically evaluating and correcting supervision reliability, we propose a Progressive Reliability-Guided Refinement (PRGR) framework. Specifically, pixel-level supervision reliability is modeled by integrating PET metabolic consistency and CT structural gradients, leveraging complementary cross-modal information without introducing additional trainable parameters. Based on the estimated reliability, the image domain is adaptively partitioned into regions with different supervision credibility levels, enabling progressive optimization with differentiated learning constraints. Furthermore, pseudo-labels are iteratively updated using reliability estimates and segmentation predictions to progressively correct erroneous supervision. Experiments on two public PET/CT datasets demonstrate that our method effectively suppresses false positives and improves boundary delineation. Quantitatively, the proposed PRGR achieves Dice scores of 0.673 and 0.596 on the STS and HeadNeck datasets, respectively, yielding absolute improvements of 1.8% and 8.4% over the second-best baseline. Furthermore, it achieves superior Precision scores of 0.776 and 0.684, demonstrating its effectiveness in narrowing the performance gap between weakly and fully supervised segmentation while reducing the dependence on pixel-level annotations.
Related
- Goal-Oriented Ordered Autoencoding for Successive Refinement in Progressive Communication
- BTSR-Net: Boundary–topology guided structural refinement for high-resolution remote sensing image segmentation
- Reliability-Aware Refinement of High-Resolution Human Mobility Origin-Destination Tables under Sparsity and Suppression
- Automated correction of training sets for biological image segmentation by convolutional networks
- Integrated workflow for univariate and multivariate evaluation of batch correction reliability
BibTeX
@article{diao2026progressive,
title = {{Progressive Reliability-Guided Refinement with Cross-Modal Consistency and Iterative Correction for Weakly Supervised PET/CT Tumor Segmentation}},
author = {Zhaoshuo Diao and Huiyan Jiang},
journal = {Mathematics},
year = {2026},
month = oct,
doi = {10.3390/math14213903},
url = {https://www.mdpi.com/2227-7390/14/21/3903},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Progressive Reliability-Guided Refinement with Cross-Modal Consistency and Iterative Correction for Weakly Supervised PET/CT Tumor Segmentation
AU - Zhaoshuo Diao
AU - Huiyan Jiang
PY - 2026
DA - 2026/10/26
JO - Mathematics
DO - 10.3390/math14213903
UR - https://www.mdpi.com/2227-7390/14/21/3903
AB - Accurate tumor segmentation in Positron Emission Tomography/Computed Tomography (PET/CT) is essential for clinical diagnosis, staging, and treatment planning. However, the reliance on time-consuming pixel-level annotations limits the practical deployment of deep learning-based segmentation methods. Weakly supervised segmentation using image-level classification labels provides a promising alternative, yet generating accurate supervision remains highly challenging. Specifically, PET provides critical metabolic information but is susceptible to physiological hypermetabolism and background noise, while CT offers anatomical structure but suffers from low soft-tissue contrast and boundary uncertainty. Consequently, static or single-modality pseudo-labels derived from Class Activation Maps (CAMs) often suffer from boundary ambiguity and false-positive activations, failing to reliably guide segmentation. To address the core problem of dynamically evaluating and correcting supervision reliability, we propose a Progressive Reliability-Guided Refinement (PRGR) framework. Specifically, pixel-level supervision reliability is modeled by integrating PET metabolic consistency and CT structural gradients, leveraging complementary cross-modal information without introducing additional trainable parameters. Based on the estimated reliability, the image domain is adaptively partitioned into regions with different supervision credibility levels, enabling progressive optimization with differentiated learning constraints. Furthermore, pseudo-labels are iteratively updated using reliability estimates and segmentation predictions to progressively correct erroneous supervision. Experiments on two public PET/CT datasets demonstrate that our method effectively suppresses false positives and improves boundary delineation. Quantitatively, the proposed PRGR achieves Dice scores of 0.673 and 0.596 on the STS and HeadNeck datasets, respectively, yielding absolute improvements of 1.8% and 8.4% over the second-best baseline. Furthermore, it achieves superior Precision scores of 0.776 and 0.684, demonstrating its effectiveness in narrowing the performance gap between weakly and fully supervised segmentation while reducing the dependence on pixel-level annotations.
SN - 2227-7390
ER -