A novel deep learning segmentation and multimodal radiomics approach for radiation proctitis prediction in cervical cancer
- 1Department of Oncology, The First Affiliated Hospital of Yangtze University, Jingzhou, Hubei China
- 2Hubei Provincial Clinical Research Center for Personalized Cancer Diagnosis and Therapy, Jingzhou, China
- 3Hubei Province Key Laboratory of Precision Radiation Oncology, Wuhan, 430022 China
Abstract
Purpose: This study aimed to develop and validate a novel interpretable framework integrating a deep segmentation network with multimodal radiomics to improve the prediction performance of radiation proctitis (RP) in cervical cancer patients undergoing radiotherapy.
Methods: This retrospective study included 650 cervical cancer patients, who were divided into training and internal device test sets, and independent device test set, based on different computed tomography (CT) scanners. A Mamba-enhanced deep learning network was constructed to enhance global modeling capability in pelvic anatomical structures, providing reliable regions of interest segmentation for radiomics and adaptive radiotherapy. In multimodal radiomics approach, foundation model features were integrated with radiomics and clinical data to mitigate inherent bias in training data, thereby enhancing the prediction of RP. Shapley Additive exPlanations (SHAP) technology was employed to explain the contributions of multimodal features to the prediction performance.
Results: The segmentation model achieved dice similarity coefficients (DSC) of 90.76% and 85.33% on the internal and independent device test sets, outperforming existing models. Ablation study confirmed the contribution of each module to segmentation performance. The multimodal radiomics model demonstrated robust performance in predicting RP, with area under the curve (AUC) values of 0.882 and 0.865 on the internal and independent device test sets, surpassing single-feature models. Additionally, SHAP analysis revealed the interactions between key features, including dosimetric and foundation model features, which contributed to improve decision interpretability.
Conclusion: This study significantly improved the robustness and accuracy of RP prediction by integrating the deep segmentation model and multimodal radiomics, providing a promising clinical applications framework for individualized radiotherapy planning.
Clinical trial number: Not applicable
Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s13014-026-02905-x.
Related
- Comparative Evaluation of Feature Selection Algorithms for Predicting Radiation-Induced Complications in Head and Neck Cancer Using Multimodal CT and MRI Radiomics
- Multiparametric MRI-based habitat and peritumoral radiomics integrating semi-automated segmentation for preoperative prediction of high-grade prostate cancer: a dual-center study
- Multimodal preoperative risk stratification of pancreatic mass lesions using clinical variables, imaging radiomics and plasma cfDNA methylation: protocol for an ambispective multicentre diagnostic prediction model study
- Combining handcrafted radiomics features and deep learning features within different regions of interest in planning CT images for the prediction of symptomatic radiation pneumonitis in locally advanced non-small cell lung cancer
- Integrative transcriptomic and computational analysis associates Ms4a6c+ macrophages with fibrosis in radiation proctitis and computationally nominates MOTS-c as a candidate modulator
BibTeX
@article{he2026novel,
title = {{A novel deep learning segmentation and multimodal radiomics approach for radiation proctitis prediction in cervical cancer}},
author = {Chuchu He and Zhichao Wang and Zhen Liu and Xiaoyan Su and Yan Hu and Jun Cai},
journal = {Radiation Oncology (London, England)},
year = {2026},
month = aug,
doi = {10.1186/s13014-026-02905-x},
url = {https://doi.org/10.1186/s13014-026-02905-x},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - A novel deep learning segmentation and multimodal radiomics approach for radiation proctitis prediction in cervical cancer
AU - Chuchu He
AU - Zhichao Wang
AU - Zhen Liu
AU - Xiaoyan Su
AU - Yan Hu
AU - Jun Cai
PY - 2026
DA - 2026/08/12
JO - Radiation Oncology (London, England)
DO - 10.1186/s13014-026-02905-x
UR - https://doi.org/10.1186/s13014-026-02905-x
AB - Purpose: This study aimed to develop and validate a novel interpretable framework integrating a deep segmentation network with multimodal radiomics to improve the prediction performance of radiation proctitis (RP) in cervical cancer patients undergoing radiotherapy. Methods: This retrospective study included 650 cervical cancer patients, who were divided into training and internal device test sets, and independent device test set, based on different computed tomography (CT) scanners. A Mamba-enhanced deep learning network was constructed to enhance global modeling capability in pelvic anatomical structures, providing reliable regions of interest segmentation for radiomics and adaptive radiotherapy. In multimodal radiomics approach, foundation model features were integrated with radiomics and clinical data to mitigate inherent bias in training data, thereby enhancing the prediction of RP. Shapley Additive exPlanations (SHAP) technology was employed to explain the contributions of multimodal features to the prediction performance. Results: The segmentation model achieved dice similarity coefficients (DSC) of 90.76% and 85.33% on the internal and independent device test sets, outperforming existing models. Ablation study confirmed the contribution of each module to segmentation performance. The multimodal radiomics model demonstrated robust performance in predicting RP, with area under the curve (AUC) values of 0.882 and 0.865 on the internal and independent device test sets, surpassing single-feature models. Additionally, SHAP analysis revealed the interactions between key features, including dosimetric and foundation model features, which contributed to improve decision interpretability. Conclusion: This study significantly improved the robustness and accuracy of RP prediction by integrating the deep segmentation model and multimodal radiomics, providing a promising clinical applications framework for individualized radiotherapy planning. Clinical trial number: Not applicable Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s13014-026-02905-x.
SN - 1748-717X
ER -