EUVIMEDOpdag EUVIMED
Crossref

Radiomics‐guided generative adversarial network for automatic primary target volume segmentation for nasopharyngeal carcinoma using computed tomography images

Juebin Jin, Jicheng Zhang, Xianwen Yu, Ziqing Xiang, Xuanxuan Zhu, Mingrou Guo, Zeshuo Zhao, WenLong Li, Heng Li, Jiayi Xu, Xiance Jin

Medical Physics · 2025-02 · 52 (2) · 1119-1132

DOI: 10.1002/mp.17493 ↗

At a glance

Journal article in Medical Physics, published online on 2024-11-13. Written by Juebin Jin and 10 co-authors. Institution: Department of Radiotherapy Center, 1st Affiliated Hospital of Wenzhou Medical University, Wenzhou, China. Major MEDLINE topics: Nasopharyngeal Carcinoma, Tomography, X-Ray Computed, Image Processing, Computer-Assisted, Nasopharyngeal Neoplasms. Cited 3 times so far (Crossref); 64 references. Funded by: National Natural Science Foundation, Key project of Zhejiang Natural Science Foundation, Key project of Zhejiang Provincial Health Science and Technology Program.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Background

Automatic primary gross tumor volume (GTVp) segmentation for nasopharyngeal carcinoma (NPC) is a quite challenging task because of the existence of similar visual characteristics between tumors and their surroundings, especially on computed tomography (CT) images with severe low contrast resolution. Therefore, most recently proposed methods based on radiomics or deep learning (DL) is difficult to achieve good results on CT datasets.

Purpose

A peritumoral radiomics‐guided generative adversarial network (PRG‐GAN) was proposed to address this challenge.

Methods

A total of 157 NPC patients with CT images was collected and divided into training, validation, and testing cohorts of 108, 9, and 30 patients, respectively. The proposed model was based on a standard GAN consisting of a generator network and a discriminator network. Morphological dilation on the initial segmentation results from GAN was first conducted to delineate annular peritumoral region, in which radiomics features were extracted as priori guide knowledge. Then, radiomics features were fused with semantic features by the discriminator's fully connected layer to achieve the voxel‐level classification and segmentation. The dice similarity coefficient (DSC), 95% Hausdorff distance (HD95), and average symmetric surface distance (ASSD) were used to evaluate the segmentation performance using a paired samples t ‐test with Bonferroni correction and Cohen's d ( d ) effect sizes. A two‐sided p ‐value of less than 0.05 was considered statistically significant.

Results

The model‐generated predictions had a high overlap ratio with the ground truth. The average DSC, HD95, and ASSD were significantly improved from 0.80 ± 0.12, 4.65 ± 4.71 mm, and 1.35 ± 1.15 mm of GAN to 0.85 ± 0.18 ( p = 0.001, d = 0.71), 4.15 ± 7.56 mm ( p = 0.002, d = 0.67), and 1.11 ± 1.65 mm ( p < 0.001, d = 0.46) of PRG‐GAN, respectively.

Conclusion

Integrating radiomics features into GAN is promising to solve unclear border limitations and increase the delineation accuracy of GTVp for patients with NPC.

Subjects

MeSH terms (MEDLINE)

Nasopharyngeal CarcinomaTomography, X-Ray ComputedImage Processing, Computer-AssistedNasopharyngeal NeoplasmsHumansAutomationTumor BurdenDeep LearningGenerative Adversarial NetworksRadiomics

Author keywords

adversarial learningcomputed tomography (CT)nasopharyngeal carcinoma segmentationradiomics

Authors and institutions

  • Department of Radiotherapy Center, 1st Affiliated Hospital of Wenzhou Medical University, Wenzhou, China.
  • Department of Medical Engineering, 2nd Affiliated Hospital of Wenzhou Medical University, Wenzhou, China.
  • School of Basic Medical Science, Wenzhou Medical University, Wenzhou, China.

Funding

  • National Natural Science Foundation
  • Key project of Zhejiang Natural Science Foundation
  • Key project of Zhejiang Provincial Health Science and Technology Program
  • Major project of Wenzhou Science and Technology Bureau
  • Wenzhou Science and Technology Bureau
  • Wenzhou Municipal Science and Technology Bureau

Cite

APA

Jin, J., Zhang, J., Yu, X., Xiang, Z., Zhu, X., Guo, M., Zhao, Z., Li, W., Li, H., Xu, J., & Jin, X. (2025). Radiomics‐guided generative adversarial network for automatic primary target volume segmentation for nasopharyngeal carcinoma using computed tomography images. Medical Physics, 52(2), 1119-1132. https://doi.org/10.1002/mp.17493

Vancouver

Jin J, Zhang J, Yu X, Xiang Z, Zhu X, Guo M, et al. Radiomics‐guided generative adversarial network for automatic primary target volume segmentation for nasopharyngeal carcinoma using computed tomography images. Medical Physics. 2025;52(2):1119-1132. doi:10.1002/mp.17493

BibTeX
@article{Jin2025,
  author = {Jin, Juebin and Zhang, Jicheng and Yu, Xianwen and Xiang, Ziqing and Zhu, Xuanxuan and Guo, Mingrou and Zhao, Zeshuo and Li, WenLong and Li, Heng and Xu, Jiayi and Jin, Xiance},
  title = {Radiomics‐guided generative adversarial network for automatic primary target volume segmentation for nasopharyngeal carcinoma using computed tomography images},
  journal = {Medical Physics},
  year = {2025},
  volume = {52},
  number = {2},
  pages = {1119--1132},
  doi = {10.1002/mp.17493}
}
Start for free →