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

IF 3.2 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Medical physics Pub Date : 2024-11-13 DOI:10.1002/mp.17493

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

{"title":"Radiomics-guided generative adversarial network for automatic primary target volume segmentation for nasopharyngeal carcinoma using computed tomography images","authors":"Juebin Jin, Jicheng Zhang, Xianwen Yu, Ziqing Xiang, Xuanxuan Zhu, Mingrou Guo, Zeshuo Zhao, WenLong Li, Heng Li, Jiayi Xu, Xiance Jin","doi":"10.1002/mp.17493","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background</h3>\n \n <p>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.</p>\n </section>\n \n <section>\n \n <h3> Purpose</h3>\n \n <p>A peritumoral radiomics-guided generative adversarial network (PRG-GAN) was proposed to address this challenge.</p>\n </section>\n \n <section>\n \n <h3> Methods</h3>\n \n <p>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 <i>t</i>-test with Bonferroni correction and Cohen's d (<i>d</i>) effect sizes. A two-sided <i>p</i>-value of less than 0.05 was considered statistically significant.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>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 (<i>p</i> = 0.001, <i>d</i> = 0.71), 4.15 ± 7.56 mm (<i>p</i> = 0.002, <i>d</i> = 0.67), and 1.11 ± 1.65 mm (<i>p</i> < 0.001, <i>d</i> = 0.46) of PRG-GAN, respectively.</p>\n </section>\n \n <section>\n \n <h3> Conclusion</h3>\n \n <p>Integrating radiomics features into GAN is promising to solve unclear border limitations and increase the delineation accuracy of GTVp for patients with NPC.</p>\n </section>\n </div>","PeriodicalId":18384,"journal":{"name":"Medical physics","volume":"52 2","pages":"1119-1132"},"PeriodicalIF":3.2000,"publicationDate":"2024-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Medical physics","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/mp.17493","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}

引用次数: 0

Abstract

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.

查看原文本刊更多论文

使用计算机断层扫描图像对鼻咽癌自动进行原发靶体积分割的放射线组学引导生成对抗网络。

背景：鼻咽癌（NPC）的原发肿瘤体积（GTVp）自动分割是一项颇具挑战性的任务，因为肿瘤与其周围环境之间存在相似的视觉特征，尤其是在对比度分辨率极低的计算机断层扫描（CT）图像上。因此，最近提出的大多数基于放射组学或深度学习（DL）的方法很难在 CT 数据集上取得良好效果：方法：共收集了 157 位鼻咽癌患者的 CT 图像，并将其分为训练组、验证组和测试组，每组分别有 108 位、9 位和 30 位患者。提出的模型基于标准的 GAN，由生成器网络和判别器网络组成。首先对 GAN 的初始分割结果进行形态扩张，以划定环状瘤周区域，在此过程中提取放射组学特征作为先验指导知识。然后，通过判别器的全连接层将放射组学特征与语义特征融合，实现体素级分类和分割。采用带 Bonferroni 校正的配对样本 t 检验和 Cohen's d（d）效应大小，使用骰子相似系数（DSC）、95% Hausdorff 距离（HD95）和平均对称面距离（ASSD）来评估分割性能。双侧 p 值小于 0.05 即为具有统计学意义：结果：模型生成的预测结果与地面实况的重叠率很高。平均 DSC、HD95 和 ASSD 从 GAN 的 0.80 ± 0.12、4.65 ± 4.71 mm 和 1.35 ± 1.15 mm 显著提高到 0.85 ± 0.18 (p = 0.001, d = 0.71)、4.15 ± 7.56 mm (p = 0.002, d = 0.67) 和 1.11 ± 1.65 mm (p 结论：将放射组学特征整合到 GAN 中，可显著提高 DSC、HD95 和 ASSD 的平均值：将放射组学特征整合到 GAN 中有望解决边界不清的限制，并提高鼻咽癌患者 GTVp 划分的准确性。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Medical physics 医学-核医学

CiteScore

6.80

自引率

15.80%

发文量

660

审稿时长

1.7 months

期刊介绍： Medical Physics publishes original, high impact physics, imaging science, and engineering research that advances patient diagnosis and therapy through contributions in 1) Basic science developments with high potential for clinical translation 2) Clinical applications of cutting edge engineering and physics innovations 3) Broadly applicable and innovative clinical physics developments Medical Physics is a journal of global scope and reach. By publishing in Medical Physics your research will reach an international, multidisciplinary audience including practicing medical physicists as well as physics- and engineering based translational scientists. We work closely with authors of promising articles to improve their quality.