Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images

Abstract Early detection and precise preoperative staging of early gastric cancer (EGC) are critical. Therefore, this study aims to develop a deep learning model using portal venous phase CT images to accurately distinguish EGC without lymph node metastasis. This study included 3164 patients with ga...

Full description

Saved in:
Bibliographic Details
Main Authors: Jingyang He, Jingli Xu, Wujie Chen, Mengxuan Cao, Jiaqing Zhang, Qing Yang, Enze Li, Ruolan Zhang, Yahang Tong, Yanqiang Zhang, Chen Gao, Qianyu Zhao, Zhiyuan Xu, Lijing Wang, Xiangdong Cheng, Guoliang Zheng, Siwei Pan, Can Hu
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Precision Oncology
Online Access:https://doi.org/10.1038/s41698-025-01055-9
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849388169437380608
author Jingyang He
Jingli Xu
Wujie Chen
Mengxuan Cao
Jiaqing Zhang
Qing Yang
Enze Li
Ruolan Zhang
Yahang Tong
Yanqiang Zhang
Chen Gao
Qianyu Zhao
Zhiyuan Xu
Lijing Wang
Xiangdong Cheng
Guoliang Zheng
Siwei Pan
Can Hu
author_facet Jingyang He
Jingli Xu
Wujie Chen
Mengxuan Cao
Jiaqing Zhang
Qing Yang
Enze Li
Ruolan Zhang
Yahang Tong
Yanqiang Zhang
Chen Gao
Qianyu Zhao
Zhiyuan Xu
Lijing Wang
Xiangdong Cheng
Guoliang Zheng
Siwei Pan
Can Hu
author_sort Jingyang He
collection DOAJ
description Abstract Early detection and precise preoperative staging of early gastric cancer (EGC) are critical. Therefore, this study aims to develop a deep learning model using portal venous phase CT images to accurately distinguish EGC without lymph node metastasis. This study included 3164 patients with gastric cancer (GC) who underwent radical surgery at two medical centers in China from 2006 to 2019. Moreover, 2.5D radiomic data and multi-instance learning (MIL) were novel approaches applied in this study. By basing the selection of features on 2.5D radiomic data and MIL, the ResNet101 model combined with the XGBoost model represented a satisfactory performance for diagnosing pT1N0 GC. Furthermore, the 2.5D MIL-based model demonstrated a markedly superior predictive performance compared to traditional radiomics models and clinical models. We first constructed a deep learning prediction model based on 2.5D radiomics and MIL for effectively diagnosing pT1N0 GC patients, which provides valuable information for the individualized treatment selection.
format Article
id doaj-art-ab979a75715243fdbaea0f5b99673d2f
institution Kabale University
issn 2397-768X
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Precision Oncology
spelling doaj-art-ab979a75715243fdbaea0f5b99673d2f2025-08-20T03:42:23ZengNature Portfolionpj Precision Oncology2397-768X2025-07-019111110.1038/s41698-025-01055-9Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT imagesJingyang He0Jingli Xu1Wujie Chen2Mengxuan Cao3Jiaqing Zhang4Qing Yang5Enze Li6Ruolan Zhang7Yahang Tong8Yanqiang Zhang9Chen Gao10Qianyu Zhao11Zhiyuan Xu12Lijing Wang13Xiangdong Cheng14Guoliang Zheng15Siwei Pan16Can Hu17Department of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Radiology, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Radiology, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Radiology, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesZhejiang Hospital of Traditional Chinese MedicineDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Cancer Hospital of China Medical University, Cancer Hospital of Dalian University of Technology, Liaoning Cancer Hospital & InstituteDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesDepartment of Gastric Surgery, Zhejiang Cancer Hospital, Hangzhou Institute of Medicine (HIM), Chinese Academy of SciencesAbstract Early detection and precise preoperative staging of early gastric cancer (EGC) are critical. Therefore, this study aims to develop a deep learning model using portal venous phase CT images to accurately distinguish EGC without lymph node metastasis. This study included 3164 patients with gastric cancer (GC) who underwent radical surgery at two medical centers in China from 2006 to 2019. Moreover, 2.5D radiomic data and multi-instance learning (MIL) were novel approaches applied in this study. By basing the selection of features on 2.5D radiomic data and MIL, the ResNet101 model combined with the XGBoost model represented a satisfactory performance for diagnosing pT1N0 GC. Furthermore, the 2.5D MIL-based model demonstrated a markedly superior predictive performance compared to traditional radiomics models and clinical models. We first constructed a deep learning prediction model based on 2.5D radiomics and MIL for effectively diagnosing pT1N0 GC patients, which provides valuable information for the individualized treatment selection.https://doi.org/10.1038/s41698-025-01055-9
spellingShingle Jingyang He
Jingli Xu
Wujie Chen
Mengxuan Cao
Jiaqing Zhang
Qing Yang
Enze Li
Ruolan Zhang
Yahang Tong
Yanqiang Zhang
Chen Gao
Qianyu Zhao
Zhiyuan Xu
Lijing Wang
Xiangdong Cheng
Guoliang Zheng
Siwei Pan
Can Hu
Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
npj Precision Oncology
title Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
title_full Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
title_fullStr Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
title_full_unstemmed Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
title_short Development of a deep learning model for T1N0 gastric cancer diagnosis using 2.5D radiomic data in preoperative CT images
title_sort development of a deep learning model for t1n0 gastric cancer diagnosis using 2 5d radiomic data in preoperative ct images
url https://doi.org/10.1038/s41698-025-01055-9
work_keys_str_mv AT jingyanghe developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT jinglixu developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT wujiechen developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT mengxuancao developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT jiaqingzhang developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT qingyang developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT enzeli developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT ruolanzhang developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT yahangtong developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT yanqiangzhang developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT chengao developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT qianyuzhao developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT zhiyuanxu developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT lijingwang developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT xiangdongcheng developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT guoliangzheng developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT siweipan developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages
AT canhu developmentofadeeplearningmodelfort1n0gastriccancerdiagnosisusing25dradiomicdatainpreoperativectimages