Artificial intelligence efficiently predicts gastric lesions, Helicobacter pylori infection and lymph node metastasis upon endoscopic images.

IF 7 2区医学 Q1 ONCOLOGY

Chinese Journal of Cancer Research Pub Date : 2024-10-30 DOI:10.21147/j.issn.1000-9604.2024.05.03

Ruixin Yang, Jialin Zhang, Fengsheng Zhan, Chao Yan, Sheng Lu, Zhenggang Zhu, Kang An, Jing Sun, Yingyan Yu

{"title":"Artificial intelligence efficiently predicts gastric lesions, Helicobacter pylori infection and lymph node metastasis upon endoscopic images.","authors":"Ruixin Yang, Jialin Zhang, Fengsheng Zhan, Chao Yan, Sheng Lu, Zhenggang Zhu, Kang An, Jing Sun, Yingyan Yu","doi":"10.21147/j.issn.1000-9604.2024.05.03","DOIUrl":null,"url":null,"abstract":"Objective: Medical images have been increased rapidly in digital medicine era, presenting an opportunity for the intervention of artificial intelligence (AI). In order to explore the value of convolutional neural network (CNN) algorithms in endoscopic images, we developed an AI-assisted comprehensive analysis system for endoscopic images and explored its performance in clinical real scenarios.Methods: A total of 6,270 white light endoscopic images from 516 cases were used to train 14 different CNN models. The images were divided into training set, validation set and test set according to 7:1:2 for exploring the possibility of discrimination of gastric cancer (GC) and benign lesions (nGC), gastric ulcer (GU) and ulcerated cancer (UCa), early gastric cancer (EGC) and nGC, infection of Helicobacter pylori (Hp) and no infection of Hp (noHp), as well as metastasis and no-metastasis at perigastric lymph nodes.Results: Among the 14 CNN models, EfficientNetB7 revealed the best performance on two-category of GC and nGC [accuracy: 96.40% and area under the curve (AUC)=0.9959], GU and UCa (accuracy: 90.84% and AUC=0.8155), EGC and nGC (accuracy: 97.88% and AUC=0.9943), and Hp and noHp (accuracy: 83.33% and AUC=0.9096). Whereas, InceptionV3 model showed better performance on predicting metastasis and no-metastasis of perigastric lymph nodes for EGC (accuracy: 79.44% and AUC=0.7181). In addition, the integrated analysis of endoscopic images and gross images of gastrectomy specimens was performed on 95 cases by EfficientNetB7 and RFB-SSD object detection model, resulting in 100% of predictive accuracy in EGC.Conclusions: Taken together, this study integrated image sources from endoscopic examination and gastrectomy of gastric tumors and incorporated the advantages of different CNN models. The AI-assisted diagnostic system will play an important role in the therapeutic decision-making of EGC.","PeriodicalId":9882,"journal":{"name":"Chinese Journal of Cancer Research","volume":"36 5","pages":"489-502"},"PeriodicalIF":7.0000,"publicationDate":"2024-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11555197/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Chinese Journal of Cancer Research","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.21147/j.issn.1000-9604.2024.05.03","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ONCOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Objective: Medical images have been increased rapidly in digital medicine era, presenting an opportunity for the intervention of artificial intelligence (AI). In order to explore the value of convolutional neural network (CNN) algorithms in endoscopic images, we developed an AI-assisted comprehensive analysis system for endoscopic images and explored its performance in clinical real scenarios.

Methods: A total of 6,270 white light endoscopic images from 516 cases were used to train 14 different CNN models. The images were divided into training set, validation set and test set according to 7:1:2 for exploring the possibility of discrimination of gastric cancer (GC) and benign lesions (nGC), gastric ulcer (GU) and ulcerated cancer (UCa), early gastric cancer (EGC) and nGC, infection of Helicobacter pylori (Hp) and no infection of Hp (noHp), as well as metastasis and no-metastasis at perigastric lymph nodes.

Results: Among the 14 CNN models, EfficientNetB7 revealed the best performance on two-category of GC and nGC [accuracy: 96.40% and area under the curve (AUC)=0.9959], GU and UCa (accuracy: 90.84% and AUC=0.8155), EGC and nGC (accuracy: 97.88% and AUC=0.9943), and Hp and noHp (accuracy: 83.33% and AUC=0.9096). Whereas, InceptionV3 model showed better performance on predicting metastasis and no-metastasis of perigastric lymph nodes for EGC (accuracy: 79.44% and AUC=0.7181). In addition, the integrated analysis of endoscopic images and gross images of gastrectomy specimens was performed on 95 cases by EfficientNetB7 and RFB-SSD object detection model, resulting in 100% of predictive accuracy in EGC.

Conclusions: Taken together, this study integrated image sources from endoscopic examination and gastrectomy of gastric tumors and incorporated the advantages of different CNN models. The AI-assisted diagnostic system will play an important role in the therapeutic decision-making of EGC.

查看原文本刊更多论文

人工智能通过内窥镜图像有效预测胃部病变、幽门螺旋杆菌感染和淋巴结转移。

目的：数字医学时代，医学图像迅速增加，为人工智能（AI）的介入提供了契机。为了探索卷积神经网络（CNN）算法在内窥镜图像中的应用价值，我们开发了一套人工智能辅助的内窥镜图像综合分析系统，并探索了其在临床真实场景中的表现：共使用了 516 个病例的 6270 张白光内窥镜图像来训练 14 个不同的 CNN 模型。方法：采用 516 个病例的 6270 张白光内窥镜图像训练 14 个不同的 CNN 模型，按照 7:1:2 的比例将图像分为训练集、验证集和测试集，以探索区分胃癌（GC）和良性病变（nGC）、胃溃疡（GU）和溃疡癌（UCa）、早期胃癌（EGC）和 nGC、幽门螺杆菌感染（Hp）和未感染幽门螺杆菌（noHp）以及胃周淋巴结转移和未转移的可能性：在 14 个 CNN 模型中，EfficientNetB7 在 GC 和 nGC 两类[准确率：96.40%，曲线下面积（AUC）=0.9959]、GU 和 UCa（准确率：90.84%，AUC=0.8155）、EGC 和 nGC（准确率：97.88%，AUC=0.9943）以及 Hp 和 noHp（准确率：83.33%，AUC=0.9096）方面表现最佳。而 InceptionV3 模型在预测 EGC 胃周淋巴结转移和无转移方面表现更佳（准确率：79.44%，AUC=0.7181）。此外，通过EfficientNetB7和RFB-SSD对象检测模型对95例胃切除术标本的内镜图像和大体图像进行了综合分析，结果对EGC的预测准确率为100%：综上所述，本研究整合了胃肿瘤内窥镜检查和胃切除术的图像源，并融合了不同 CNN 模型的优势。人工智能辅助诊断系统将在 EGC 的治疗决策中发挥重要作用。

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

Chinese Journal of Cancer Research 医学-肿瘤学

自引率

9.80%

发文量

1726

审稿时长

4.5 months

期刊介绍： Chinese Journal of Cancer Research (CJCR; Print ISSN: 1000-9604; Online ISSN:1993-0631) is published by AME Publishing Company in association with Chinese Anti-Cancer Association.It was launched in March 1995 as a quarterly publication and is now published bi-monthly since February 2013. CJCR is published bi-monthly in English, and is an international journal devoted to the life sciences and medical sciences. It publishes peer-reviewed original articles of basic investigations and clinical observations, reviews and brief communications providing a forum for the recent experimental and clinical advances in cancer research. This journal is indexed in Science Citation Index Expanded (SCIE), PubMed/PubMed Central (PMC), Scopus, SciSearch, Chemistry Abstracts (CA), the Excerpta Medica/EMBASE, Chinainfo, CNKI, CSCI, etc.