A novel artificial intelligence-based endoscopic ultrasonography diagnostic system for diagnosing the invasion depth of early gastric cancer.

IF 6.9 2区医学 Q1 GASTROENTEROLOGY & HEPATOLOGY

Journal of Gastroenterology Pub Date : 2024-07-01 Epub Date: 2024-05-07 DOI:10.1007/s00535-024-02102-1

Ryotaro Uema, Yoshito Hayashi, Takashi Kizu, Takumi Igura, Hideharu Ogiyama, Takuya Yamada, Risato Takeda, Kengo Nagai, Takuya Inoue, Masashi Yamamoto, Shinjiro Yamaguchi, Takashi Kanesaka, Takeo Yoshihara, Minoru Kato, Shunsuke Yoshii, Yoshiki Tsujii, Shinichiro Shinzaki, Tetsuo Takehara

{"title":"A novel artificial intelligence-based endoscopic ultrasonography diagnostic system for diagnosing the invasion depth of early gastric cancer.","authors":"Ryotaro Uema, Yoshito Hayashi, Takashi Kizu, Takumi Igura, Hideharu Ogiyama, Takuya Yamada, Risato Takeda, Kengo Nagai, Takuya Inoue, Masashi Yamamoto, Shinjiro Yamaguchi, Takashi Kanesaka, Takeo Yoshihara, Minoru Kato, Shunsuke Yoshii, Yoshiki Tsujii, Shinichiro Shinzaki, Tetsuo Takehara","doi":"10.1007/s00535-024-02102-1","DOIUrl":null,"url":null,"abstract":"Background: We developed an artificial intelligence (AI)-based endoscopic ultrasonography (EUS) system for diagnosing the invasion depth of early gastric cancer (EGC), and we evaluated the performance of this system.Methods: A total of 8280 EUS images from 559 EGC cases were collected from 11 institutions. Within this dataset, 3451 images (285 cases) from one institution were used as a development dataset. The AI model consisted of segmentation and classification steps, followed by the CycleGAN method to bridge differences in EUS images captured by different equipment. AI model performance was evaluated using an internal validation dataset collected from the same institution as the development dataset (1726 images, 135 cases). External validation was conducted using images collected from the other 10 institutions (3103 images, 139 cases).Results: The area under the curve (AUC) of the AI model in the internal validation dataset was 0.870 (95% CI: 0.796-0.944). Regarding diagnostic performance, the accuracy/sensitivity/specificity values of the AI model, experts (n = 6), and nonexperts (n = 8) were 82.2/63.4/90.4%, 81.9/66.3/88.7%, and 68.3/60.9/71.5%, respectively. The AUC of the AI model in the external validation dataset was 0.815 (95% CI: 0.743-0.886). The accuracy/sensitivity/specificity values of the AI model (74.1/73.1/75.0%) and the real-time diagnoses of experts (75.5/79.1/72.2%) in the external validation dataset were comparable.Conclusions: Our AI model demonstrated a diagnostic performance equivalent to that of experts.","PeriodicalId":16059,"journal":{"name":"Journal of Gastroenterology","volume":" ","pages":"543-555"},"PeriodicalIF":6.9000,"publicationDate":"2024-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11217111/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Gastroenterology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s00535-024-02102-1","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/5/7 0:00:00","PubModel":"Epub","JCR":"Q1","JCRName":"GASTROENTEROLOGY & HEPATOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: We developed an artificial intelligence (AI)-based endoscopic ultrasonography (EUS) system for diagnosing the invasion depth of early gastric cancer (EGC), and we evaluated the performance of this system.

Methods: A total of 8280 EUS images from 559 EGC cases were collected from 11 institutions. Within this dataset, 3451 images (285 cases) from one institution were used as a development dataset. The AI model consisted of segmentation and classification steps, followed by the CycleGAN method to bridge differences in EUS images captured by different equipment. AI model performance was evaluated using an internal validation dataset collected from the same institution as the development dataset (1726 images, 135 cases). External validation was conducted using images collected from the other 10 institutions (3103 images, 139 cases).

Results: The area under the curve (AUC) of the AI model in the internal validation dataset was 0.870 (95% CI: 0.796-0.944). Regarding diagnostic performance, the accuracy/sensitivity/specificity values of the AI model, experts (n = 6), and nonexperts (n = 8) were 82.2/63.4/90.4%, 81.9/66.3/88.7%, and 68.3/60.9/71.5%, respectively. The AUC of the AI model in the external validation dataset was 0.815 (95% CI: 0.743-0.886). The accuracy/sensitivity/specificity values of the AI model (74.1/73.1/75.0%) and the real-time diagnoses of experts (75.5/79.1/72.2%) in the external validation dataset were comparable.

Conclusions: Our AI model demonstrated a diagnostic performance equivalent to that of experts.

Abstract Image

查看原文本刊更多论文

基于人工智能的新型内镜超声诊断系统，用于诊断早期胃癌的侵犯深度。

背景：我们开发了一种基于人工智能（AI）的内镜超声（EUS）系统，用于诊断早期胃癌（EGC）的侵袭深度，并对该系统的性能进行了评估：方法：我们从 11 家机构共收集了 559 例 EGC 的 8280 张 EUS 图像。在该数据集中，来自一家机构的 3451 张图像（285 例）被用作开发数据集。人工智能模型包括分割和分类步骤，然后使用 CycleGAN 方法弥合不同设备采集的 EUS 图像之间的差异。人工智能模型的性能是通过与开发数据集来自同一机构的内部验证数据集（1726 张图像，135 个病例）进行评估的。外部验证使用从其他 10 家机构收集的图像（3103 幅图像，139 个病例）：内部验证数据集中人工智能模型的曲线下面积（AUC）为 0.870（95% CI：0.796-0.944）。在诊断性能方面，人工智能模型、专家（n = 6）和非专家（n = 8）的准确性/敏感性/特异性值分别为 82.2/63.4/90.4%、81.9/66.3/88.7% 和 68.3/60.9/71.5%。外部验证数据集中人工智能模型的AUC为0.815（95% CI：0.743-0.886）。在外部验证数据集中，人工智能模型（74.1/73.1/75.0%）和专家实时诊断（75.5/79.1/72.2%）的准确性/敏感性/特异性值相当：我们的人工智能模型显示出了与专家相当的诊断性能。

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

Journal of Gastroenterology 医学-胃肠肝病学

CiteScore

12.20

自引率

1.60%

发文量

审稿时长

4-8 weeks

期刊介绍： The Journal of Gastroenterology, which is the official publication of the Japanese Society of Gastroenterology, publishes Original Articles (Alimentary Tract/Liver, Pancreas, and Biliary Tract), Review Articles, Letters to the Editors and other articles on all aspects of the field of gastroenterology. Significant contributions relating to basic research, theory, and practice are welcomed. These publications are designed to disseminate knowledge in this field to a worldwide audience, and accordingly, its editorial board has an international membership.