Surgical Insight-guided Deep Learning for Colorectal Lesion Management.

IF 1.1 4区医学 Q3 SURGERY

Surgical Laparoscopy, Endoscopy & Percutaneous Techniques Pub Date : 2024-12-01 DOI:10.1097/SLE.0000000000001298

Ozan Can Tatar, Anil Çubukçu

{"title":"Surgical Insight-guided Deep Learning for Colorectal Lesion Management.","authors":"Ozan Can Tatar, Anil Çubukçu","doi":"10.1097/SLE.0000000000001298","DOIUrl":null,"url":null,"abstract":"Background: Colonoscopy stands as a pivotal diagnostic tool in identifying gastrointestinal diseases, including potentially malignant tumors. The procedure, however, faces challenges in the precise identification of lesions during visual inspections. The recent strides in AI and machine learning technologies have opened avenues for enhanced medical imaging analysis, including in the field of colonoscopy.Methods: In this study, we developed and evaluated a deep learning (DL) model, ColoNet, for detecting lesions in colonoscopic images. We analyzed 1760 images from 306 patients who underwent colorectal surgery between 2009 and 2022, meeting specific inclusion criteria. These images were used to train and validate ColoNet, employing the YOLOv8 architecture and various data augmentation techniques. Deep learning metrics are assessed via YOLO architecture and trained model diagnostic accuracy was assessed via sensitivity, specifity, positive predictive value, and negative predictive value.Results: Our results from the validation dataset revealed a precision of 0.79604, a recall of 0.78086, an mAP50 of 0.83243, and an mAP50-95 of 0.4439. In addition, on a separate real-time dataset of 91 images consisting both healthy and suspect lesions, ColoNet achieved a sensitivity of 70.73%, specificity of 92.00%, positive predictive value (PPV) of 87.88%, and negative predictive value (NPV) of 79.31%. The positive and negative likelihood ratios were 8.84 and 0.32, respectively, with an overall accuracy of 82.42%.Conclusions: In conclusion, our model has demonstrated promising results, indicating its potential as a valuable tool to assist surgeons during colonoscopy procedures. Its ability to detect suspicious lesions with potential malignancy offers a noteworthy advancement in the early diagnosis and management of colorectal cancers. Further multicentric, prospective research and validation are warranted to fully realize its clinical applicability and impact.","PeriodicalId":22092,"journal":{"name":"Surgical Laparoscopy, Endoscopy & Percutaneous Techniques","volume":"34 6","pages":"559-565"},"PeriodicalIF":1.1000,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Surgical Laparoscopy, Endoscopy & Percutaneous Techniques","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1097/SLE.0000000000001298","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"SURGERY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Colonoscopy stands as a pivotal diagnostic tool in identifying gastrointestinal diseases, including potentially malignant tumors. The procedure, however, faces challenges in the precise identification of lesions during visual inspections. The recent strides in AI and machine learning technologies have opened avenues for enhanced medical imaging analysis, including in the field of colonoscopy.

Methods: In this study, we developed and evaluated a deep learning (DL) model, ColoNet, for detecting lesions in colonoscopic images. We analyzed 1760 images from 306 patients who underwent colorectal surgery between 2009 and 2022, meeting specific inclusion criteria. These images were used to train and validate ColoNet, employing the YOLOv8 architecture and various data augmentation techniques. Deep learning metrics are assessed via YOLO architecture and trained model diagnostic accuracy was assessed via sensitivity, specifity, positive predictive value, and negative predictive value.

Results: Our results from the validation dataset revealed a precision of 0.79604, a recall of 0.78086, an mAP50 of 0.83243, and an mAP50-95 of 0.4439. In addition, on a separate real-time dataset of 91 images consisting both healthy and suspect lesions, ColoNet achieved a sensitivity of 70.73%, specificity of 92.00%, positive predictive value (PPV) of 87.88%, and negative predictive value (NPV) of 79.31%. The positive and negative likelihood ratios were 8.84 and 0.32, respectively, with an overall accuracy of 82.42%.

Conclusions: In conclusion, our model has demonstrated promising results, indicating its potential as a valuable tool to assist surgeons during colonoscopy procedures. Its ability to detect suspicious lesions with potential malignancy offers a noteworthy advancement in the early diagnosis and management of colorectal cancers. Further multicentric, prospective research and validation are warranted to fully realize its clinical applicability and impact.

查看原文本刊更多论文

求助全文

约1分钟内获得全文求助全文

来源期刊

Surgical Laparoscopy, Endoscopy & Percutaneous Techniques 医学-外科

CiteScore

2.00

自引率

10.00%

发文量

103

审稿时长

3-8 weeks

期刊介绍： Surgical Laparoscopy Endoscopy & Percutaneous Techniques is a primary source for peer-reviewed, original articles on the newest techniques and applications in operative laparoscopy and endoscopy. Its Editorial Board includes many of the surgeons who pioneered the use of these revolutionary techniques. The journal provides complete, timely, accurate, practical coverage of laparoscopic and endoscopic techniques and procedures; current clinical and basic science research; preoperative and postoperative patient management; complications in laparoscopic and endoscopic surgery; and new developments in instrumentation and technology.