Jin Rong Tan, Yan Gao, Raghavan Raghuraman, Daniel Ting, Kang Min Wong, Lionel Tim-Ee Cheng, Hong Choon Oh, Siang Hiong Goh, Yet Yen Yan
{"title":"深度学习算法在髋关节术后植入物切口分类和定位中的应用。","authors":"Jin Rong Tan, Yan Gao, Raghavan Raghuraman, Daniel Ting, Kang Min Wong, Lionel Tim-Ee Cheng, Hong Choon Oh, Siang Hiong Goh, Yet Yen Yan","doi":"10.1007/s00256-024-04692-6","DOIUrl":null,"url":null,"abstract":"<p><strong>Objective: </strong>This study aims to explore the feasibility of employing convolutional neural networks for detecting and localizing implant cutouts on anteroposterior pelvic radiographs.</p><p><strong>Materials and methods: </strong>The research involves the development of two Deep Learning models. Initially, a model was created for image-level classification of implant cutouts using 40191 pelvic radiographs obtained from a single institution. The radiographs were partitioned into training, validation, and hold-out test datasets in a 6/2/2 ratio. Performance metrics including the area under the receiver operator characteristics curve (AUROC), sensitivity, and specificity were calculated using the test dataset. Additionally, a second object detection model was trained to localize implant cutouts within the same dataset. Bounding box visualizations were generated on images predicted as cutout-positive by the classification model in the test dataset, serving as an adjunct for assessing algorithm validity.</p><p><strong>Results: </strong>The classification model had an accuracy of 99.7%, sensitivity of 84.6%, specificity of 99.8%, AUROC of 0.998 (95% CI: 0.996, 0.999) and AUPRC of 0.774 (95% CI: 0.646, 0.880). From the pelvic radiographs predicted as cutout-positive, the object detection model could achieve 95.5% localization accuracy on true positive images, but falsely generated 14 results from the 15 false-positive predictions.</p><p><strong>Conclusion: </strong>The classification model showed fair accuracy for detection of implant cutouts, while the object detection model effectively localized cutout. This serves as proof of concept of using a deep learning-based approach for classification and localization of implant cutouts from pelvic radiographs.</p>","PeriodicalId":21783,"journal":{"name":"Skeletal Radiology","volume":" ","pages":"67-75"},"PeriodicalIF":1.9000,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Application of deep learning algorithms in classification and localization of implant cutout for the postoperative hip.\",\"authors\":\"Jin Rong Tan, Yan Gao, Raghavan Raghuraman, Daniel Ting, Kang Min Wong, Lionel Tim-Ee Cheng, Hong Choon Oh, Siang Hiong Goh, Yet Yen Yan\",\"doi\":\"10.1007/s00256-024-04692-6\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><strong>Objective: </strong>This study aims to explore the feasibility of employing convolutional neural networks for detecting and localizing implant cutouts on anteroposterior pelvic radiographs.</p><p><strong>Materials and methods: </strong>The research involves the development of two Deep Learning models. Initially, a model was created for image-level classification of implant cutouts using 40191 pelvic radiographs obtained from a single institution. The radiographs were partitioned into training, validation, and hold-out test datasets in a 6/2/2 ratio. Performance metrics including the area under the receiver operator characteristics curve (AUROC), sensitivity, and specificity were calculated using the test dataset. Additionally, a second object detection model was trained to localize implant cutouts within the same dataset. Bounding box visualizations were generated on images predicted as cutout-positive by the classification model in the test dataset, serving as an adjunct for assessing algorithm validity.</p><p><strong>Results: </strong>The classification model had an accuracy of 99.7%, sensitivity of 84.6%, specificity of 99.8%, AUROC of 0.998 (95% CI: 0.996, 0.999) and AUPRC of 0.774 (95% CI: 0.646, 0.880). From the pelvic radiographs predicted as cutout-positive, the object detection model could achieve 95.5% localization accuracy on true positive images, but falsely generated 14 results from the 15 false-positive predictions.</p><p><strong>Conclusion: </strong>The classification model showed fair accuracy for detection of implant cutouts, while the object detection model effectively localized cutout. This serves as proof of concept of using a deep learning-based approach for classification and localization of implant cutouts from pelvic radiographs.</p>\",\"PeriodicalId\":21783,\"journal\":{\"name\":\"Skeletal Radiology\",\"volume\":\" \",\"pages\":\"67-75\"},\"PeriodicalIF\":1.9000,\"publicationDate\":\"2026-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Skeletal Radiology\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1007/s00256-024-04692-6\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"2024/5/21 0:00:00\",\"PubModel\":\"Epub\",\"JCR\":\"Q2\",\"JCRName\":\"ORTHOPEDICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Skeletal Radiology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s00256-024-04692-6","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/5/21 0:00:00","PubModel":"Epub","JCR":"Q2","JCRName":"ORTHOPEDICS","Score":null,"Total":0}
Application of deep learning algorithms in classification and localization of implant cutout for the postoperative hip.
Objective: This study aims to explore the feasibility of employing convolutional neural networks for detecting and localizing implant cutouts on anteroposterior pelvic radiographs.
Materials and methods: The research involves the development of two Deep Learning models. Initially, a model was created for image-level classification of implant cutouts using 40191 pelvic radiographs obtained from a single institution. The radiographs were partitioned into training, validation, and hold-out test datasets in a 6/2/2 ratio. Performance metrics including the area under the receiver operator characteristics curve (AUROC), sensitivity, and specificity were calculated using the test dataset. Additionally, a second object detection model was trained to localize implant cutouts within the same dataset. Bounding box visualizations were generated on images predicted as cutout-positive by the classification model in the test dataset, serving as an adjunct for assessing algorithm validity.
Results: The classification model had an accuracy of 99.7%, sensitivity of 84.6%, specificity of 99.8%, AUROC of 0.998 (95% CI: 0.996, 0.999) and AUPRC of 0.774 (95% CI: 0.646, 0.880). From the pelvic radiographs predicted as cutout-positive, the object detection model could achieve 95.5% localization accuracy on true positive images, but falsely generated 14 results from the 15 false-positive predictions.
Conclusion: The classification model showed fair accuracy for detection of implant cutouts, while the object detection model effectively localized cutout. This serves as proof of concept of using a deep learning-based approach for classification and localization of implant cutouts from pelvic radiographs.
期刊介绍:
Skeletal Radiology provides a forum for the dissemination of current knowledge and information dealing with disorders of the musculoskeletal system including the spine. While emphasizing the radiological aspects of the many varied skeletal abnormalities, the journal also adopts an interdisciplinary approach, reflecting the membership of the International Skeletal Society. Thus, the anatomical, pathological, physiological, clinical, metabolic and epidemiological aspects of the many entities affecting the skeleton receive appropriate consideration.
This is the Journal of the International Skeletal Society and the Official Journal of the Society of Skeletal Radiology and the Australasian Musculoskelelal Imaging Group.