Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach

Abdominal computed tomography (CT) is a frequently used imaging modality for evaluating gastrointestinal diseases. The detection of colorectal cancer is often realized using CT before a more invasive colonoscopy. When a CT exam is performed for indications other than colorectal evaluation, the tortu...

Full description

Bibliographic Details
Main Authors: Prasan Kumar Sahoo, Pushpanjali Gupta, Ying-Chieh Lai, Sum-Fu Chiang, Jeng-Fu You, Djeane Debora Onthoni, Yih-Jong Chern
Format: Article
Language:English
Published: MDPI AG 2023-08-01
Series:Bioengineering
Subjects:
Online Access:https://www.mdpi.com/2306-5354/10/8/972
_version_ 1797585472757694464
author Prasan Kumar Sahoo
Pushpanjali Gupta
Ying-Chieh Lai
Sum-Fu Chiang
Jeng-Fu You
Djeane Debora Onthoni
Yih-Jong Chern
author_facet Prasan Kumar Sahoo
Pushpanjali Gupta
Ying-Chieh Lai
Sum-Fu Chiang
Jeng-Fu You
Djeane Debora Onthoni
Yih-Jong Chern
author_sort Prasan Kumar Sahoo
collection DOAJ
description Abdominal computed tomography (CT) is a frequently used imaging modality for evaluating gastrointestinal diseases. The detection of colorectal cancer is often realized using CT before a more invasive colonoscopy. When a CT exam is performed for indications other than colorectal evaluation, the tortuous structure of the long, tubular colon makes it difficult to analyze the colon carefully and thoroughly. In addition, the sensitivity of CT in detecting colorectal cancer is greatly dependent on the size of the tumor. Missed incidental colon cancers using CT are an emerging problem for clinicians and radiologists; consequently, the automatic localization of lesions in the CT images of unprepared bowels is needed. Therefore, this study used artificial intelligence (AI) to localize colorectal cancer in CT images. We enrolled 190 colorectal cancer patients to obtain 1558 tumor slices annotated by radiologists and colorectal surgeons. The tumor sites were double-confirmed via colonoscopy or other related examinations, including physical examination or image study, and the final tumor sites were obtained from the operation records if available. The localization and training models used were RetinaNet, YOLOv3, and YOLOv8. We achieved an F1 score of 0.97 (±0.002), a mAP of 0.984 when performing slice-wise testing, 0.83 (±0.29) sensitivity, 0.97 (±0.01) specificity, and 0.96 (±0.01) accuracy when performing patient-wise testing using our derived model YOLOv8 with hyperparameter tuning.
first_indexed 2024-03-11T00:06:39Z
format Article
id doaj.art-977ce2e493b34e8cad5d66f2f992eb15
institution Directory Open Access Journal
issn 2306-5354
language English
last_indexed 2024-03-11T00:06:39Z
publishDate 2023-08-01
publisher MDPI AG
record_format Article
series Bioengineering
spelling doaj.art-977ce2e493b34e8cad5d66f2f992eb152023-11-19T00:18:34ZengMDPI AGBioengineering2306-53542023-08-0110897210.3390/bioengineering10080972Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning ApproachPrasan Kumar Sahoo0Pushpanjali Gupta1Ying-Chieh Lai2Sum-Fu Chiang3Jeng-Fu You4Djeane Debora Onthoni5Yih-Jong Chern6Department of Computer Science and Information Engineering, Chang Gung University, Guishan, Taoyuan 33302, TaiwanDepartment of Computer Science and Information Engineering, Chang Gung University, Guishan, Taoyuan 33302, TaiwanDepartment of Medical Imaging and Intervention, Chang Gung Memorial Hospital, Linkou, New Taipei City 33305, TaiwanDivision of Colon and Rectal Surgery, Chang Gung Memorial Hospital, Linkou, New Taipei City 33305, TaiwanDivision of Colon and Rectal Surgery, Chang Gung Memorial Hospital, Linkou, New Taipei City 33305, TaiwanDepartment of Computer Science and Information Engineering, Chang Gung University, Guishan, Taoyuan 33302, TaiwanDivision of Colon and Rectal Surgery, Chang Gung Memorial Hospital, Linkou, New Taipei City 33305, TaiwanAbdominal computed tomography (CT) is a frequently used imaging modality for evaluating gastrointestinal diseases. The detection of colorectal cancer is often realized using CT before a more invasive colonoscopy. When a CT exam is performed for indications other than colorectal evaluation, the tortuous structure of the long, tubular colon makes it difficult to analyze the colon carefully and thoroughly. In addition, the sensitivity of CT in detecting colorectal cancer is greatly dependent on the size of the tumor. Missed incidental colon cancers using CT are an emerging problem for clinicians and radiologists; consequently, the automatic localization of lesions in the CT images of unprepared bowels is needed. Therefore, this study used artificial intelligence (AI) to localize colorectal cancer in CT images. We enrolled 190 colorectal cancer patients to obtain 1558 tumor slices annotated by radiologists and colorectal surgeons. The tumor sites were double-confirmed via colonoscopy or other related examinations, including physical examination or image study, and the final tumor sites were obtained from the operation records if available. The localization and training models used were RetinaNet, YOLOv3, and YOLOv8. We achieved an F1 score of 0.97 (±0.002), a mAP of 0.984 when performing slice-wise testing, 0.83 (±0.29) sensitivity, 0.97 (±0.01) specificity, and 0.96 (±0.01) accuracy when performing patient-wise testing using our derived model YOLOv8 with hyperparameter tuning.https://www.mdpi.com/2306-5354/10/8/972colorectal cancerdeep learninglocalizationcomputed tomography
spellingShingle Prasan Kumar Sahoo
Pushpanjali Gupta
Ying-Chieh Lai
Sum-Fu Chiang
Jeng-Fu You
Djeane Debora Onthoni
Yih-Jong Chern
Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
Bioengineering
colorectal cancer
deep learning
localization
computed tomography
title Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
title_full Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
title_fullStr Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
title_full_unstemmed Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
title_short Localization of Colorectal Cancer Lesions in Contrast-Computed Tomography Images via a Deep Learning Approach
title_sort localization of colorectal cancer lesions in contrast computed tomography images via a deep learning approach
topic colorectal cancer
deep learning
localization
computed tomography
url https://www.mdpi.com/2306-5354/10/8/972
work_keys_str_mv AT prasankumarsahoo localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT pushpanjaligupta localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT yingchiehlai localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT sumfuchiang localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT jengfuyou localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT djeanedeboraonthoni localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach
AT yihjongchern localizationofcolorectalcancerlesionsincontrastcomputedtomographyimagesviaadeeplearningapproach