Multicentered Data Based Polyp Detection Using Colonoscopy Images Using DNN

Multicentered Data Based Polyp Detection Using Colonoscopy Images Using DNN Abdus Salam Department of Electrical & Computer Engineering Rajshahi University of Engineering & Technology Rajshahi Bangladesh Department of Electrical Engineering Qatar University Doha Qatar Moajjem Hossain Chowdhury Department of Electrical Engineering Qatar University Doha Qatar M. Murugappan Intelligent Signal Processing (ISP) Research Lab, Department of Electronics and Communication Engineering Kuwait College of Science and Technology Doha Kuwait Department of Electronics and Communication Engineering Vels Institute of Sciences, Technology, and Advanced Studies Chennai Tamil Nadu India https://orcid.org/0000-0002-5839-4589 Muhammad E. H. Chowdhury Department of Electrical Engineering Qatar University Doha Qatar ABSTRACT

The diagnosis and screening of colon polyps are essential for the early detection of colorectal cancer. Polyps can be identified through colonoscopies before becoming cancerous, making accurate detection and prompt intervention critical for colorectal health. A comprehensive evaluation of deep learning models using colonoscopy images and comparisons with state‐of‐the‐art models is presented in this study. A total of 7900 still and video sequence images from the PolypGen multicenter data set were used to train cutting‐edge object detection models, including YOLOv5, YOLOv7, YOLOv8, and F‐RCNN + ResNet101. In terms of accuracy, precision, recall, and mAP, the YOLOv8x model achieved the best performance with an F1 score of 0.9058, accuracy of 0.949, precision of 0.863, and mAP@0.5. The robustness of the model was further confirmed across varying patient demographics and conditions using the external Kvasir data set. To enhance interpretability, the EigenCam explainable AI (XAI) technique was used, offering visual insights into the model's decision‐making process by highlighting the most influential regions in the input images.
06 05 2025 07 2025 e70123 10.1002/ima.70123 2 10.1002/crossmark_policy onlinelibrary.wiley.com true 2024-05-22 2025-05-18 2025-06-05 http://onlinelibrary.wiley.com/termsAndConditions#vor 10.1002/ima.70123 https://onlinelibrary.wiley.com/doi/10.1002/ima.70123 https://onlinelibrary.wiley.com/doi/pdf/10.1002/ima.70123 10.1053/j.gastro.2012.06.001 10.7326/0003‐4819‐139‐12‐200312160‐00005 10.7326/0003‐4819‐119‐8‐199310150‐00010 10.3322/CA.2007.0018 10.1093/ANNONC/MDW235 10.1053/J.GASTRO.2018.08.063 10.1148/RADIOL.11101887/‐/DC1 10.1046/j.1365‐2168.2002.02120.x 10.1056/NEJMoa1100370 10.1038/ajg.2014.435 10.1016/j.measurement.2022.111485 10.1007/s00521‐023‐08407‐1 10.1186/S12916‐019‐1426‐2 10.1007/s44230‐024‐00067‐1 10.1038/AJG.2017.174 10.1109/CISP.2015.7407907 10.1109/TITB.2003.813794 10.1109/JSEN.2020.3036005 X.Yang E.Song G.Ma et al. “YOLO‐OB: An Improved Anchor‐Free Real‐Time Multiscale Colon Polyp Detector in Colonoscopy ”2023. 10.1007/s10489‐022‐04299‐1 10.1016/J.MEDIA.2022.102625 10.1109/ACCESS.2019.2921027 10.3390/APP9122404 10.1016/J.BSPC.2019.04.007 D.Jha N. K.Tomar V.Sharma et al. “PolypDB: A Curated Multi‐Center Dataset for Development of AI Algorithms in Colonoscopy ”2024. 10.1038/s41597‐023‐01981‐y 10.1038/s41598‐024‐66642‐5 10.1109/ACCESS.2022.3171238 10.1371/JOURNAL.PONE.0250632 10.1007/S11042‐023‐17138‐3 10.1016/J.COMPBIOMED.2021.105031 10.1109/IWSSIP48289.2020.9145130 10.1007/978-3-030-00889-5_1 10.1038/s41597‐020‐00622‐y 10.1109/ACCESS.2021.3063716 CEUR Workshop Proceedings Alam S. 2882 2020 Automatic Polyp Segmentation Using U‐Net‐ResNet50 10.3390/s21165315 2019 13th International Symposium on Medical Information and Communication Technology Qadir H. A. 1 2019 10.1016/j.compeleceng.2020.106959 10.1016/J.ESWA.2023.119741 10.3390/app10238501 10.1016/j.tige.2021.09.004 Advances in Neural Information Processing Systems Ren S. 28 2015 Faster R‐CNN: Towards Real‐Time Object Detection With Region Proposal Networks 10.1109/CVPR.2014.81 10.1109/ICCV.2015.169 R.Girshick “Fast R‐CNN”2015. K.SimonyanandA.Zisserman “Very Deep Convolutional Networks for Large‐Scale Image Recognition ”2014. 10.3348/KJR.2019.0651 10.1109/CVPR.2016.91 C.‐Y.Wang A.Bochkovskiy andH.‐Y. M.Liao “Scaled‐YOLOv4: Scaling Cross Stage Partial Network ”n.d. 10.1109/ACCESS.2018.2856402 10.1109/CVPR.2016.319 10.1109/ICCV.2017.74 10.1109/WACV.2018.00097 10.1109/IJCNN48605.2020.9206626 10.1109/ACCESS.2025.3536621 10.1016/j.patrec.2023.03.003