Saved in:
Bibliographic Details
Main Authors: Peter, Ojonugwa Oluwafemi Ejiga, Oluwapemiisin, Akingbola, Chetachi, Amalahu, Opeyemi, Adeniran, Khalifa, Fahmi, Rahman, Md Mahmudur
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.06170
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916887675273216
author Peter, Ojonugwa Oluwafemi Ejiga
Oluwapemiisin, Akingbola
Chetachi, Amalahu
Opeyemi, Adeniran
Khalifa, Fahmi
Rahman, Md Mahmudur
author_facet Peter, Ojonugwa Oluwafemi Ejiga
Oluwapemiisin, Akingbola
Chetachi, Amalahu
Opeyemi, Adeniran
Khalifa, Fahmi
Rahman, Md Mahmudur
contents Colonoscopy is a vital tool for the early diagnosis of colorectal cancer, which is one of the main causes of cancer-related mortality globally; hence, it is deemed an essential technique for the prevention and early detection of colorectal cancer. The research introduces a unique multidirectional architectural framework to automate polyp detection within colonoscopy images while helping resolve limited healthcare dataset sizes and annotation complexities. The research implements a comprehensive system that delivers synthetic data generation through Stable Diffusion enhancements together with detection and segmentation algorithms. This detection approach combines Faster R-CNN for initial object localization while the Segment Anything Model (SAM) refines the segmentation masks. The faster R-CNN detection algorithm achieved a recall of 93.08% combined with a precision of 88.97% and an F1 score of 90.98%.SAM is then used to generate the image mask. The research evaluated five state-of-the-art segmentation models that included U-Net, PSPNet, FPN, LinkNet, and MANet using ResNet34 as a base model. The results demonstrate the superior performance of FPN with the highest scores of PSNR (7.205893) and SSIM (0.492381), while UNet excels in recall (84.85%) and LinkNet shows balanced performance in IoU (64.20%) and Dice score (77.53%).
format Preprint
id arxiv_https___arxiv_org_abs_2508_06170
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Synthetic Data-Driven Multi-Architecture Framework for Automated Polyp Segmentation Through Integrated Detection and Mask Generation
Peter, Ojonugwa Oluwafemi Ejiga
Oluwapemiisin, Akingbola
Chetachi, Amalahu
Opeyemi, Adeniran
Khalifa, Fahmi
Rahman, Md Mahmudur
Computer Vision and Pattern Recognition
Artificial Intelligence
Colonoscopy is a vital tool for the early diagnosis of colorectal cancer, which is one of the main causes of cancer-related mortality globally; hence, it is deemed an essential technique for the prevention and early detection of colorectal cancer. The research introduces a unique multidirectional architectural framework to automate polyp detection within colonoscopy images while helping resolve limited healthcare dataset sizes and annotation complexities. The research implements a comprehensive system that delivers synthetic data generation through Stable Diffusion enhancements together with detection and segmentation algorithms. This detection approach combines Faster R-CNN for initial object localization while the Segment Anything Model (SAM) refines the segmentation masks. The faster R-CNN detection algorithm achieved a recall of 93.08% combined with a precision of 88.97% and an F1 score of 90.98%.SAM is then used to generate the image mask. The research evaluated five state-of-the-art segmentation models that included U-Net, PSPNet, FPN, LinkNet, and MANet using ResNet34 as a base model. The results demonstrate the superior performance of FPN with the highest scores of PSNR (7.205893) and SSIM (0.492381), while UNet excels in recall (84.85%) and LinkNet shows balanced performance in IoU (64.20%) and Dice score (77.53%).
title Synthetic Data-Driven Multi-Architecture Framework for Automated Polyp Segmentation Through Integrated Detection and Mask Generation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2508.06170