Publication

A stacking-based artificial intelligence framework for an effective detection and localization of colon polyps

View document

Bibliographic Details
Summary:Polyp detection through colonoscopy is a widely used method to prevent colorectal cancer. The automation of this process aided by artificial intelligence allows faster and improved detection of polyps that can be missed during a standard colonoscopy. In this work, we propose implementing different object detection algorithms for polyp detection. To improve the mean average precision (mAP) of the detection, we combine the baseline models through a stacking approach. The experiments demonstrate the potential of this new methodology, which can reduce the workload for oncologists and increase the precision of the localization of polyps. Our proposal achieves an mAP of 0.86, translated into an improvement of 34.9% compared to the best baseline model and 28.8% with respect to the weighted boxes fusion ensemble technique.
Main Authors:Albuquerque, Carina
Other Authors:Henriques, Roberto; Castelli, Mauro
Subject:Humans Artificial Intelligence Colonic Polyps/diagnosis Algorithms Colon Colonoscopy/methods Colorectal Neoplasms/diagnosis General SDG 3 - Good Health and Well-being
Year:2022
Country:Portugal
Document type:article
Access type:open access
Associated institution:Universidade Nova de Lisboa
Language:English
Origin:Repositório Institucional da UNL
Description
Summary:Polyp detection through colonoscopy is a widely used method to prevent colorectal cancer. The automation of this process aided by artificial intelligence allows faster and improved detection of polyps that can be missed during a standard colonoscopy. In this work, we propose implementing different object detection algorithms for polyp detection. To improve the mean average precision (mAP) of the detection, we combine the baseline models through a stacking approach. The experiments demonstrate the potential of this new methodology, which can reduce the workload for oncologists and increase the precision of the localization of polyps. Our proposal achieves an mAP of 0.86, translated into an improvement of 34.9% compared to the best baseline model and 28.8% with respect to the weighted boxes fusion ensemble technique.

Funded activities

Loading funded projects...