Automatic colonic polyp detection using integration of modified deep residual convolutional neural network and ensemble learning approaches. (July 2021)

• Record Type: Journal Article
• Authors: Liew, Win Sheng; Tang, Tong Boon; Lin, Cheng-Hung; Lu, Cheng-Kai
• Is Part Of: Computer methods and programs in biomedicine. Volume 206(2021)
• Publication Date: 2021-07
• Subjects: colorectal cancer (CRC) -- polyps -- deep residual network -- principal component analysis -- AdaBoost ensemble learning; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2021.106114
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: Existing computer-aided diagnosis techniques still have room to improve further, especially when high sensitivity is an important concern. A new combination of modified deep residual network, principal component analysis (PCA), and ensemble learning methods has been developed to detect colonic polyps during the colonoscopy screening automatically. A deep residual network is modified to improve the performance while reducing network complexity, making the network lightweight. A dimension reduction technique, namely PCA, is used together with AdaBoost to prevent overfitting. The proposed method achieved a Matthews Correlation Coefficient (MCC) of 0.9819 with accuracy, sensitivity, precision, and specificity are 99.10%, 98.82%, 99.37%, and 99.38%, respectively, which outperforming the state-of-the-art techniques. Abstract: Background and Objective: The increased incidence of colorectal cancer (CRC) and its mortality rate have attracted interest in the use of artificial intelligence (AI) based computer-aided diagnosis (CAD) tools to detect polyps at an early stage. Although these CAD tools have thus far achieved a good accuracy level to detect polyps, they still have room to improve further (e.g. sensitivity). Therefore, a new CAD tool is developed in this study to detect colonic polyps accurately. Methods: In this paper, we propose a novel approach to distinguish colonic polyps by integrating several techniques, including a modified deep residual network, principal component analysis and AdaBoost ensemble learning. A powerful deep residual network architecture, ResNet-50, was investigated to reduce the computational time by altering its architecture. To keep the interference to a minimum, median filter, image thresholding, contrast enhancement, and normalisation techniques were exploited on the endoscopic images to train the classification model. Three publicly available datasets, i.e., Kvasir, ETIS-LaribPolypDB, and CVC-ClinicDB, were merged to train the model, which included images with and without polyps. Results: The proposed approach trained with a combination of three datasets achieved Matthews Correlation Coefficient (MCC) of 0.9819 with accuracy, sensitivity, precision, and specificity of 99.10%, 98.82%, 99.37%, and 99.38%, respectively. Conclusions: These results show that our method could repeatedly classify endoscopic images automatically and could be used to effectively develop computer-aided diagnostic tools for early CRC detection.