@article{scholars14799,
            year = {2021},
             doi = {10.1016/j.cmpb.2021.106114},
          volume = {206},
            note = {cited By 22},
       publisher = {Elsevier Ireland Ltd},
         journal = {Computer Methods and Programs in Biomedicine},
           title = {Automatic colonic polyp detection using integration of modified deep residual convolutional neural network and ensemble learning approaches},
            issn = {01692607},
          author = {Liew, W. S. and Tang, T. B. and Lin, C.-H. and Lu, C.-K.},
        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. {\^A}{\copyright} 2021},
        keywords = {Classification (of information); Computer aided diagnosis; Convolutional neural networks; Deep neural networks; Diseases; Endoscopy; Image enhancement; Median filters; Network architecture; Principal component analysis, Adaboost ensemble learning; Colonic polyp detection; Colonic polyps; Colorectal cancer; Computer-aided; Deep residual network; Diagnosis tools; Endoscopic image; Polyp; Principal-component analysis, Adaptive boosting, Article; automation; cancer incidence; classification; colon polyp; colonoscopy; colorectal cancer; computer assisted diagnosis; contrast enhancement; convolutional neural network; correlation coefficient; diagnostic accuracy; diagnostic test accuracy study; dimensionality reduction; early cancer diagnosis; feature extraction; human; learning algorithm; mortality rate; principal component analysis; residual neural network; sensitivity and specificity; artificial intelligence; colon polyp; diagnostic imaging; machine learning, Artificial Intelligence; Colonic Polyps; Diagnosis, Computer-Assisted; Humans; Machine Learning; Neural Networks, Computer},
             url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85105593805&doi=10.1016\%2fj.cmpb.2021.106114&partnerID=40&md5=9c0357281484fc0b8733c207d0ba55df}
}