CNN-Based Colon Cancer Recognition Model
Abdulfattah E. Ba Alawi, Ferhat Bozkurt · 2023
Digital pathology is being used extensively for the diagnosis of tumors. Disappointingly, the existing approaches are still constrained whenever confronted with a resolution, a size of images, and a lack of extensively cleaned datasets. Further, the recognition accuracy mostly does not reach high scores. In terms of Deep Learning (DL) approaches' capacity to handle extensive applications, such an approach appears to be an absorbing solution for both categorization and tissue segmentation in histopathology images. The present study concentrates on the application of deep learning models in the classification of the context of histopathology data and the recognition of colon cancer. In this, cutting-edge Fully Convolutional Network (CNN) models such as DenseNet121, EfficientNetB7, EfficientNetB1, EfficientNetB2, and DenseNet201 have been evaluated for the recognition of Colon Cancer. The assessment of the algorithms used for the proposed CNN-based colon cancer detection model ensures reliable classification findings with EfficientNetB2 attaining up to 99.9994% in terms of accuracy.