<p>The cervical cancer occurs in the internal cervix region of the women patients, and its symptoms are delivered very slowly. Its detection is quite a cumbersome task. Thus, a cervical concentrated convolutional neural network (CCCNN) is proposed in the given paper using an external feature-based deep learning algorithm. This methodology requires training of the cervical images in both benign and malignant case images through enhancement, Non-Subsampled Contourlet Transform (NSCT), and external feature map construction. This produces individual sequences for healthy and cancerous cervical images from the dataset. Then, these trained sequences are fed into the testing block of the trained neural network. This block receives the test cervical image and enhances the internal pixels and transforms them into slices using NSCT. It is followed by the Improved Morphological Segmentation Algorithm (IMSA), which segments the cancer pixels in the abnormal cervical image more accurately than the conventional segmentation algorithms. The proposed work is validated on two independent cervical image datasets with respect to various performance evaluation metrics and the results are highly promising as compared to existing recent studies.</p>