


Volume 20 No 22 (2022)
Download PDF
Binary Classified Convolutional Neural Network Model for Liver Cancer Tissue Identification at Early Stages
Ankita, Kamal Malik
Abstract
Cancer is the abnormal growth of cells in a particular part of a body which forms lumps. These lumps are commonly called tumours. Liver cancer is a deadly and aggressive disease that grows rapidly in the body. Diagnosis of liver cancer with early detection and high accuracy can improve patient survival time. Histopathological image analysis is time-consuming and ambitious to manually detect cancer cells. In this work, a methodology is applied to PET – CT images which help to identify cancer cells as well as predict prognosis. First, a median image filter is applied to eliminate noise from the input CT images before applying them to the model. In this paper, a trained ConvNet model is implemented to classify a patient's liver tumour. Deep learning resolves the complex problem and challenge of cancer recognition using neural networks. Nevertheless, deep learning is a powerful technique which uses neural networks to enable the classification of abdominal hepatic lesion images. This model partially solves the dimensionality problem. For this proposed model, CT images of 214 liver patients were collected and the data was split in the 80:20 ratios for training and testing. The model achieves a classification accuracy of 95.35% with a validation loss of 0.20% using a deep learning-based CNN model.
Keywords
CT images, Liver Lesion, Median Filter and Convolutional Neural Network.
Copyright
Copyright © Neuroquantology
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Articles published in the Neuroquantology are available under Creative Commons Attribution Non-Commercial No Derivatives Licence (CC BY-NC-ND 4.0). Authors retain copyright in their work and grant IJECSE right of first publication under CC BY-NC-ND 4.0. Users have the right to read, download, copy, distribute, print, search, or link to the full texts of articles in this journal, and to use them for any other lawful purpose.