AUTHOR=Khan Abdullah , Khan Asfandyar , Ullah Muneeb , Alam Muhammad Mansoor , Bangash Javed Iqbal , Suud Mazliham Mohd 

TITLE=A computational classification method of breast cancer images using the VGGNet model

JOURNAL=Frontiers in Computational Neuroscience

VOLUME=Volume 16 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2022.1001803

DOI=10.3389/fncom.2022.1001803

ISSN=1662-5188

ABSTRACT=Cancer is one of the most common diseases in the world. Breast cancer is the most common disease among women, where abnormal cells grow in an uncontrolled fashion. It is a very challenging task to detect and classify breast cancer. Therefore, different computational methods like KNN, SVM, Multi-Layer Perceptron, Decision Tree, and genetic algorithms were used for the detection and classification of breast cancer in the modern computational world. But all the techniques have their own limitations in terms of accuracy. Similarly, this research proposed a new VGGNet-based Convolutional Neural Network (CNN) model. The existing VGG-16 model consists of 16 layers, which causes overfitting on learning and testing data. This study enhances the existing models and reduces the number of layers down to 12, which consists of 6 convolutional, 3 maxpooling, 1 flattening, and 2 fully connected layers. The performance of the proposed model enhanced VGGNet is checked in terms of different parameters such as accuracy, loss, recall, precision, f-measure, specificity, and sensitivity. They are compared with the existing models such as CNN and LeNet, and models based on CNN on the breast cancer dataset for the aim of classification. Overall, experimental results show that the proposed VGGNet performed well on the breast cancer classification task in terms of different parameters.