1Department of Electronics and Communication Engineering, Mahendra Institute of Technology (Autonomous), Namakkal, India
2Department of Electrical and Electronics Engineering, Mahendra Engineering College (Autonomous), Namakkal, India.
Article Publishing History
Received: 15/10/2020
Accepted After Revision: 29/12/2020
Breast cancer is the largest detection of cancer among women worldwide. Advancement in computer-aided diagnosis (CAD) makes it easy to detect and to classify benign and malignant images, henceforth to increase the life span of women. But fine-tuning of the accuracy of the existing CAD system comes to the limelight with the available resources. In recent study shows deep convolutional network provides greater accuracy. In this paper, we use deep CNN to extract the features with AlexNet. Then we Fine-tuned the various parameters to improve the accuracy with various optimizers and learning rates to classify the malignant and benign masses with CBIS-DDSM (Curated Breast Imaging Subset of DDSM) dataset. The two classifiers used the Support vector machine (SVM) and the Extreme Learning Machine (ELM) which provides an accuracy of 97.36% and 100% respectively.
Dcnn, Alexnet, Adam, Mammogram, Mass Classification.