A new parallel deep learning algorithm for breast cancer classification

Now diagnostic methods with the help of machine learning have been able to help doctors in this field. One of the most important of these methods is deep learning, which has gotten good answers in images containing cancer. Increasing the accuracy of deep neural network classifiers can increase the diagnosis of breast cancer. In this paper, we have tried to achieve higher accuracy than non-parallel models with the help of a parallel model of a deep neural network. The proposed method is a parallel hybrid method combining AlexNet and VGGNet networks applied in parallel to mammographic images. The database used in this article is INBreast. The results obtained from this method show a ۴% increase compared to some other classification models so that in the type of density ۱, it has achieved about ۹۹.۷%. In the case of other densities, an accuracy of nearly ۹۹% has been obtained.