Hyperparameter Tuned Deep Hybrid Denoising Autoencoder Breast Cancer Classification on Digital Mammograms

Breast Cancer (BC) is considered the most commonly scrutinized cancer in women worldwide, affecting one eight a lifetime. Mammography screening becomes such standard method that helpful identifying suspicious masses’ malignancy of BC at an initial level. However, prior identification masses mammograms was still challenging for extremely dense and breast categories needs effective automatic mechanisms helping radiotherapists diagnosis. Deep learning (DL) techniques were broadly utilized medical imaging applications, particularly mass classification. The advancements DL field paved way highly intellectual self-reliant computer-aided diagnosis (CAD) systems since capability Machine Learning (ML) constantly improving. This paper presents new Hyperparameter Tuned Hybrid Denoising Autoencoder Classification (HTDHDAE-BCC) on Digital Mammograms. presented HTDHDAE-BCC model examines mammogram images BC. In model, stage image preprocessing carried out using average median filter. addition, deep convolutional neural network-based Inception v4 employed to generate feature vectors. parameter tuning process uses binary spider monkey optimization (BSMO) algorithm. exploits chameleon swarm (CSO) with DHDAE experimental analysis performed MIAS database. outcomes demonstrate betterments over other recent approaches.