Breast Cancer Detection and Classification Using Deep CNN Techniques

Breast cancer is a commonly diagnosed disease in women. Early detection, personalized treatment approach, and better understanding are necessary for patients to survive. In this work, deep learning network traditional convolution were both employed with the Digital Database Screening Mammography (DDSM) dataset. images subjected background removal followed by Wiener filtering contrast limited histogram equalization (CLAHE) filter image restoration. Wavelet packet decomposition (WPD) using Daubechies wavelet level 3 (db3) was improve smoothness of images. For breast recognition, these preprocessed first fed neural networks, namely GoogleNet AlexNet Adam. Root mean square propagation (RMSprop) stochastic gradient descent momentum (SGDM) optimizers used different rates, such as 0.01, 0.001, 0.0001. As medical imaging necessitates presence discriminative features classification, pretrained architectures extract complicated from increase recognition rate. latter part study, particle swarm optimization-based multi-layer perceptron (PSO-MLP) ant colony (ACO-MLP) statistical features, skewness, kurtosis, variance, entropy, contrast, correlation, energy, homogeneity, mean, which extracted image. The performance compared AlexNet, PSO-MLP, ACO-MLP terms accuracy, loss rate, runtime found achieve an accuracy 99% lower rate 0.1547 lowest run time 4.14 minutes.