Histopathological Image based Skin Cancer Classification Using CNN

The dataset consisting of H&E-stained digital microscopic histology images which was collected from 354 cancer patients on A.H.R.I. The dataset has four Classes (Benign,BCC,SCC and Melanoma). The dataset is divided into benign and malignant tumor of biopsies. Small patches were extracted at four magnifications of ×40, ×100, ×200, and ×400. The benign tumors were classified into eight subclasses which were Melanocytic Nevus, Prome, Pilomatrixomas, Sebonlaric Keratosis (Pigmented), Trico blastoma, Verica, Dermatofibroma, Mull scum Contagiosum, Infundibular Cyst and the malignant tumors were also classified into three subclasses which were basal cell carcinoma (BCC), squamous cell carcinoma (SCC) and Melanoma sometimes called Malignant Melanoma (MM). The dataset contains 16,099 total histology images with data augmentation and without data augmentation 4,357 total histology images. The owner of this Dataset is Alemu Abate Asheber MSc Software Engineer) . There is no third party in this dataset.