Multi-class Classification of Skin Diseases using Pre-trained DenseNet Architecture on Dermoscopy Images

Authors:

Abstract
Dermoscopy is a non-invasive procedure used to visualize skin lesions at a magnified level. In this research, are trained DenseNet model is used for the classification of seven class skin disease dermoscopy images. The proposed method is simulated using different confusion matrix parameters on the HAMIOOOO dataset. The extracted features are then used to classify the images into one of several predefined skin disease categories. A sizable collection of dermoscopy images is used to train the CNN model images and among the classification of the seven classes of the skin achieved accuracy as high as 98% for one of the classes and the overall accuracy of the model after training is 88.78. The results of the research demonstrate the achievement of using CNNs for classifying skin diseases from dermoscopy images, which has potential applications in medical diagnosis and treatment planning.

This publication has 19 references indexed in Scilit: