Skin Lesion Detection and Classification Using Convolutional Neural Network for Deep Feature Extraction and Support Vector Machine

Pigmented skin lesion identification is essential for detecting harmful pathologies related to this large organ, especially cancer. An analysis of the different methods and projects developed diagnose these illnesses throughout years showed that they had become very useful tools identify melanoma, dermatofibroma, basal cell carcinoma, among other types cancer, are seen through use new computer-aided technologies. The most common diagnosis based on dermoscopy dermatologist expertise can improve accuracy with image detection techniques classification by computer. Therefore, study aims develop software models able detect classify following work images obtained from HAM10000 dataset, a database 10000 previously tested validated research use. main process divided into three relevant parts: segmentation, feature extraction (FE) using ten pre-trained Convolutional Neural Networks (CNNs), Support Vector Machine (SVM) establish model. According results, performed well segmentation step, showing average accuracies between 80.67% (Xception) 90% (Alexnet). In contrast without where no method reached 60%. AlexNet plus SVM model minor running time presented higher rate (90.34%) correct seven categories cutaneous lesions taken account.