Dermatologist Level Dermoscopy Skin Cancer Classification Using Different Deep Learning Convolutional Neural Networks Algorithms

TL;DR

This study evaluates various deep learning CNN architectures on a large dermoscopy dataset, demonstrating that these models outperform dermatologists in classifying eight skin diseases with high accuracy.

Contribution

It compares multiple pre-trained CNN models on a large skin lesion dataset, showing superior performance over dermatologists in skin cancer classification.

Findings

Deep learning models outperform dermatologists by at least 11%.

ResNet 152 achieves 94.40% ROC AUC for melanoma.

DenseNet 201 achieves 99.30% ROC AUC for basal cell carcinoma.

Abstract

In this paper, the effectiveness and capability of convolutional neural networks have been studied in the classification of 8 skin diseases. Different pre-trained state-of-the-art architectures (DenseNet 201, ResNet 152, Inception v3, InceptionResNet v2) were used and applied on 10135 dermoscopy skin images in total (HAM10000: 10015, PH2: 120). The utilized dataset includes 8 diagnostic categories - melanoma, melanocytic nevi, basal cell carcinoma, benign keratosis, actinic keratosis and intraepithelial carcinoma, dermatofibroma, vascular lesions, and atypical nevi. The aim is to compare the ability of deep learning with the performance of highly trained dermatologists. Overall, the mean results show that all deep learning models outperformed dermatologists (at least 11%). The best ROC AUC values for melanoma and basal cell carcinoma are 94.40% (ResNet 152) and 99.30% (DenseNet 201) versus 82.26% and 88.82% of dermatologists, respectively. Also, DenseNet 201 had the highest macro and micro averaged AUC values for overall classification (98.16%, 98.79%, respectively).