Abstract:Diagnosis of skin lesions is a very challenging task due to high inter-class similarities and intra-class variations between lesions in terms of color, size, site and appearance. As a result, as is the case with many types of cancer, early detection of skin cancer is vital for survival. Advances in artificial intelligence, in particular, deep learning have enabled to design and implementation of intelligence-based lesion detection and classification systems that are based on visible light images that are capable of performing early and accurate diagnosis of different skin diseases. In most cases, the precision of these methods has reached a level of accuracy that is comparable to that achieved by a qualified dermatologist. This work presents potential skin lesion classification solutions based on the datasets taken from the most recent publicly available “Skin Lesion Analysis Towards Melanoma Detection” grand challenges ISIC 2018. The proposed classification approach leverages convolutional neural networks (CNN) and transfer learning to enhance skin classification. Different pre-trained models were applied, including VGG-Net, ResNet50, InceptionV3, Xception and DenseNet121. Additionally, the heavy class imbalance is examined as a critical problem for this dataset and multiple balancing techniques, such as weight balancing and data augmentation, are considered. Finally, an ensemble approach is evaluated by combining and averaging several CNN architectures to classify the seven different types of skin lesion. The experimental results indicate that the proposed frameworks exhibit promising results when compared with ISIC 2018 challenge live leaderboard.

Araguaia Medical Vision Lab at ISIC 2017 Skin Lesion Classification Challenge

Disease Classification within Dermascopic Images Using features extracted by ResNet50 and classification through Deep Forest

Image Classification of Melanoma, Nevus and Seborrheic Keratosis by Deep Neural Network Ensemble

Ensemble of Deep Convolutional Neural Network for Skin Lesion Classification in Dermoscopy Images

Deep Learning for Skin Lesion Classification

Deep-CLASS at ISIC Machine Learning Challenge 2018

Incorporating the Knowledge of Dermatologists to Convolutional Neural Networks for the Diagnosis of Skin Lesions

Transfer Learning for Melanoma Detection: Participation in ISIC 2017 Skin Lesion Classification Challenge

Automated Skin Lesion Classification Using Ensemble of Deep Neural Networks in ISIC 2018: Skin Lesion Analysis Towards Melanoma Detection Challenge

Skin Lesion Diagnosis Using Convolutional Neural Networks

Skin Lesion Analysis Towards Melanoma Detection Using Deep Learning Network

ISIC 2017 - Skin Lesion Analysis Towards Melanoma Detection

Skin Lesion Classification Using Hybrid Deep Neural Networks

SkinLesNet: Classification of Skin Lesions and Detection of Melanoma Cancer Using a Novel Multi-Layer Deep Convolutional Neural Network

Skin lesion classification using ensembles of multi-resolution EfficientNets with meta data

A Novel Hybrid Deep Learning Approach for Skin Lesion Segmentation and Classification

Random Forests and VGG-NET: An Algorithm for the ISIC 2017 Skin Lesion Classification Challenge

Skin Lesion Analysis Towards Melanoma Detection via End-to-end Deep Learning of Convolutional Neural Networks

Seven Class Classification of Skin Lesions by using Incremental Convolutional Neural Network in Python

Skin disease identification from dermoscopy images using deep convolutional neural network

WonDerM: Skin Lesion Classification with Fine-tuned Neural Networks