Computer-Aided Diagnosis of Skin Diseases Using Deep Neural Networks

Abstract

Propensity of skin diseases to manifest in a variety of forms, lack and maldistribution of qualified dermatologists, and exigency of timely and accurate diagnosis call for automated Computer-Aided Diagnosis (CAD). This study aims at extending previous works on CAD for dermatology by exploring the potential of Deep Learning to classify hundreds of skin diseases, improving classification performance, and utilizing disease taxonomy. We trained state-of-the-art Deep Neural Networks on two of the largest publicly available skin image datasets, namely DermNet and ISIC Archive, and also leveraged disease taxonomy, where available, to improve classification performance of these models. On DermNet we establish new state-of-the-art with 80% accuracy and 98% Area Under the Curve (AUC) for classification of 23 diseases. We also set precedence for classifying all 622 unique sub-classes in this dataset and achieved 67% accuracy and 98% AUC. On ISIC Archive we classified all 7 diseases with 93% average accuracy and 99% AUC. This study shows that Deep Learning has great potential to classify a vast array of skin diseases with near-human accuracy and far better reproducibility. It can have a promising role in practical real-time skin disease diagnosis by assisting physicians in large-scale screening using clinical or dermoscopic images.

Malignant Melanoma Classification Using Deep Learning: Datasets, Performance Measurements, Challenges and Opportunities

“…Dugonik et al [94] proposed an ensemble method consist of ResNet, DenseNet, SE-ResNext,and NasNet. The efficiency of proposed method tested on ISIC Archive and Dermnet.…”

“…The research articles [47], [48], [54], [55], [59], [61], [64], [68], [73], [75] and [78] In this review [67], [70], [72], [76] and [94] use this dataset to check the accuracy of their methods e) ISBI 2017 Dataset Challenge [37]: This dataset was used by [47], [50], [58], [59], [63], [64], [71] and [77] for testing their proposed methods. It contained 2,000 dermoscopic images in which 374 were melanomas, 254 were seborrheic keratoses, and 1,372 benign nevi images.…”

An Adaptive Federated Machine Learning-Based Intelligent System for Skin Disease Detection: A Step toward an Intelligent Dermoscopy Device

“…In [30], a hybrid approach was designed by using a combination of shallow learning-and deep learning-based pretrained models, such as AlexNet and support vector machine. A recent study [31] highlighted the enormous potential of deep understanding to detect skin diseases with human-like diagnosis accuracy or better. Furthermore, this study urged the utilization of deep learning-based real-time intelligent healthcare systems for clinical utilization.…”

Skin Lesion Classification Based on Deep Convolutional Neural Networks Architectures

“…The skin lesion is referred to as anomalies in skin appearances such as visible signs of sore, abnormal lump, or colored skin color. Physicians may use machine learning techniques to recognize and classify skin lesions in images before making decisions that would affect patients' health [8,9]. In general, there are four primary machine learning steps in the detection and diagnosis of melanoma cancer, comprising preprocessing of images, segmentation, feature extraction, and classification of images capturing the lesions (see Fig.…”