Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network

Abstract

Although there have been reports of the successful diagnosis of skin disorders using deep learning, unrealistically large clinical image datasets are required for artificial intelligence (AI) training. We created datasets of standardized nail images using a region-based convolutional neural network (R-CNN) trained to distinguish the nail from the background.We used R-CNN to generate training datasets of 49,567 images, which we then used to fine-tune the ResNet-152 and VGG-19 models. The validation datasets comprised 100 and 194 images from Inje University (B1 and B2 datasets, respectively), 125 images from Hallym University (C dataset), and 939 images from Seoul National University (D dataset).The AI (ensemble model; ResNet-152 + VGG-19 + feedforward neural networks) results showed test sensitivity/specificity/ area under the curve values of (96.0 / 94.7 / 0.98), (82.7 / 96.7 / 0.95), (92.3 / 79.3 / 0.93), (87.7 / 69.3 / 0.82) for the B1, B2, C, and D datasets.With a combination of the B1 and C datasets, the AI Youden index was significantly (p = 0.01) higher than that of 42 dermatologists doing the same assessment manually. For B1+C and B2+ D dataset combinations, almost none of the dermatologists performed as well as the AI. By training with a dataset comprising 49,567 images, we achieved a diagnostic accuracy for onychomycosis using deep learning that was superior to that of most of the dermatologists who participated in this study.

Machine Learning in Dermatology: Current Applications, Opportunities, and Limitations

“…Again, the combined performance was on par with dermatologists and outperformed non-dermatologists. R-CNNs have been used in fracture detection in radiology [55] and the detection of the nail plate in onychomycosis [56]. Algorithms could potentially detect and diagnosis skin cancer without any preselection of suspicious lesions by dermatologists.…”

“…This is a notable limitation considering that the incidence of scalp psoriasis is 45-56% and nail psoriasis is 23-27% among psoriatic patients [57]. CNNs have been created for other diseases, such as atopic dermatitis [58], onychomycosis [56], and rosacea [59]. To classify onychomycosis, Han et al [56] used a R-CNN to generate a training datasets of 49,567 images of nails and found that a combination of their datasets performed better than dermatologists.…”

Research Techniques Made Simple: Deep Learning for the Classification of Dermatological Images

“…All of these works have reported the equivalence between computer and human diagnosis. Besides skin cancer detection, DL is also being successfully applied to other areas of dermatology, such as the monitoring of wound healing (Shenoy et al, 2018), the classification of ulcers (Goyal et al, 2018), and onychomycosis (Han et al, 2018b).…”

Lesion-Based Convolutional Neural Network in Diagnosis of Early Gastric Cancer

“…9 Since then, numerous authors have studied medical image analysis using CNN with deep learning method. [12][13][14] Most examinations and diagnoses of GI tract diseases are performed using endoscopy. 1,15 Therefore, accurate analysis of endoscopic images is important, and the application of CNN can be considerably useful.…”