Generation of a Melanoma and Nevus Data Set From Unstandardized Clinical Photographs on the Internet

Abstract

ImportanceArtificial intelligence (AI) training for diagnosing dermatologic images requires large amounts of clean data. Dermatologic images have different compositions, and many are inaccessible due to privacy concerns, which hinder the development of AI.ObjectiveTo build a training data set for discriminative and generative AI from unstandardized internet images of melanoma and nevus.Design, Setting, and ParticipantsIn this diagnostic study, a total of 5619 (CAN5600 data set) and 2006 (CAN2000 data set; a manually revised subset of CAN5600) cropped lesion images of either melanoma or nevus were semiautomatically annotated from approximately 500 000 photographs on the internet using convolutional neural networks (CNNs), region-based CNNs, and large mask inpainting. For unsupervised pretraining, 132 673 possible lesions (LESION130k data set) were also created with diversity by collecting images from 18 482 websites in approximately 80 countries. A total of 5000 synthetic images (GAN5000 data set) were generated using the generative adversarial network (StyleGAN2-ADA; training, CAN2000 data set; pretraining, LESION130k data set).Main Outcomes and MeasuresThe area under the receiver operating characteristic curve (AUROC) for determining malignant neoplasms was analyzed. In each test, 1 of the 7 preexisting public data sets (total of 2312 images; including Edinburgh, an SNU subset, Asan test, Waterloo, 7-point criteria evaluation, PAD-UFES-20, and MED-NODE) was used as the test data set. Subsequently, a comparative study was conducted between the performance of the EfficientNet Lite0 CNN on the proposed data set and that trained on the remaining 6 preexisting data sets.ResultsThe EfficientNet Lite0 CNN trained on the annotated or synthetic images achieved higher or equivalent mean (SD) AUROCs to the EfficientNet Lite0 trained using the pathologically confirmed public data sets, including CAN5600 (0.874 [0.042]; P = .02), CAN2000 (0.848 [0.027]; P = .08), and GAN5000 (0.838 [0.040]; P = .31 [Wilcoxon signed rank test]) and the preexisting data sets combined (0.809 [0.063]) by the benefits of increased size of the training data set.Conclusions and RelevanceThe synthetic data set in this diagnostic study was created using various AI technologies from internet images. A neural network trained on the created data set (CAN5600) performed better than the same network trained on preexisting data sets combined. Both the annotated (CAN5600 and LESION130k) and synthetic (GAN5000) data sets could be shared for AI training and consensus between physicians.

JAMA Dermatology—The Year in Review, 2023

“…Vodrahalli et al 11 highlighted the capacity of a machine-learning algorithm trained on retrospective telemedicine images to identify poor-quality images and a clinical pilot study of a patient-facing application that was associated with a significantly reduced number of poor-quality images submitted by patients to their dermatologists. Cho et al 12 report the use of several AI modalities, including generative AI, to build comprehensive image datasets for dermatologic education. The application of generative AI to enhance educational materials has powerful, complex implications for our image-rich specialty.…”

High-Fidelity Synthetic Face Generation for Rosacea Skin Condition from Limited Data

“…Although there have been a few studies [27][28][29][30][31][32] on generating synthetic images of skin cancer lesions using various types of GANs architecture, the images were captured through a dermatoscope and other imaging devices that focus only on a specific locality i.e., cancerous regions of the skin. Carrasco et al [33] and Cho et al [34] explored the generation of cancerous skin lesion images using the StyleGAN2-ADA architecture. Carrasco et al [33] employed a substantial dataset comprising 37,648 images in both conditional and unconditional settings.…”

“…Carrasco et al [33] employed a substantial dataset comprising 37,648 images in both conditional and unconditional settings. On the other hand, Cho et al [34] focused on creating a melanocytic lesion dataset using non-standardized Internet images, annotating approximately 500,000 photographs to develop a diverse and extensive dataset.…”

“…The main objective of Cho et al's [34] study was to explore the possibility of image generation using images scrapped from various online sources where data are not structured. They created a diverse LESION130k dataset of potential lesions and generated 5000 synthetic images with StyleGAN2-ADA, illustrating the potential of AI in diversifying medical image datasets from various sources.…”