Cancer-Net SCa-Synth: An Open Access Synthetically Generated 2D Skin Lesion Dataset for Skin Cancer Classification

Chi-en Amy Tai,Oustan Ding,Alexander Wong

2024-11-08

Abstract:In the United States, skin cancer ranks as the most commonly diagnosed cancer, presenting a significant public health issue due to its high rates of occurrence and the risk of serious complications if not caught early. Recent advancements in dataset curation and deep learning have shown promise in quick and accurate detection of skin cancer. However, current open-source datasets have significant class imbalances which impedes the effectiveness of these deep learning models. In healthcare, generative artificial intelligence (AI) models have been employed to create synthetic data, addressing data imbalance in datasets by augmenting underrepresented classes and enhancing the overall quality and performance of machine learning models. In this paper, we build on top of previous work by leveraging new advancements in generative AI, notably Stable Diffusion and DreamBooth. We introduce Cancer-Net SCa-Synth, an open access synthetically generated 2D skin lesion dataset for skin cancer classification. Further analysis on the data effectiveness by comparing the ISIC 2020 test set performance for training with and without these synthetic images for a simple model highlights the benefits of leveraging synthetic data to improve performance. Cancer-Net SCa-Synth is publicly available at <a class="link-external link-https" href="https://github.com/catai9/Cancer-Net-SCa-Synth" rel="external noopener nofollow">this https URL</a> as part of a global open-source initiative for accelerating machine learning for cancer care.

Computer Vision and Pattern Recognition,Machine Learning

What problem does this paper attempt to address?

This paper aims to solve the problem of dataset class imbalance in skin cancer detection. The currently publicly available datasets have a significant class imbalance when training deep - learning models, which hinders the effectiveness and performance of the models. To this end, the researchers used generative artificial intelligence techniques (specifically Stable Diffusion and DreamBooth) to create a synthetic 2D skin lesion image dataset named Cancer - Net SCa - Synth to improve model performance in skin cancer classification tasks. By comparing the performance on the ISIC 2020 test set, the positive effect of using synthetic data on improving model performance is demonstrated. The Cancer - Net SCa - Synth dataset contains 10,000 images, with 5,000 benign lesions and 5,000 malignant melanomas each, achieving a balanced distribution of classes. The research results show that combined training with the Cancer - Net SCa - Synth and ISIC 2020 training sets can significantly improve the model's performance on private and public scores, increasing the AUC values by more than 0.04 and 0.09 respectively.