Department of Dermatology, University of California, Irvine, 1 Medical Plaza Drive, Irvine, CA, 92617, USA.
Arch Dermatol Res. 2022 May;314(4):325-327. doi: 10.1007/s00403-021-02236-9. Epub 2021 May 15.
Machine learning (ML) has been increasingly utilized for skin cancer screening, primarily of melanomas but also of non-melanoma skin cancers (NMSC).
This study presents the first quantitative review of the success of these techniques in NMSC screening.
A primary literature search was conducted using PubMed, MEDLINE, and arXiv, capturing all articles involving ML techniques and NMSC screening.
52 articles were included for quantitative analysis, resulting in a mean sensitivity of 89.2% (n = 52, 95% confidence interval (CI) 87.0-91.3) and a mean specificity of 81.1% (n = 44, 95% CI 74.5-87.8) for ML algorithms in the diagnosis of NMSC. Studies were further grouped by skin cancer type, algorithm type, diagnostic gold standard, data set source, and data set size.
There is insufficient evidence to conclude that an ML algorithm is superior at NMSC screening than a trained dermatologist utilizing dermoscopy for either BCC or SCC. Given that the studies included in this review were performed in silico, further study in the form of randomized clinical trials are needed to further elucidate the role of NMSC screening algorithms in dermatology.
机器学习（ML）已越来越多地用于皮肤癌筛查，主要用于黑色素瘤，但也用于非黑色素瘤皮肤癌（NMSC）。
本研究首次对这些技术在 NMSC 筛查中的成功进行了定量评估。
使用 PubMed、MEDLINE 和 arXiv 进行了初步文献检索，捕获了所有涉及 ML 技术和 NMSC 筛查的文章。
对 52 篇文章进行了定量分析，结果显示，用于 NMSC 诊断的 ML 算法的平均敏感性为 89.2%（n=52，95%置信区间[CI]为 87.0-91.3），特异性为 81.1%（n=44，95%CI 为 74.5-87.8）。研究进一步按皮肤癌类型、算法类型、诊断金标准、数据集来源和数据集大小进行分组。
没有足够的证据表明 ML 算法在 NMSC 筛查方面优于利用皮肤镜检查的训练有素的皮肤科医生，无论是用于 BCC 还是 SCC。鉴于本综述中包含的研究是在计算机上进行的，需要进一步进行随机临床试验研究，以进一步阐明 NMSC 筛查算法在皮肤科中的作用。
