Bassi Arshpreet, Krance Saffire H, Pucchio Aidan, Pur Daiana R, Miranda Rafael N, Felfeli Tina
Schulich School of Medicine & Dentistry, Western University, London, Ontario, Canada.
School of Medicine, Queen's University, Kingston, Ontario, Canada.
Clin Ophthalmol. 2022 Aug 30;16:2895-2908. doi: 10.2147/OPTH.S377358. eCollection 2022.
This study aims to identify the available literature describing the utilization of artificial intelligence (AI) as a clinical tool in uveal diseases.
A comprehensive literature search was conducted in 5 electronic databases, finding studies relating to AI and uveal diseases.
After screening 10,258 studies,18 studies met the inclusion criteria. Uveal melanoma (44%) and uveitis (56%) were the two uveal diseases examined. Ten studies (56%) used complex AI, while 13 studies (72%) used regression methods. Lactate dehydrogenase (LDH), found in 50% of studies concerning uveal melanoma, was the only biomarker that overlapped in multiple studies. However, 94% of studies highlighted that the biomarkers of interest were significant.
This study highlights the value of using complex and simple AI tools as a clinical tool in uveal diseases. Particularly, complex AI methods can be used to weigh the merit of significant biomarkers, such as LDH, in order to create staging tools and predict treatment outcomes.
本研究旨在识别描述人工智能（AI）作为葡萄膜疾病临床工具应用的现有文献。
在5个电子数据库中进行了全面的文献检索，查找与AI和葡萄膜疾病相关的研究。
在筛选了10258项研究后，18项研究符合纳入标准。所研究的两种葡萄膜疾病为葡萄膜黑色素瘤（44%）和葡萄膜炎（56%）。10项研究（56%）使用了复杂AI，而13项研究（72%）使用了回归方法。乳酸脱氢酶（LDH）在50%的葡萄膜黑色素瘤相关研究中出现，是多项研究中唯一重叠的生物标志物。然而，94%的研究强调感兴趣的生物标志物具有重要意义。
本研究突出了使用复杂和简单AI工具作为葡萄膜疾病临床工具的价值。特别是，复杂AI方法可用于权衡重要生物标志物（如LDH）的价值，以创建分期工具并预测治疗结果。
