人工智能在OCTA图像分析和眼部疾病诊断中的应用指南(2024)

Guidelines for the application of artificial intelligence in optical coherence tomography angiography image analysis and ocular disease diagnosis(2024)

光学相干断层扫描血管成像(OCTA)是一种无创成像技术,可提供三维、信息丰富的血管图像。大量研究表明,OCTA技术在影像生物标志物量化、诊断和监测方面具有独特的优势,因此在实验及临床研究中得到了迅速的应用。图像分析工具可快速、准确地量化OCTA的血管和病理特征,从而大大提高了OCTA成像的价值。近年来,人工智能(AI)已成为最强大的图像分析方法,特别是基于深度学习的图像分析可提供各种情况下的精确测量,包括不同的疾病和眼部区域。在此,中国医药教育协会眼科影像与智能医疗分会和国际转化医学会眼科专业委员会组织专家总结了国内外AI在OCTA图像分析和疾病诊断中的应用,其中包括脉络膜新生血管等病变的准确检测、视网膜灌注的精确量化以及可靠的疾病诊断,并分析目前面临的挑战和发展方向,经过多轮讨论和修改,形成了AI在OCTA图像分析和眼部疾病诊断中的应用指南,该指南旨在为临床提供新的见解和参考。

Optical coherence tomography angiography(OCTA)is a non-invasive imaging technique that provides three-dimensional,informative vascular images.Numerous studies have shown that OCTA technology has unique advantages in biomarker quantification,diagnosis and monitoring,and has therefore been rapidly applied in experiments and clinical studies.Image analysis tools can quickly and accurately quantify vascular and pathological features,greatly improving the value of OCTA imaging.In recent years,artificial intelligence(AI)has become the most powerful image analysis method,especially deep learning-based image analysis,which can provide accurate measurements in various situations,including different diseases and eye regions.The Ophthalmic Imaging and Intelligent Medicine Branch of Chinese Medicine Education Association and the Ophthalmology Committee of International Association of Translational Medicine designated experts to summarize the application of AI in OCTA image analysis and disease diagnosis at home and abroad,including accurate detection of choroidal neovascularization and other lesions,accurate quantification of retinal perfusion,and reliable disease diagnosis.They also analyzed the current challenges and development directions.After multiple rounds of discussion and revisions,they drafted the guidelines for the application of AI in OCTA image analysis and ocular disease diagnosis,aiming to provide new insights and references for clinical practice.