Optical coherence tomography angiography in neovascular age-related macular degeneration: comprehensive review of advancements and future perspective

Eye (Lond). 2024 Aug 15. doi: 10.1038/s41433-024-03295-8. Online ahead of print.

Abstract

Optical coherence tomography angiography (OCTA) holds promise in enhancing the care of various retinal vascular diseases, including neovascular age-related macular degeneration (nAMD). Given nAMD's vascular nature and the distinct vasculature of macular neovascularization (MNV), detailed analysis is expected to gain significance. Research in artificial intelligence (AI) indicates that en-face OCTA views may offer superior predictive capabilities than spectral domain optical coherence tomography (SD-OCT) images, highlighting the necessity to identify key vascular parameters. Analyzing vasculature could facilitate distinguishing MNV subtypes and refining diagnosis. Future studies correlating OCTA parameters with clinical data might prompt a revised classification system. However, the combined utilization of qualitative and quantitative OCTA biomarkers to enhance the accuracy of diagnosing disease activity remains underdeveloped. Discrepancies persist regarding the optimal biomarker for indicating an active lesion, warranting comprehensive prospective studies for validation. AI holds potential in extracting valuable insights from the vast datasets within OCTA, enabling researchers and clinicians to fully exploit its OCTA imaging capabilities. Nevertheless, challenges pertaining to data quantity and quality pose significant obstacles to AI advancement in this field. As OCTA gains traction in clinical practice and data volume increases, AI-driven analysis is expected to further augment diagnostic capabilities.

摘要: 相干光断层扫描血管造影 (OCTA) 有望提升各种视网膜血管疾病的治疗效果, 包括新生血管性老年性黄斑变性 (nAMD) 。鉴于 nAMD 的新生血管特性和黄斑新生血管 (MNV) 独特的血管结构, 进行详细分析至关重要。人工智能 (AI) 研究表明, 相比于谱域相干光断层扫描 (SD-OCT) 图像, en-face OCTA 图像具有更强的预测能力, 强调了识别关键血管参数的必要性。血管分析有助于区分 MNV 亚型及完善诊断。联合 OCTA 参数与临床数据的研究可促使修正MNV分类系统的构建。联合使用定性及定量 的OCTA 生物标志物来提高诊断疾病活动性的准确性仍有待发展。对于提示病变的活动性的最佳生物标志物, 目前仍有争议, 需进行前瞻性研究来证实。 人工智能有望从 OCTA 的庞大数据库中提取有价值的信息, 使研究人员和临床医生能够充分利用其 OCTA 成像功能。然而, 数据数量和质量方面的挑战对人工智能在这一领域的发展具有阻碍。随着 OCTA 在临床实践的应用和数据量的增加, 人工智能驱动的分析有望进一步提高其诊断能力。.

Publication types

  • Review