研究目的
To determine whether a qualitative approach toward evaluating optical coherence tomography (OCT) imaging improves the ability to detect glaucomatous damage compared to a conventional metric of global circumpapillary retinal nerve fiber layer (cpRNFL) thickness.
研究成果
Qualitative evaluation of OCT imaging results is more accurate than global cpRNFL thickness for detecting glaucomatous damage, with high sensitivity and specificity. It leverages comprehensive OCT data and is feasible for clinical use, though further validation is needed for broader applicability.
研究不足
An observer with extensive experience performed all grading, which may not be generalizable to less experienced clinicians. The study did not include eyes with conditions like high myopia or macrodiscs, and interobserver agreement was not fully established.
1:Experimental Design and Method Selection:
The study used a diagnostic accuracy design to compare qualitative evaluation of OCT imaging with a conventional metric (global cpRNFL thickness) for detecting glaucomatous damage. Receiver operating characteristic (ROC) curves and statistical tests (Wald test with bootstrap resampling) were employed to assess performance.
2:Sample Selection and Data Sources:
Participants included 394 healthy eyes and 272 glaucoma eyes (156 perimetric glaucoma, 116 suspected glaucoma) from a prospective study. Healthy participants were from a reference dataset provided by Topcon, Inc. Eyes were categorized based on visual field test history using a Humphrey Field Analyzer.
3:List of Experimental Equipment and Materials:
Spectral-domain OCT device (3D OCT-2000; Topcon, Inc.), visual field analyzer (Humphrey Field Analyzer II-I; Carl Zeiss Meditec, Inc.), customized one-page reports generated from OCT scans.
4:Experimental Procedures and Operational Workflow:
OCT volume scans of the optic disc and macula were obtained. Customized reports were created and qualitatively graded by a masked specialist for the probability of optic neuropathy. Visual field tests were conducted and analyzed for reliability and abnormalities.
5:Data Analysis Methods:
Statistical analysis included ROC curve generation, sensitivity and specificity calculations at 95% specificity, and comparison using bootstrap resampling (1000 resamples).
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容