Abstract
Purpose: :
Spectral-domain optical coherence tomography (SD-OCT) has several advantages over time-domain OCT (TD-OCT): higher resolution, ability to sample a larger retinal area in shorter time, post acquisition editing of the scan location, and precise registration of the scan locations for further follow up in subsequent visits. The present study describes the correlation between TD-OCT and SD-OCT measurements in normal eyes and in eyes with diabetic macular edema (DME)
Methods: :
Eighty eyes (40 subjects) with no known ocular diseases and 38 eyes with DME (23 subjects) were scanned by Stratus® (TD-OCT) and Spectralis® (SD-OCT) apparatuses. Each eye was scanned on each OCT machine, 5 minutes apart by the same operator. Normal eyes were scanned twice for reproducibility analysis. Measurements were taken from each of five subfields: central (C), nasal (N), temporal (T), superior (S), and inferior (I). Statistical analyses were calculated for each subfield measurements in each study group. A correlative equation was derived and used to predict SD-OCT values from TD-OCT measurements. The relationship between the predicted and actual values was analyzed.
Results: :
The mean subfield measurements among normal eyes were 188µm (C), 266 (N), 267 (T), 255 (S), and 266 (I), for TD-OCT, compared to 264µm (C), 340 (N), 340 (T), 327 (S), and 337 (I) for SD-OCT. The correlation coefficient (r) between TD and SD values in different subfields ranged between 0.89 and 0.95 in normal eyes and between 0.90 and 0.96 in eyes with DME. The reproducibility parameters were 3.45µm (±4.9) for TD-OCT compared to 1.36µm (±1.35) for SD-OCT. The predicted SD-OCT values, derived according to the polynomial equation Y= -0.0005X2 +1.21X +52.69 (where Y = SD-OCT value and X = TD-OCT measurement) correlated strongly to the actual values, with r = 0.95 and a difference in the means of 6.2µm.
Conclusions: :
Intra-operator reproducibility among healthy eyes is high in both types of OCT, with Spectralis® yielding nearly no difference in repeated measurements. Retinal thickness measurements from both apparatuses can be converted interchangeably with high degree of accuracy in normal eyes and in eyes with DME provided the scans are taken in precisely corresponding locations. Regression equation generated from this study can provide appropriate comparison of acquired data in standard care and clinical trial assessment
Keywords: imaging/image analysis: clinical • diabetic retinopathy • clinical (human) or epidemiologic studies: systems/equipment/techniques