Novel Method to Establish Retinal Nerve Fiber Layer (RNFL) Thickness Measurement Comparability Between Time Domain Optical Coherence Tomography (TD-OCT) and Spectral Domain OCT (SD-OCT)

Abstract

Purpose: : TD-OCT has been commonly used in clinical practice, producing a large inventory of circular scan data. SD-OCT produces 3D data volumes. The purpose of this study was to create a robust technique that makes TD-OCT circular scans comparable with 3D SD-OCT volumes.

Methods: : Eleven eyes of 11 healthy subjects were assessed. Nine circumpapillary TD-OCT Fast RNFL Scans (Stratus OCT, Carl Zeiss Meditec, Dublin, CA) each set with 1 center and 8 displaced circles in known distance/directions, and one SD-OCT scan (Cirrus HD-OCT, Carl Zeiss Meditec, Dublin, CA; Optic Disc Cube 200x200) were obtained. The location of the TD-OCT circle scan was automatically detected within the SD-OCT volume scan using a software program of our own design. Mixed effect models were used to estimate the spatial deviation of the detected scan location on the SD-OCT 3D volume from the actual TD-OCT scan location to assess algorithm performance. Global and sectoral RNFL thickness measurements with or without scan location matching were also compared.

Results: : The difference [95% confidence interval] in scan circle center location between TD- and SD-OCT was 2.34 [1.24-3.44] pixels (70.20 [37.20-103.20] µm on the retina). There were significant differences in RNFL thickness measurements when SD-OCT/TD-OCT scan locations were not matched and no significant differences when they were matched, except for the superior quadrant and the 9, 12, and 6 o’clock sectors.

Conclusions: : Our novel method of location matching bridges the gap between line and volume scan data, providing follow-up comparability between TD- and SD-OCT RNFL measurements.

Keywords: imaging/image analysis: clinical • nerve fiber layer • image processing