Purpose:
To determine whether the density of B-scans in a spectral domain OCT volume/cube scan can be reduced without reducing the accuracy of retinal boundary segmentation and retinal thickness maps.
Methods:
Raw OCT data was exported from consecutive patients with retinal disease who underwent imaging using the Cirrus HD-OCT (version 4.0 software) and 512 x 128 Macular Cube protocol. OCT scans were reviewed by certified Doheny Image Reading Center Graders to identify 58 cases in which accurate automated retinal boundary segmentation was observed on all 128 scans. These cases were then re-processed to sequentially remove B-scans (removing every other scan, then every third, and so forth to a density of only every 16th scan) and both the segmentation boundaries and retinal thickness maps were regenerated. Retinal thickness measurements derived using only a subset of scans were compared to measurements using all 128 B-scans, and differences for the foveal central subfield (FCS) thickness and total macular volume were computed.
Results:
The mean absolute error in FCS retinal thickness measurement increased as the density of B-scans decreased, though the error remained less than 20 microns when only 32 (1 of 4) B-scans were used (FIGURE). Mean absolute error in retinal volume demonstrated a similar trend, with an error of less than 0.3 ml until fewer than 32 B-scans were used.
Conclusions:
Retinal thickness measurements may change as B-scan density is reduced in volume SDOCT acquisitions, though on average, dramatic changes are not observed until B-scan density is reduced below 32 (every fourth) B-scans. This information may be of value in design of scanning protocols for Cirrus OCT for use in future clinical trials.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • imaging/image analysis: clinical • retina