Abstract
Purpose: :
To develop a method which allows automatically segmentation of the choroidal scleral interface (CSI) based on polarization sensitive optical coherence tomography (PS-OCT) data, in order to measure choroidal thickness.
Methods: :
We acquired data sets of healthy human volunteers with a novel swept source PS-OCT system working at a central wavelength of 1060 nm and an A-Scan rate of 100 kHz. High definition intensity, retardation and axis orientation B-scans were calculated, by using advanced averaging algorithms for PS-OCT data. RPE and CSI were automatically segmented in the high definition B-Scans using polarization sensitive data and in a further step choroidal thickness values were calculated. The RPE was segmented using degree of polarization uniformity (DOPU) values and for automatically segmenting the CSI a new algorithm was developed. Based on the fact that the sclera is birefringent while the choroid is not, the algorithm searches for retardation gradients in the region posterior the RPE.
Results: :
Several 2D measurements were performed in healthy human volunteers. A good reproducibility within each volunteer was shown, when comparing the segmentation results for RPE and CSI of different measurements. For the choroidal thickness values a thickness difference of approximately 120 µm between the nasal and the temporal areas was observed, which is in agreement with previous findings.
Conclusions: :
First tests in healthy human volunteers showed, that automated segmentation of the CSI using PS-OCT data is possible for high definition B-Scans. Several measurements within the same volunteer showed that the new algorithm is capable of producing a reproducible segmentation of the CSI. Choroidal thickness values were calculated based on the automated segmentation for RPE and CSI and showed good agreement with previous findings.
Keywords: choroid • retina • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)