Purchase this article with an account.
Scott Walter, Dilraj Singh Grewal, David Cunefare, Paramjit K Bhullar, Sina Farsiu, Tamer H Mahmoud; Topography of the Vitreoretinal Interface in Vitreomacular Traction Syndrome. Invest. Ophthalmol. Vis. Sci. 2016;57(12):4078.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate vitreoretinal interface topography changes in vitreomacular traction (VMT) syndrome, using radial and raster high-density optical coherence tomography (OCT) scan protocols.
Longitudinal, observational study of patients with VMT followed with two OCT scanning protocols: a 3x3mm 24-line radial scan and a 6x6mm 61-line raster scan. Manual segmentation of the hyaloid and automated segmentation of the internal limiting membrane (ILM), retinal pigment epithelium (RPE), and Bruch’s membrane was performed using the Duke Optical Coherence Tomography Retinal Analysis Program (Figure 1). Scans were excluded if the hyaloid was not visualized on ≥3 consecutive B-scans. The primary outcome measures were volume of the subhyaloid space and surface area of vitreomacular adhesion (VMA), measured within 1.5mm of the fovea.
Thirty-two radial and 30 raster scans from 13 eyes of 10 patients were graded providing for 15 longitudinal comparisons with an average inter-scan interval of 7.9 months (range 1-18 months). Manual segmentation of the hyaloid was possible in over 99% (1236/1248) of individual B-scans, and hyaloid elevation was interpolated from adjacent B-scans in the remaining 1%. Volume of the subhyaloid space correlated with decreasing VMA area (r2= -0.5, p=0.0001), average retinal thickness within the area of VMA (r2=0.38, p=0.006), and maximal elevation of the RPE above Bruch’s membrane (r2=0.28, p=0.04). Both primary outcome measures were highly correlated (r2 >0.99, p<0.00001) for concurrent radial and raster scans. However, raster scans failed to detect VMA in 10% of instances where it was detected on radial scans. Radial scanning also provided finer topographic detail of the vitreomacular interface (Figure 2). Volume of the subhyaloid space increased at an average rate of 0.003mm3 per month, and surface area of VMA decreased at -0.006mm2 per month. There was an approximately 4-fold difference in the magnitude of changes observed in VMA area with radial versus raster scanning (-0.045 vs. -0.011mm2; p=0.06).
Hyaloid segmentation of OCT images, combined with automated retinal segmentation algorithms, allows precise monitoring of vitreoretinal interface topography in VMT syndrome. Radial scans provide finer topographic detail and appear to be more sensitive for the detection of VMA and changes in VMA area over time.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
Segmentation of the hyaloid and retinal layers.
Hyaloid-ILM volume maps
This PDF is available to Subscribers Only