Purpose:
To visualize comprehensively the retinal capillary network in 3D using high speed optical coherence tomography with high lateral resolution and to analyze the network for healthy and diseased cases.
Methods:
We employ CMOS equipped ultra-high speed OCT that allows for imaging speeds reaching 160.000 A-scans/sec. A tdensely sampled volume of 500x500x768 pixels is recorded in only 1.5 sec, minimizing transverse motion blurring and preserving the integrity of small structural details. We combine such ultra-high speed system with broadband light source, to achieve as well ultra high axial resolution (1.5 µm in tissue). The capillary network in the parafoveal region is imaged of healthy and diseased subjects using a high-density sampling. We apply then refined image-processing algorithms, based on a probabilistic kernel, in order to segment the capillary network. This gives us the capability to quantify the amount of micro-vascularization per volume for the tissue surrounding the fovea. Having the segmented capillary network at hand, we also employ fractional dimension analysis using a standard sandbox algorithm.
Results:
The capillary network as well as the foveal avasculature zone of 3 healthy volunteers has been segmented and the vascular density and the fractional dimension quantitatively analyzed. Since those results are not based on Doppler analysis of phase resolved OCT techniques they are less sensitive to motion artifacts, and in particular are independent of otherwise strict scanning and sampling issues. Moreover they yield high resolution 3D capillary images with a quality, that has not yet been achieved with any other technique.
Conclusions:
Comprehensive 3D images of capillary patterns are the basis for analysis of vessel density and vessel tree structure using fractional dimensionality. Since retinal diseases will manifest at an early stage in capillary disorders we believe that such in-vivo studies of capillary details will help for a better understanding as well as early diagnosis of such major retinal diseases
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • blood supply • retina