Purchase this article with an account.
Roman V. Kuranov, Ian Rubinoff, Yuanbo Wang, Lisa Beckmann, Xian Zhang, Amani A Fawzi, Hiroshi Ishikawa, Joel Schuman, Hao F Zhang; Enhanced imaging of the outer retina layers using speckle-reduced visible-light optical coherence tomography. Invest. Ophthalmol. Vis. Sci. 2019;60(9):1297. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Non-invasive delineation of the outer retina layers including Bruch’s membrane, RPE and ellipsoid (IS/OS junction) is beneficial for many ocular diseases such as AMD, retinal drug toxicity and Central Serous Chorioretinopathy (CSC). It is also beneficial for studying mouse models of ocular diseases, where retinal layers are even harder to delineate. However, achieving this goal remains challenging.Commercial and research near infrared optical coherence tomography (NIR-OCT) do not possess the micron resolution and high contrast required to identify fine details of the outer retina layers. Although visible light (vis-) OCT has the theoretical resolution and contrast for such imaging, in practice it is limited by speckle noise. To this end, we developed a simple, motion-insensitive speckle reduction algorithm using vis-OCT.
Here we developed a speckle-reduction method for vis-OCT based on the deflection of the scanning beam to the direction orthogonal to the frame acquisition line to acquire uncorrelated A-scans for speckle averaging. The increased efficiency of the scanning pattern allowed us to reduce the optimal deflection amplitude and thus mitigate distortion of the image. Additionally, we improved optimization metric to identify optimal deflection amplitude. We tested our method by imaging human volunteers and wild-type mice.
Using a mouse eye phantom we determined optimal deflection amplitude in the range of 50 µm-75 µm. We also reduced the number of independent A-scans by factor of 2 required for optimal speckle mitigation. The improvement of the quality of the images was estimated using contrast to noise ratio (CNR), equivalent number of looks (ENL) and structural similarity metric (SSIM). We were able to significantly improve quality of the images and delineate all retinal layers in both mouse (Figure 1) and human (Figure 2) subjects.
We developed and tested in-vivo a new non-invasive imaging tool based on vis-OCT that has unprecedented resolution and contrast for highly-detailed imaging of the inner and outer retina. We believe that speckle-reduced vis-OCT will have significant impact on ophthalmic research and clinical applications.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
Fig. 1. Conventional (a) and speckle reduced (b) images of the mice retina.
Fig. 2. Conventional (a) and speckle reduced (b) images of the human retina. Here COS and ROS are cone and rod outer segments respectively.
This PDF is available to Subscribers Only