Retinal layer thicknesses retrieved with different segmentation algorithms from optical coherence tomography scans acquired under different signal-to-noise ratio conditions

Abstract

Glaucomatous damage can be quantified by measuring the thickness of different retinal layers. However, poor image quality may hamper the accuracy of the layer thickness measurement. We determined the effect of poor image quality (low signal-to-noise ratio) on the different layer thicknesses and compared different segmentation algorithms regarding their robustness against this degrading effect. For this purpose, we performed OCT measurements in the macular area of healthy subjects and degraded the image quality by employing neutral density filters. We also analysed OCT scans from glaucoma patients with different disease severity. The algorithms used were: The Canon HS-100’s built-in algorithm, DOCTRAP, IOWA, and FWHM, an approach we developed. We showed that the four algorithms used were all susceptible to noise at a varying degree, depending on the retinal layer assessed, and the results between different algorithms were not interchangeable. The algorithms also differed in their ability to differentiate between young healthy eyes and older glaucoma eyes and failed to accurately separate different glaucoma stages from each other.

U-Shaped Effect of Blood Pressure on Structural OCT Metrics and Retinal Perfusion in Ophthalmologically Healthy Subjects

“…1 A), a region that has been shown to be advantageous over the commonly used 5-mm diameter ROI for mRNFL measurements. 37 It also reports peripapillary RNFL (pRNFL) thickness at a 3.45-mm diameter circle centered at the ONH ( Fig. 1 B).…”

U-shaped effect of blood pressure on structural OCT metrics and retinal blood flow autoregulation in ophthalmologically healthy subjects

“…The device automatically segments and quantifies macular RNFL (mRNFL) thickness and ganglion cell-inner plexiform layer (GCIPL) within a 10-mm diameter circular region of interest (ROI) centered at the fovea ( Figure 1A), a region which has been shown to be advantageous over the commonly used 5-mm diameter ROI for mRNFL measurements. 37 It also reports peripapillary RNFL (pRNFL) thickness at a 3.45-mm diameter circle centered at the ONH ( Figure 1B). We further subdivided pRNFL into temporal, superior, nasal, and inferior.…”

Influence of signal‐to‐noise ratio, glaucoma stage and segmentation algorithm on OCT usability for quantifying layer thicknesses in the peripapillary retina

“…First, we (back) transformed the logarithmically compressed intensity values; for this, we assumed that the 0–255 grayscale values in the 8‐bit image correspond to a 50 dB dynamic range in the (squared) spectrometer values, with the (squared) spectrometer output being linearly scaled. (Heikka, Cense, & Jansonius, 2020; Heikka, Ometto, et al, 2020) The B‐scan image is shown in Figure 1b; Figure 1c presents the corresponding A‐scans from the two locations marked in Figure 1b, based on both the greyscale values from the B‐scan image (blue) and the linearly scaled data (black). The RNFLT was calculated as the full width half maximum (FWHM) of the first peak in the linearly scaled A‐scan (marked with horizontal red line in Figure 1c).…”

“…Our previous work described the FWHM algorithm for segmenting the macular area. (Heikka, Cense, & Jansonius, 2020) Shortly, we developed the FWHM algorithm using MATLAB (Mathworks, version R2018a, Natick, Massachusetts: The MathWorks Inc.). First, we (back) transformed the logarithmically compressed intensity values; for this, we assumed that the 0-255 grayscale values in the 8-bit image correspond to a 50 dB dynamic range in the (squared) spectrometer values, with the (squared) spectrometer output being linearly scaled.…”

“…(Jansonius et al, 2016; de Kinkelder et al, 2013) Layer thickness estimations also vary between different OCT manufacturers (de Kinkelder et al, 2013), primarily due to differences in segmentation approaches rather than differences in hardware. (Heikka, Cense, & Jansonius, 2020; Heikka, Ometto, et al, 2020) In an earlier study, we focused on the influence of SNR on retinal layer thickness assessment with different algorithms in the macular area. (Heikka, Ometto, et al, 2020) In the current study, we focus on the peripapillary area.…”