Modular deep neural networks for automatic quality control of retinal optical coherence tomography scans.

Retinal optical coherence tomography (OCT) with intraretinal layer segmentation is increasingly used not only in ophthalmology but also for neurological diseases such as multiple sclerosis (MS). Signal quality influences segmentation results, and high-quality OCT images are needed for accurate segmentation and quantification of subtle intraretinal layer changes. Among others, OCT image quality depends on the ability to focus, patient compliance and operator skills. Current criteria for OCT quality define acceptable image quality, but depend on manual rating by experienced graders and are time consuming and subjective. In this paper, we propose and validate a standardized, grader-independent, real-time feedback system for automatic quality assessment of retinal OCT images. We defined image quality criteria for scan centering, signal quality and image completeness based on published quality criteria and typical artifacts identified by experienced graders when inspecting OCT images. We then trained modular neural networks on OCT data with manual quality grading to analyze image quality features. Quality analysis by a combination of these trained networks generates a comprehensive quality report containing quantitative results. We validated the approach against quality assessment according to the OSCAR-IB criteria by an experienced grader. Here, 100 OCT files with volume, circular and radial scans, centered on optic nerve head and macula, were analyzed and classified. A specificity of 0.96, a sensitivity of 0.97 and an accuracy of 0.97 as well as a Matthews correlation coefficient of 0.93 indicate a high rate of correct classification. Our method shows promising results in comparison to manual OCT grading and may be useful for real-time image quality analysis or analysis of large data sets, supporting standardized application of image quality criteria.

Normative Data and Minimally Detectable Change for Inner Retinal Layer Thicknesses Using a Semi-automated OCT Image Segmentation Pipeline.

Neurodegenerative and neuroinflammatory diseases regularly cause optic nerve and retinal damage. Evaluating retinal changes using optical coherence tomography (OCT) in diseases like multiple sclerosis has thus become increasingly relevant. However, intraretinal segmentation, a necessary step for interpreting retinal changes in the context of these diseases, is not standardized and often requires manual correction. Here we present a semi-automatic intraretinal layer segmentation pipeline and establish normative values for retinal layer thicknesses at the macula, including dependencies on age, sex, and refractive error. Spectral domain OCT macular 3D volume scans were obtained from healthy participants using a Heidelberg Engineering Spectralis OCT. A semi-automated segmentation tool (SAMIRIX) based on an interchangeable third-party segmentation algorithm was developed and employed for segmentation, correction, and thickness computation of intraretinal layers. Normative data is reported from a 6 mm Early Treatment Diabetic Retinopathy Study (ETDRS) circle around the fovea. An interactive toolbox for the normative database allows surveying for additional normative data. We cross-sectionally evaluated data from 218 healthy volunteers (144 females/74 males, age 36.5 ± 12.3 years, range 18-69 years). Average macular thickness (MT) was 313.70 ± 12.02 µm, macular retinal nerve fiber layer thickness (mRNFL) 39.53 ± 3.57 µm, ganglion cell and inner plexiform layer thickness (GCIPL) 70.81 ± 4.87 µm, and inner nuclear layer thickness (INL) 35.93 ± 2.34 µm. All retinal layer thicknesses decreased with age. MT and GCIPL were associated with sex, with males showing higher thicknesses. Layer thicknesses were also positively associated with each other. Repeated-measurement reliability for the manual correction of automatic intraretinal segmentation results was excellent, with an intra-class correlation coefficient >0.99 for all layers. The SAMIRIX toolbox can simplify intraretinal segmentation in research applications, and the normative data application may serve as an expandable reference for studies, in which normative data cannot be otherwise obtained.

Optic nerve head three-dimensional shape analysis.

We present a method for optic nerve head (ONH) 3-D shape analysis from retinal optical coherence tomography (OCT). The possibility to noninvasively acquire in vivo high-resolution 3-D volumes of the ONH using spectral domain OCT drives the need to develop tools that quantify the shape of this structure and extract information for clinical applications. The presented method automatically generates a 3-D ONH model and then allows the computation of several 3-D parameters describing the ONH. The method starts with a high-resolution OCT volume scan as input. From this scan, the model-defining inner limiting membrane (ILM) as inner surface and the retinal pigment epithelium as outer surface are segmented, and the Bruch's membrane opening (BMO) as the model origin is detected. Based on the generated ONH model by triangulated 3-D surface reconstruction, different parameters (areas, volumes, annular surface ring, minimum distances) of different ONH regions can then be computed. Additionally, the bending energy (roughness) in the BMO region on the ILM surface and 3-D BMO-MRW surface area are computed. We show that our method is reliable and robust across a large variety of ONH topologies (specific to this structure) and present a first clinical application.

Temporal visual resolution and disease severity in MS.

OBJECTIVE: To examine temporal visual resolution assessed as critical flicker frequency (CFF) in patients with MS and to investigate associations with visual system damage and general disability and cognitive function. METHODS: Thirty-nine patients with MS and 31 healthy controls (HCs) were enrolled in this cross-sectional study and underwent CFF testing, high- and low-contrast visual acuity, alertness and information processing speed using the paced auditory serial addition task (PASAT), and retinal optical coherence tomography (OCT). In patients with MS, visual evoked potentials (VEPs) and Expanded Disability Status Scale (EDSS) scores were assessed. RESULTS: CFF in patients with MS (mean ± SD: 40.9 ± 4.4 Hz) was lower than in HCs (44.8 ± 4.4 Hz, p < 0.001). There was no significant CFF difference between eyes with and without previous optic neuritis (ON). CFF was not associated with visual acuity, VEP latency, the peripapillary retinal nerve fiber layer thickness, and the combined ganglion cell and inner plexiform layer volume. Instead, reduced CFF was associated with worse EDSS scores (r2 = 0.26, p < 0.001) and alertness (r2 = 0.42, p = 0.00042) but not with PASAT (p = 0.33). CONCLUSION: CFF reduction in MS occurs independently of ON and structural visual system damage. Its association with the EDSS score and alertness suggests that CFF reflects global disease processes and higher cortical processing rather than focal optic nerve or retinal damage.

Active contour method for ILM segmentation in ONH volume scans in retinal OCT.

The optic nerve head (ONH) is affected by many neurodegenerative and autoimmune inflammatory conditions. Optical coherence tomography can acquire high-resolution 3D ONH scans. However, the ONH's complex anatomy and pathology make image segmentation challenging. This paper proposes a robust approach to segment the inner limiting membrane (ILM) in ONH volume scans based on an active contour method of Chan-Vese type, which can work in challenging topological structures. A local intensity fitting energy is added in order to handle very inhomogeneous image intensities. A suitable boundary potential is introduced to avoid structures belonging to outer retinal layers being detected as part of the segmentation. The average intensities in the inner and outer region are then rescaled locally to account for different brightness values occurring among the ONH center. The appropriate values for the parameters used in the complex computational model are found using an optimization based on the differential evolution algorithm. The evaluation of results showed that the proposed framework significantly improved segmentation results compared to the commercial solution.