Diagnostic Value of Inter-Eye Difference Metrics on OCT for Myelin Oligodendrocyte Glycoprotein Antibody-Associated Optic Neuritis.

BACKGROUND AND OBJECTIVES: The 2022 International Consortium for Optic Neuritis diagnostic criteria for optic neuritis (ON) include optical coherence tomography (OCT). The diagnostic value of intereye difference (IED) metrics is high for ON in patients with multiple sclerosis and aquaporin-4 antibody seropositive neuromyelitis optica spectrum disorders, but unknown in myelin oligodendrocyte glycoprotein antibody-associated ON (MOG-ON). METHODS: A multicenter validation study was conducted on the published IED cutoff values (>4% or >4 µm in the macular ganglion cell and inner plexiform layer [mGCIP] or >5% or >5 µm in the peripapillary retinal nerve fiber layer [pRNFL]) in individuals with MOG-ON and age-matched and sex-matched healthy controls (HCs). Structural data were acquired with Spectralis spectral-domain OCT >6 months after ON. We calculated sensitivity, specificity, and receiver operating characteristics for both intereye percentage (IEPD) and absolute difference (IEAD). RESULTS: A total of 66 individuals were included (MOG-ON N = 33; HCs N = 33). ON was unilateral in 20 and bilateral in 13 subjects. In the pooled analysis, the mGCIP IEPD was most sensitive (92%), followed by the mGCIP IEAD (88%) and pRNFL (84%). The same pattern was found for the specificity (mGCIP IEPD 82%, IEAD 82%; pRNFL IEPD 82%, IEAD 79%).In subgroup analyses, the diagnostic sensitivity was higher in subjects with unilateral ON (>99% for all metrics) compared with bilateral ON (61%-78%). DISCUSSION: In individuals with MOG-ON, the diagnostic accuracy of OCT-based IED metrics for ON was high, especially of mGCIP IEPD. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that the intereye difference on OCT can distinguish between those with MOG and normal controls.

Retinal Changes in Double-Antibody Seronegative Neuromyelitis Optica Spectrum Disorders.

BACKGROUND AND OBJECTIVES: To systematically describe the clinical picture of double-antibody seronegative neuromyelitis optica spectrum disorders (DN-NMOSD) with specific emphasis on retinal involvement. METHODS: Cross-sectional data of 25 people with DN-NMOSD (48 eyes) with and without a history of optic neuritis (ON) were included in this study along with data from 25 people with aquaporin-4 antibody seropositive neuromyelitis optica spectrum disorder (AQP4-NMOSD, 46 eyes) and from 25 healthy controls (HCs, 49 eyes) for comparison. All groups were matched for age and sex and included from the collaborative retrospective study of retinal optical coherence tomography (OCT) in neuromyelitis optica (CROCTINO). Participants underwent OCT with central postprocessing and local neurologic examination and antibody testing. Retinal neurodegeneration was quantified as peripapillary retinal nerve fiber layer thickness (pRNFL) and combined ganglion cell and inner plexiform layer thickness (GCIPL). RESULTS: This DN-NMOSD cohort had a history of [median (inter-quartile range)] 6 (5; 9) attacks within their 5 ± 4 years since onset. Myelitis and ON were the most common attack types. In DN-NMOSD eyes after ON, pRNFL (p < 0.001) and GCIPL (p = 0.023) were thinner compared with eyes of HCs. Even after only one ON episode, DN-NMOSD eyes already had considerable neuroaxonal loss compared with HCs. In DN-NMOSD eyes without a history of ON, pRNFL (p = 0.027) and GCIPL (p = 0.022) were also reduced compared with eyes of HCs. However, there was no difference in pRNFL and GCIPL between DN-NMOSD and AQP4-NMOSD for the whole group and for subsets with a history of ON and without a history of ON-as well as between variances of retinal layer thicknesses. DISCUSSION: DN-NMOSD is characterized by severe retinal damage after ON and attack-independent retinal neurodegeneration. Most of the damage occurs during the first ON episode, which highlights the need for better diagnostic markers in DN-NMOSD to facilitate an earlier diagnosis as well as for effective and early treatments. In this study, people with DN-NMOSD presented with homogeneous clinical and imaging findings potentially suggesting a common retinal pathology in these patients.

Individual Prognostication of Disease Activity and Disability Worsening in Multiple Sclerosis With Retinal Layer Thickness z Scores.

BACKGROUND AND OBJECTIVES: Retinal optical coherence tomography (OCT) provides promising prognostic imaging biomarkers for future disease activity in multiple sclerosis (MS). However, raw OCT-derived measures have multiple dependencies, supporting the need for establishing reference values adjusted for possible confounders. The purpose of this study was to investigate the capacity for age-adjusted z scores of OCT-derived measures to prognosticate future disease activity and disability worsening in people with MS (PwMS). METHODS: We established age-adjusted OCT reference data using generalized additive models for location, scale, and shape for peripapillary retinal nerve fiber layer (pRNFL) and ganglion cell-inner plexiform layer (GCIP) thicknesses, involving 910 and 423 healthy eyes, respectively. Next, we transformed the retinal layer thickness of PwMS from 3 published studies into age-adjusted z scores (pRNFL-z and GCIP-z) based on the reference data. Finally, we investigated the association of pRNFL-z or GCIP-z as predictors with future confirmed disability worsening (Expanded Disability Status Scale score increase) or disease activity (failing of the no evidence of disease activity [NEDA-3] criteria) as outcomes. Cox proportional hazards models or logistic regression analyses were applied according to the original studies. Optimal cutoffs were identified using the Akaike information criterion as well as location with the log-rank and likelihood-ratio tests. RESULTS: In the first cohort (n = 863), 172 PwMS (24%) had disability worsening over a median observational period of 2.0 (interquartile range [IQR]:1.0-3.0) years. Low pRNFL-z (≤-2.04) were associated with an increased risk of disability worsening (adjusted hazard ratio (aHR) [95% CI] = 2.08 [1.47-2.95], p = 3.82e-5). In the second cohort (n = 170), logistic regression analyses revealed that lower pRNFL-z showed a higher likelihood for disability accumulation at the two-year follow-up (reciprocal odds ratio [95% CI] = 1.51[1.06-2.15], p = 0.03). In the third cohort (n = 78), 46 PwMS (59%) did not maintain the NEDA-3 status over a median follow-up of 2.0 (IQR: 1.9-2.1) years. PwMS with low GCIP-z (≤-1.03) had a higher risk of showing disease activity (aHR [95% CI] = 2.14 [1.03-4.43], p = 0.04). Compared with raw values with arbitrary cutoffs, applying the z score approach with optimal cutoffs showed better performance in discrimination and calibration (higher Harrell's concordance index and lower integrated Brier score). DISCUSSION: In conclusion, our work demonstrated reference cohort-based z scores that account for age, a major driver for disease progression in MS, to be a promising approach for creating OCT-derived measures useable across devices and toward individualized prognostication.

Multiscale networks in multiple sclerosis.

Complex diseases such as Multiple Sclerosis (MS) cover a wide range of biological scales, from genes and proteins to cells and tissues, up to the full organism. In fact, any phenotype for an organism is dictated by the interplay among these scales. We conducted a multilayer network analysis and deep phenotyping with multi-omics data (genomics, phosphoproteomics and cytomics), brain and retinal imaging, and clinical data, obtained from a multicenter prospective cohort of 328 patients and 90 healthy controls. Multilayer networks were constructed using mutual information for topological analysis, and Boolean simulations were constructed using Pearson correlation to identified paths within and among all layers. The path more commonly found from the Boolean simulations connects protein MK03, with total T cells, the thickness of the retinal nerve fiber layer (RNFL), and the walking speed. This path contains nodes involved in protein phosphorylation, glial cell differentiation, and regulation of stress-activated MAPK cascade, among others. Specific paths identified were subsequently analyzed by flow cytometry at the single-cell level. Combinations of several proteins (GSK3AB, HSBP1 or RS6) and immune cells (Th17, Th1 non-classic, CD8, CD8 Treg, CD56 neg, and B memory) were part of the paths explaining the clinical phenotype. The advantage of the path identified from the Boolean simulations is that it connects information about these known biological pathways with the layers at higher scales (retina damage and disability). Overall, the identified paths provide a means to connect the molecular aspects of MS with the overall phenotype.

Interactions of optic radiation lesions with retinal and brain atrophy in early multiple sclerosis.

OBJECTIVE: Retrograde trans-synaptic neuroaxonal degeneration is considered a key pathological factor of subclinical retinal neuroaxonal damage in multiple sclerosis (MS). We aim to evaluate the longitudinal association of optic radiation (OR) lesion activity with retinal neuroaxonal damage and its role in correlations between retinal and brain atrophy in people with clinically isolated syndrome and early MS (pweMS). METHODS: Eighty-five pweMS were retrospectively screened from a prospective cohort (Berlin CIS cohort). Participants underwent 3T magnetic resonance imaging (MRI) for OR lesion volume and brain atrophy measurements and optical coherence tomography (OCT) for retinal layer thickness measurements. All pweMS were followed with serial OCT and MRI over a median follow-up of 2.9 (interquartile range: 2.6-3.4) years. Eyes with a history of optic neuritis prior to study enrollment were excluded. Linear mixed models were used to analyze the association of retinal layer thinning with changes in OR lesion volume and brain atrophy. RESULTS: Macular ganglion cell-inner plexiform layer (GCIPL) thinning was more pronounced in pweMS with OR lesion volume increase during follow-up compared to those without (Difference: -0.82 µm [95% CI:-1.49 to -0.15], p = 0.018). Furthermore, GCIPL thinning correlated with both OR lesion volume increase (ß [95% CI] = -0.27 [-0.50 to -0.03], p = 0.028) and brain atrophy (ß [95% CI] = 0.47 [0.25 to 0.70], p < 0.001). Correlations of GCIPL changes with brain atrophy did not differ between pweMS with or without OR lesion increase ( η p 2 = 5.92e-7 , p = 0.762). INTERPRETATION: Faster GCIPL thinning rate is associated with increased OR lesion load. Our results support the value of GCIPL as a sensitive biomarker reflecting both posterior visual pathway pathology and global brain neurodegeneration.

Predicting disease severity in multiple sclerosis using multimodal data and machine learning.

BACKGROUND: Multiple sclerosis patients would benefit from machine learning algorithms that integrates clinical, imaging and multimodal biomarkers to define the risk of disease activity. METHODS: We have analysed a prospective multi-centric cohort of 322 MS patients and 98 healthy controls from four MS centres, collecting disability scales at baseline and 2 years later. Imaging data included brain MRI and optical coherence tomography, and omics included genotyping, cytomics and phosphoproteomic data from peripheral blood mononuclear cells. Predictors of clinical outcomes were searched using Random Forest algorithms. Assessment of the algorithm performance was conducted in an independent prospective cohort of 271 MS patients from a single centre. RESULTS: We found algorithms for predicting confirmed disability accumulation for the different scales, no evidence of disease activity (NEDA), onset of immunotherapy and the escalation from low- to high-efficacy therapy with intermediate to high-accuracy. This accuracy was achieved for most of the predictors using clinical data alone or in combination with imaging data. Still, in some cases, the addition of omics data slightly increased algorithm performance. Accuracies were comparable in both cohorts. CONCLUSION: Combining clinical, imaging and omics data with machine learning helps identify MS patients at risk of disability worsening.

Discrimination of multiple sclerosis using OCT images from two different centers.

BACKGROUND: Multiple sclerosis (MS) is one of the most prevalent chronic inflammatory diseases caused by demyelination and axonal damage in the central nervous system. Structural retinal imaging via optical coherence tomography (OCT) shows promise as a noninvasive biomarker for monitoring of MS. There are successful reports regarding the application of Artificial Intelligence (AI) in the analysis of cross-sectional OCTs in ophthalmologic diseases. However, the alteration of thicknesses of various retinal layers in MS is noticeably subtle compared to other ophthalmologic diseases. Therefore, raw cross-sectional OCTs are replaced with multilayer segmented OCTs for discrimination of MS and healthy controls (HCs). METHODS: To conform to the principles of trustworthy AI, interpretability is provided by visualizing the regional layer contribution to classification performance with the proposed occlusion sensitivity approach. The robustness of the classification is also guaranteed by showing the effectiveness of the algorithm while being tested on the new independent dataset. The most discriminative features from different topologies of the multilayer segmented OCTs are selected by the dimension reduction method. Support vector machine (SVM), random forest (RF), and artificial neural network (ANN) are used for classification. Patient-wise cross-validation (CV) is utilized to evaluate the performance of the algorithm, where the training and test folds contain records from different subjects. RESULTS: The most discriminative topology is determined to square with a size of 40 pixels and the most influential layers are the ganglion cell and inner plexiform layer (GCIPL) and inner nuclear layer (INL). Linear SVM resulted in 88% Accuracy (with standard deviation (std) = 0.49 in 10 times of execution to indicate the repeatability), 78% precision (std=1.48), and 63% recall (std=1.35) in the discrimination of MS and HCs using macular multilayer segmented OCTs. CONCLUSION: The proposed classification algorithm is expected to help neurologists in the early diagnosis of MS. This paper distinguishes itself from other studies by employing two distinct datasets, which enhances the robustness of its findings in comparison with previous studies with lack of external validation. This study aims to circumvent the utilization of deep learning methods due to the limited quantity of the available data and convincingly demonstrates that favorable outcomes can be achieved without relying on deep learning techniques.

Visually Evoked Potential as Prognostic Biomarker for Neuroaxonal Damage in Multiple Sclerosis From a Multicenter Longitudinal Cohort.

BACKGROUND AND OBJECTIVES: With the increasing use of visually evoked potentials (VEPs) as quantitative outcome parameters for myelin in clinical trials, an in-depth understanding of longitudinal VEP latency changes and their prognostic potential for subsequent neuronal loss will be required. In this longitudinal multicenter study, we evaluated the association and prognostic potential of VEP latency for retinal neurodegeneration, measured by optical coherence tomography (OCT), in relapsing-remitting MS (RRMS). METHODS: We included 293 eyes of 147 patients with RRMS (age [years, median ± SD] 36 ± 10, male sex 35%, F/U [years, median {IQR} 2.1 {1.5-3.9}]): 41 eyes had a history of optic neuritis (ON) ≥6 months before baseline (CHRONIC-ON), and 252 eyes had no history of ON (CHRONIC-NON). P100 latency (VEP), macular combined ganglion cell and inner plexiform layer volume (GCIPL), and peripapillary retinal nerve fiber layer thickness (pRNFL) (OCT) were quantified. RESULTS: P100 latency change over the first year predicted subsequent GCIPL loss (36 months) across the entire chronic cohort (p = 0.001) and in (and driven by) the CHRONIC-NON subset (p = 0.019) but not in the CHRONIC-ON subset (p = 0.680). P100 latency and pRNFL were correlated at baseline (CHRONIC-NON p = 0.004, CHRONIC-ON p < 0.001), but change in P100 latency and pRNFL were not correlated. P100 latency did not differ longitudinally between protocols or centers. DISCUSSION: VEP in non-ON eyes seems to be a promising marker of demyelination in RRMS and of potential prognostic value for subsequent retinal ganglion cell loss. This study also provides evidence that VEP may be a useful and reliable biomarker for multicenter studies.

Diagnostic value of intereye difference metrics for optic neuritis in aquaporin-4 antibody seropositive neuromyelitis optica spectrum disorders.

BACKGROUND: The novel optic neuritis (ON) diagnostic criteria include intereye differences (IED) of optical coherence tomography (OCT) parameters. IED has proven valuable for ON diagnosis in multiple sclerosis but has not been evaluated in aquaporin-4 antibody seropositive neuromyelitis optica spectrum disorders (AQP4+NMOSD). We evaluated the diagnostic accuracy of intereye absolute (IEAD) and percentage difference (IEPD) in AQP4+NMOSD after unilateral ON >6 months before OCT as compared with healthy controls (HC). METHODS: Twenty-eight AQP4+NMOSD after unilateral ON (NMOSD-ON), 62 HC and 45 AQP4+NMOSD without ON history (NMOSD-NON) were recruited by 13 centres as part of the international Collaborative Retrospective Study on retinal OCT in Neuromyelitis Optica study. Mean thickness of peripapillary retinal nerve fibre layer (pRNFL) and macular ganglion cell and inner plexiform layer (GCIPL) were quantified by Spectralis spectral domain OCT. Threshold values of the ON diagnostic criteria (pRNFL: IEAD 5 µm, IEPD 5%; GCIPL: IEAD: 4 µm, IEPD: 4%) were evaluated using receiver operating characteristics and area under the curve (AUC) metrics. RESULTS: The discriminative power was high for NMOSD-ON versus HC for IEAD (pRNFL: AUC 0.95, specificity 82%, sensitivity 86%; GCIPL: AUC 0.93, specificity 98%, sensitivity 75%) and IEPD (pRNFL: AUC 0.96, specificity 87%, sensitivity 89%; GCIPL: AUC 0.94, specificity 96%, sensitivity 82%). The discriminative power was high/moderate for NMOSD-ON versus NMOSD-NON for IEAD (pRNFL: AUC 0.92, specificity 77%, sensitivity 86%; GCIP: AUC 0.87, specificity 85%, sensitivity 75%) and for IEPD (pRNFL: AUC 0.94, specificity 82%, sensitivity 89%; GCIP: AUC 0.88, specificity 82%, sensitivity 82%). CONCLUSIONS: Results support the validation of the IED metrics as OCT parameters of the novel diagnostic ON criteria in AQP4+NMOSD.

Structure-function correlates of vision loss in neuromyelitis optica spectrum disorders.

Optic neuritis (ON) in neuromyelitis optica spectrum disorders (NMOSD) regularly leads to more profound vision loss compared to multiple sclerosis (MS) and myelin-oligodendrocyte-glycoprotein-antibody associated disease (MOGAD). Here we investigate ON-related vision loss in NMOSD compared to MS and MOGAD in order to identify neuroaxonal and retinal contributors to visual dysfunction. In this retrospective study we included patients with aquaporin-4-antibody seropositive NMOSD (n = 28), MOGAD (n = 14), MS (n = 29) and controls (n = 14). We assessed optic nerve damage and fovea morphometry by optical coherence tomography. Visual function was assessed as high (HCVA) and low contrast visual acuity (LCVA), and visual fields' mean deviation (MD). In all diseases, lower visual function was associated with peripapillary retinal nerve fiber layer (pRNFL) and ganglion cell and inner plexiform layer (GCIP) thinning following a broken stick model, with pRNFL and GCIP cutoff point at ca. 60 µm. HCVA loss per µm pRNFL and GCIP thinning was stronger in NMOSD compared with MOGAD. Foveal inner rim volume contributed to MD and LCVA in NMOSD eyes, only. Together these data supports that visual dysfunction in NMOSD is associated with neuroaxonal damage beyond the effect seen in MS and MOGAD. A primary retinopathy, respectively Müller cell pathology, may contribute to this effect.

Prior optic neuritis detection on peripapillary ring scans using deep learning.

BACKGROUND: The diagnosis of multiple sclerosis (MS) requires demyelinating events that are disseminated in time and space. Peripapillary retinal nerve fiber layer (pRNFL) thickness as measured by optical coherence tomography (OCT) distinguishes eyes with a prior history of acute optic neuritis (ON) and may provide evidence to support a demyelinating attack. OBJECTIVE: To investigate whether a deep learning (DL)-based network can distinguish between eyes with prior ON and healthy control (HC) eyes using peripapillary ring scans. METHODS: We included 1033 OCT scans from 415 healthy eyes (213 HC subjects) and 510 peripapillary ring scans from 164 eyes with prior acute ON (140 patients with MS). Data were split into 70% training, 15% validation, and 15% test data. We included 102 OCT scans from 80 healthy eyes (40 HC) and 61 scans from 40 ON eyes (31 MS patients) from an independent second center. Receiver operating characteristic curve analyses with area under the curve (AUC) were used to investigate performance. RESULTS: We used a dilated residual convolutional neural network for the classification. The final network had an accuracy of 0.85 and an AUC of 0.86, whereas pRNFL only had an AUC of 0.77 in recognizing ON eyes. Using data from a second center, the network achieved an accuracy of 0.77 and an AUC of 0.90 compared to pRNFL, which had an AUC of 0.84. INTERPRETATION: DL-based disease classification of prior ON is feasible and has the potential to outperform thickness-based classification of eyes with and without history of prior ON.

Serum glial fibrillary acidic protein correlates with retinal structural damage in aquaporin-4 antibody positive neuromyelitis optica spectrum disorder.

BACKGROUND: Aquaporin-4 immunoglobulin-G positive (AQP4-IgG+) neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune astrocytopathy associated with optic neuritis (ON). Myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) is an oligodendrocytopathy with a similar phenotype. Serum glial fibrillary acidic protein (sGFAP), an astrocyte-derived protein, is associated with disease severity in AQP4-IgG+ NMOSD. Serum neurofilament light (sNfL) indicates neuroaxonal damage. The objective was to investigate the association of sGFAP and sNfL with subclinical afferent visual system damage in clinically stable AQP4-IgG+ NMOSD and MOGAD patients. METHODS: In this cross-sectional study, clinically stable patients with AQP4-IgG+ NMOSD (N = 33) and MOGAD (N = 16), as diseased controls, underwent sGFAP and sNfL measurements by single molecule array, retinal optical coherence tomography and visually evoked potentials. RESULTS: Higher sGFAP concentrations were associated with thinner ganglion cell-inner plexiform layer (ß (95% confidence interval (CI)) = -0.75 (-1.23 to -0.27), p = 0.007) and shallower fovea (average pit depth: ß (95%CI) = -0.59 (-0.63 to -0.55), p = 0.020) in NMOSD non-ON eyes. Participants with pathological P100 latency had higher sGFAP (median [interquartile range]: 131.32 [81.10-179.34] vs. 89.50 [53.46-121.91] pg/ml, p = 0.024). In MOGAD, sGFAP was not associated with retinal structural or visual functional measures. CONCLUSIONS: The association of sGFAP with structural and functional markers of afferent visual system damage in absence of ON suggests that sGFAP may be a sensitive biomarker for chronic disease severity in clinically stable AQP4-IgG+ NMOSD.

Longitudinal Retinal Changes in MOGAD.

OBJECTIVE: Patients with myelin oligodendrocyte glycoprotein antibody (MOG-IgG)-associated disease (MOGAD) suffer from severe optic neuritis (ON) leading to retinal neuro-axonal loss, which can be quantified by optical coherence tomography (OCT). We assessed whether ON-independent retinal atrophy can be detected in MOGAD. METHODS: Eighty patients with MOGAD and 139 healthy controls (HCs) were included. OCT data was acquired with (1) Spectralis spectral domain OCT (MOGAD: N = 66 and HCs: N = 103) and (2) Cirrus high-definition OCT (MOGAD: N = 14 and HCs: N = 36). Macular combined ganglion cell and inner plexiform layer (GCIPL) and peripapillary retinal nerve fiber layer (pRNFL) were quantified. RESULTS: At baseline, GCIPL and pRNFL were lower in MOGAD eyes with a history of ON (MOGAD-ON) compared with MOGAD eyes without a history of ON (MOGAD-NON) and HCs (p < 0.001). MOGAD-NON eyes had lower GCIPL volume compared to HCs (p < 0.001) in the Spectralis, but not in the Cirrus cohort. Longitudinally (follow-up up to 3 years), MOGAD-ON with ON within the last 6-12 months before baseline exhibited greater pRNFL thinning than MOGAD-ON with an ON greater than 12 months ago (p < 0.001). The overall MOGAD cohort did not exhibit faster GCIPL thinning compared with the HC cohort. INTERPRETATION: Our study suggests the absence of attack-independent retinal damage in patients with MOGAD. Yet, ongoing neuroaxonal damage or edema resolution seems to occur for up to 12 months after ON, which is longer than what has been reported with other ON forms. These findings support that the pathomechanisms underlying optic nerve involvement and the evolution of OCT retinal changes after ON is distinct in patients with MOGAD. ANN NEUROL 2022;92:476-485.

Intraretinal Layer Segmentation Using Cascaded Compressed U-Nets.

Reliable biomarkers quantifying neurodegeneration and neuroinflammation in central nervous system disorders such as Multiple Sclerosis, Alzheimer's dementia or Parkinson's disease are an unmet clinical need. Intraretinal layer thicknesses on macular optical coherence tomography (OCT) images are promising noninvasive biomarkers querying neuroretinal structures with near cellular resolution. However, changes are typically subtle, while tissue gradients can be weak, making intraretinal segmentation a challenging task. A robust and efficient method that requires no or minimal manual correction is an unmet need to foster reliable and reproducible research as well as clinical application. Here, we propose and validate a cascaded two-stage network for intraretinal layer segmentation, with both networks being compressed versions of U-Net (CCU-INSEG). The first network is responsible for retinal tissue segmentation from OCT B-scans. The second network segments eight intraretinal layers with high fidelity. At the post-processing stage, we introduce Laplacian-based outlier detection with layer surface hole filling by adaptive non-linear interpolation. Additionally, we propose a weighted version of focal loss to minimize the foreground-background pixel imbalance in the training data. We train our method using 17,458 B-scans from patients with autoimmune optic neuropathies, i.e., multiple sclerosis, and healthy controls. Voxel-wise comparison against manual segmentation produces a mean absolute error of 2.3 µm, outperforming current state-of-the-art methods on the same data set. Voxel-wise comparison against external glaucoma data leads to a mean absolute error of 2.6 µm when using the same gold standard segmentation approach, and 3.7 µm mean absolute error in an externally segmented data set. In scans from patients with severe optic atrophy, 3.5% of B-scan segmentation results were rejected by an experienced grader, whereas this was the case in 41.4% of B-scans segmented with a graph-based reference method. The validation results suggest that the proposed method can robustly segment macular scans from eyes with even severe neuroretinal changes.

Astrocytic outer retinal layer thinning is not a feature in AQP4-IgG seropositive neuromyelitis optica spectrum disorders.

BACKGROUND: Patients with anti-aquaporin-4 antibody seropositive (AQP4-IgG+) neuromyelitis optica spectrum disorders (NMOSDs) frequently suffer from optic neuritis (ON) leading to severe retinal neuroaxonal damage. Further, the relationship of this retinal damage to a primary astrocytopathy in NMOSD is uncertain. Primary astrocytopathy has been suggested to cause ON-independent retinal damage and contribute to changes particularly in the outer plexiform layer (OPL) and outer nuclear layer (ONL), as reported in some earlier studies. However, these were limited in their sample size and contradictory as to the localisation. This study assesses outer retinal layer changes using optical coherence tomography (OCT) in a multicentre cross-sectional cohort. METHOD: 197 patients who were AQP4-IgG+ and 32 myelin-oligodendrocyte-glycoprotein antibody seropositive (MOG-IgG+) patients were enrolled in this study along with 75 healthy controls. Participants underwent neurological examination and OCT with central postprocessing conducted at a single site. RESULTS: No significant thinning of OPL (25.02±2.03 µm) or ONL (61.63±7.04 µm) were observed in patients who were AQP4-IgG+ compared with patients who were MOG-IgG+ with comparable neuroaxonal damage (OPL: 25.10±2.00 µm; ONL: 64.71±7.87 µm) or healthy controls (OPL: 24.58±1.64 µm; ONL: 63.59±5.78 µm). Eyes of patients who were AQP4-IgG+ (19.84±5.09 µm, p=0.027) and MOG-IgG+ (19.82±4.78 µm, p=0.004) with a history of ON showed parafoveal OPL thinning compared with healthy controls (20.99±5.14 µm); this was not observed elsewhere. CONCLUSION: The results suggest that outer retinal layer loss is not a consistent component of retinal astrocytic damage in AQP4-IgG+ NMOSD. Longitudinal studies are necessary to determine if OPL and ONL are damaged in late disease due to retrograde trans-synaptic axonal degeneration and whether outer retinal dysfunction occurs despite any measurable structural correlates.

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.

Retinal Optical Coherence Tomography in Neuromyelitis Optica.

BACKGROUND AND OBJECTIVES: To determine optic nerve and retinal damage in aquaporin-4 antibody (AQP4-IgG)-seropositive neuromyelitis optica spectrum disorders (NMOSD) in a large international cohort after previous studies have been limited by small and heterogeneous cohorts. METHODS: The cross-sectional Collaborative Retrospective Study on retinal optical coherence tomography (OCT) in neuromyelitis optica collected retrospective data from 22 centers. Of 653 screened participants, we included 283 AQP4-IgG-seropositive patients with NMOSD and 72 healthy controls (HCs). Participants underwent OCT with central reading including quality control and intraretinal segmentation. The primary outcome was thickness of combined ganglion cell and inner plexiform (GCIP) layer; secondary outcomes were thickness of peripapillary retinal nerve fiber layer (pRNFL) and visual acuity (VA). RESULTS: Eyes with ON (NMOSD-ON, N = 260) or without ON (NMOSD-NON, N = 241) were assessed compared with HCs (N = 136). In NMOSD-ON, GCIP layer (57.4 ± 12.2 µm) was reduced compared with HC (GCIP layer: 81.4 ± 5.7 µm, p < 0.001). GCIP layer loss (-22.7 µm) after the first ON was higher than after the next (-3.5 µm) and subsequent episodes. pRNFL observations were similar. NMOSD-NON exhibited reduced GCIP layer but not pRNFL compared with HC. VA was greatly reduced in NMOSD-ON compared with HC eyes, but did not differ between NMOSD-NON and HC. DISCUSSION: Our results emphasize that attack prevention is key to avoid severe neuroaxonal damage and vision loss caused by ON in NMOSD. Therapies ameliorating attack-related damage, especially during a first attack, are an unmet clinical need. Mild signs of neuroaxonal changes without apparent vision loss in ON-unaffected eyes might be solely due to contralateral ON attacks and do not suggest clinically relevant progression but need further investigation.

Increased Serum Neurofilament Light and Thin Ganglion Cell-Inner Plexiform Layer Are Additive Risk Factors for Disease Activity in Early Multiple Sclerosis.

OBJECTIVE: To investigate the association of combined serum neurofilament light chain (sNfL) and retinal optical coherence tomography (OCT) measurements with future disease activity in patients with early multiple sclerosis (MS). METHODS: We analyzed sNfL by single molecule array technology and performed OCT measurements in a prospective cohort of 78 patients with clinically isolated syndrome and early relapsing-remitting MS with a median (interquartile range) follow-up of 23.9 (23.3-24.7) months. Patients were grouped into those with abnormal or normal sNfL levels, defined as sNfL ≥/<80th percentile of age-corrected reference values. Likewise, patients were grouped by a median split into those with thin or thick ganglion cell and inner plexiform layer (GCIP), peripapillary retinal nerve fiber layer, and inner nuclear layer in nonoptic neuritis eyes. Outcome parameters were violation of no evidence of disease activity (NEDA-3) criteria or its components. RESULTS: Patients with abnormal baseline sNfL had a higher risk of violating NEDA-3 (hazard ratio [HR] 2.28, 95% CI 1.27-4.09, p = 0.006) and developing a new brain lesion (HR 2.47, 95% CI 1.30-4.69, p = 0.006), but not for a new relapse (HR 2.21, 95% CI 0.97-5.03, p = 0.058). Patients with both abnormal sNfL and thin GCIP had an even higher risk for NEDA-3 violation (HR 3.61, 95% CI 1.77-7.36, p = 4.2e-4), new brain lesion (HR 3.19, 95% CI 1.51-6.76, p = 0.002), and new relapse (HR 5.38, 95% CI 1.61-17.98, p = 0.006) than patients with abnormal sNfL alone. CONCLUSIONS: In patients with early MS, the presence of both abnormal sNfL and thin GCIP is a stronger risk factor for future disease activity than the presence of each parameter alone.

Lateral geniculate nucleus volume changes after optic neuritis in neuromyelitis optica: A longitudinal study.

OBJECTIVES: Lateral geniculate nucleus (LGN) volume is reduced after optic neuritis (ON) in neuromyelitis optica spectrum disorders (NMOSD). We aimed at a longitudinal assessment of LGN volume in NMOSD. METHODS: Twenty-nine patients with aquaporin 4-IgG seropositive NMOSD (age: 47.8 ± 14.6 years (y), female: n = 27, history of ON (NMO-ON): n = 17, median time since ON: 3[1.2-12.1]y) and 18 healthy controls (HC; age: 39.3 ± 15.8y; female: n = 13) were included. Median follow-up was 4.1[1.1-4.7]y for patients and 1.7[0.9-3.2]y for HC. LGN volume was measured using a multi-atlas-based approach of automated segmentation on 3 Tesla magnetic resonance images. Retinal optical coherence tomography and probabilistic tractography of the optic radiations (OR) were also performed. RESULTS: At baseline, NMO-ON patients had lower LGN volumes (395.4 ± 48.9 mm3) than patients without ON (NMO-NON: 450.7 ± 55.6 mm3; p = 0.049) and HC (444.5 ± 61.5 mm3, p = 0.025). LGN volume was associated with retinal neuroaxonal loss and microstructural OR damage. Longitudinally, there was no change in LGN volumes in the absence of ON, neither in all patients (B = -0.6, SE = 1.4, p = 0.670), nor in NMO-ON (B = -0.8, SE = 1.6, p = 0.617) and NMO-NON (B = 1.7, SE = 3.5, p = 0.650). However, in four patients with new ON during follow-up, LGN volume was reduced at last visit (median time since ON: 2.6 [1.8-3.9]y) compared to the measurement before ON (352 ± 52.7 vs. 371.1 ± 55.9 mm3; t = -3.6, p = 0.036). CONCLUSION: Although LGN volume is reduced after ON in NMOSD, this volume loss is not progressive over longer follow-up or independent of ON. Thus, our findings -at least in this relatively small cohort- do not support occult neurodegeneration of the afferent visual pathway in NMOSD.