Three "Red Lines" for Pattern Recognition-Based Differential Diagnosis Using Optical Coherence Tomography in Clinical Practice

Longitudinal evaluation of retinal neuroaxonal loss in epilepsy using optical coherence tomography

OBJECTIVE

People with epilepsy (PwE) suffer from progressive brain atrophy, which is reflected as neuroaxonal loss on the retinal level. This study aims to provide initial insight into the longitudinal dynamics of the retinal neuroaxonal loss and possible driving factors.

METHODS

PwE and healthy controls (HC; 18-55 years of age) underwent spectral domain optical coherence tomography at baseline and 7.0 ± 1.5 and 6.7 ± 1.0 months later, respectively. The change in retinal thickness/volume and annualized percentage change (APC) were calculated for the peripapillary retinal nerve fiber layer (pRNFL), the macular RNFL (mRNFL), the ganglion cell inner plexiform layer (GCIP), the inner nuclear layer, and the total macular volume (TMV). Group comparisons and multiple linear models with stepwise backward selection were performed to evaluate associations with demographic and clinical parameters.

RESULTS

PwE (n = 44, 21 females, mean age = 35.6 ± 10.9 years) revealed a significant decrease in the pRNFL, mRNFL, GCIP, and TMV thickness or volume in the study interval. When compared to HC (n = 56, 37 females, mean age = 32.7 ± 8.3 years), the APC of the pRNFL (-.98 ± 3.13%/year) and the GCIP (-1.24 ± 2.56%/year) were significantly more pronounced in PwE (p = .01 and p = .046, respectively). Of note, atrophy of the mRNFL was significantly influenced by the number of antiseizure medications (ASMs; p = .047) and increasing age of PwE (p = .03). Contradictory results, however, were revealed for the impact of seizures.

SIGNIFICANCE

In epilepsy, progression of retinal neuroaxonal loss was already detectable at short-term follow-up. PwE who receive a high number of ASMs seem to be at risk for accelerated neuroaxonal loss, stressing the importance of well-considered and effective antiseizure therapy.

Differentiating glaucoma from chiasmal compression using optical coherence tomography: the macular naso-temporal ratio

BACKGROUND/AIMS

The analysis of visual field loss patterns is clinically useful to guide differential diagnosis of visual pathway pathology. This study investigates whether a novel index of macular atrophy patterns can discriminate between chiasmal compression and glaucoma.

METHODS

A retrospective series of patients with preoperative chiasmal compression, primary open-angle glaucoma (POAG) and healthy controls. Macular optical coherence tomography (OCT) images were analysed for the macular ganglion cell and inner plexiform layer (mGCIPL) thickness. The nasal hemi-macula was compared with the temporal hemi-macula to derive the macular naso-temporal ratio (mNTR). Differences between groups and diagnostic accuracy were explored with multivariable linear regression and the area under the receiver operating characteristic curve (AUC).

RESULTS

We included 111 individuals (31 with chiasmal compression, 30 with POAG and 50 healthy controls). Compared with healthy controls, the mNTR was significantly greater in POAG cases (β=0.07, 95% CI 0.03 to 0.11, p=0.001) and lower in chiasmal compression cases (β=-0.12, 95% CI -0.16 to -0.09, p<0.001), even though overall mGCIPL thickness did not discriminate between these pathologies (p=0.36). The mNTR distinguished POAG from chiasmal compression with an AUC of 95.3% (95% CI 90% to 100%). The AUCs when comparing healthy controls to POAG and chiasmal compression were 79.0% (95% CI 68% to 90%) and 89.0% (95% CI 80% to 98%), respectively.

CONCLUSIONS

The mNTR can distinguish between chiasmal compression and POAG with high discrimination. This ratio may provide utility over-and-above previously reported sectoral thinning metrics. Incorporation of mNTR into the output of OCT instruments may aid earlier diagnosis of chiasmal compression.

Deep learning model to identify homonymous defects on automated perimetry

BACKGROUND

Homonymous visual field (VF) defects are usually an indicator of serious intracranial pathology but may be subtle and difficult to detect. Artificial intelligence (AI) models could play a key role in simplifying the detection of these defects. This study aimed to develop an automated deep learning AI model to accurately identify homonymous VF defects from automated perimetry.

METHODS

VFs performed on Humphrey field analyser (24-2 algorithm) were collected and run through an in-house optical character recognition program that extracted mean deviation data and prepared it for use in the proposed AI model. The deep learning AI model, Deep Homonymous Classifier, was developed using PyTorch framework and used convolutional neural networks to extract spatial features for binary classification. Total collected dataset underwent 7-fold cross validation for model training and evaluation. To address dataset class imbalance, data augmentation techniques and state-of-the-art loss function that uses complement cross entropy were used to train and enhance the proposed AI model.

RESULTS

The proposed model was evaluated using 7-fold cross validation and achieved an average accuracy of 87% for detecting homonymous VF defects in previously unseen VFs. Recall, which is a critical value for this model as reducing false negatives is a priority in disease detection, was found to be on average 92%. The calculated F2 score for the proposed model was 0.89 with a Cohen's kappa value of 0.70.

CONCLUSION

This newly developed deep learning model achieved an overall average accuracy of 87%, making it highly effective in identifying homonymous VF defects on automated perimetry.

Clinical review of retinotopy

Two observations made 29 years apart are the cornerstones of this review on the contributions of Dr Gordon T. Plant to understanding pathology affecting the optic nerve. The first observation laid the anatomical basis in 1990 for the interpretation of optical coherence tomography (OCT) findings in 2009. Retinal OCT offers clinicians detailed in vivo structural imaging of individual retinal layers. This has led to novel observations which were impossible to make using ophthalmoscopy. The technique also helps to re-introduce the anatomically grounded concept of retinotopy to clinical practise. This review employs illustrations of the anatomical basis for retinotopy through detailed translational histological studies and multimodal brain-eye imaging studies. The paths of the prelaminar and postlaminar axons forming the optic nerve and their postsynaptic path from the dorsal lateral geniculate nucleus to the primary visual cortex in humans are described. With the mapped neuroanatomy in mind we use OCT-MRI pairings to discuss the patterns of neurodegeneration in eye and brain that are a consequence of the hard wired retinotopy: anterograde and retrograde axonal degeneration which can, within the visual system, propagate trans-synaptically. The technical advances of OCT and MRI for the first time enable us to trace axonal degeneration through the entire visual system at spectacular resolution. In conclusion, the neuroanatomical insights provided by the combination of OCT and MRI allows us to separate incidental findings from sinister pathology and provides new opportunities to tailor and monitor novel neuroprotective strategies.

Ganglion Cell Complex Analysis: Correlations with Retinal Nerve Fiber Layer on Optical Coherence Tomography

The aim of this review is to analyze the correlations between the changes in the ganglion cell complex (GCC) and the retinal nerve fiber layer (RNFL) on optical coherence tomography in different possible situations, especially in eyes with glaucoma. For glaucoma evaluation, several studies have suggested that in the early stages, GCC analysis, especially the thickness of the infero and that of the inferotemporal GCC layers, is a more sensitive examination than circumpapillary RNFL (pRNFL). In the moderate stages of glaucoma, inferior pRNFL thinning is better correlated with the disease than in advanced cases. Another strategy for glaucoma detection is to find any asymmetry of the ganglion cell-inner plexiform layers (GCIPL) between the two macular hemifields, because this finding is a valuable indicator for preperimetric glaucoma, better than the RNFL thickness or the absolute thickness parameters of GCIPL. In preperimetric and suspected glaucoma, GCC and pRNFL have better specificity and are superior to the visual field. In advanced stages, pRNFL and later, GCC reach the floor effect. Therefore, in this stage, it is more useful to evaluate the visual field for monitoring the progression of glaucoma. In conclusion, GCC and pRNFL are parameters that can be used for glaucoma diagnosis and monitoring of the progression of the disease, with each having a higher accuracy depending on the stage of the disease.

Diagnosis and classification of optic neuritis

There is no consensus regarding the classification of optic neuritis, and precise diagnostic criteria are not available. This reality means that the diagnosis of disorders that have optic neuritis as the first manifestation can be challenging. Accurate diagnosis of optic neuritis at presentation can facilitate the timely treatment of individuals with multiple sclerosis, neuromyelitis optica spectrum disorder, or myelin oligodendrocyte glycoprotein antibody-associated disease. Epidemiological data show that, cumulatively, optic neuritis is most frequently caused by many conditions other than multiple sclerosis. Worldwide, the cause and management of optic neuritis varies with geographical location, treatment availability, and ethnic background. We have developed diagnostic criteria for optic neuritis and a classification of optic neuritis subgroups. Our diagnostic criteria are based on clinical features that permit a diagnosis of possible optic neuritis; further paraclinical tests, utilising brain, orbital, and retinal imaging, together with antibody and other protein biomarker data, can lead to a diagnosis of definite optic neuritis. Paraclinical tests can also be applied retrospectively on stored samples and historical brain or retinal scans, which will be useful for future validation studies. Our criteria have the potential to reduce the risk of misdiagnosis, provide information on optic neuritis disease course that can guide future treatment trial design, and enable physicians to judge the likelihood of a need for long-term pharmacological management, which might differ according to optic neuritis subgroups.

Visual fields and optical coherence tomography (OCT) in neuro-ophthalmology: Structure-function correlation

Visual field (VF) testing is an essential component of the neurological examination. The differential diagnosis of VF defects depends on relating this measure of afferent visual function to the structure of the visual pathway and optical coherence tomography (OCT) is an invaluable tool for detailed structural evaluation of the optic nerve and retina. This review describes the ways in which interpretation of VF and OCT can be used together to increase the accuracy of the localization of lesions along the visual pathway. Lesions of the anterior visual pathway (originating in ganglion cells or nerve fibre layer of the retina or optic nerve) will typically produce defects that respect the horizontal midline, reflecting the arcuate path of the ganglion cell axons as they travel to the optic nerve. OCT of peripapillary retinal nerve fibre layer and ganglion cell complex (GCC) will typically demonstrate irreversible thinning in compressive and demyelinating lesions affecting anterior visual pathway. Chiasmal lesions produce highly localizable VF defects (junctional scotoma and bitemporal hemianopia) which correspond to the thinning of nasal portion of GCC. Lesions of the optic tract result in incongruous homonymous hemianopia on VF with corresponding hemianopic thinning on GCC developing within months. Lesions affecting optic radiations usually produce more congruous homonymous VF defects and can also produce homonymous thinning on GCC, however, this takes much longer to develop as trans-synaptic degeneration at the lateral geniculate body must occur.

Re-evaluating diabetic papillopathy using optical coherence tomography and inner retinal sublayer analysis

Background/Objectives

To re-evaluate diabetic papillopathy using optical coherence tomography (OCT) for quantitative analysis of the peripapillary retinal nerve fibre layer (pRNFL), macular ganglion cell layer (mGCL) and inner nuclear layer (mINL) thickness.

Subjects/Methods

In this retrospective observational case series between June 2008 and July 2019 at Moorfields Eye hospital, 24 eyes of 22 patients with diabetes and optic disc swelling with confirmed diagnosis of NAION or diabetic papillopathy by neuro-ophthalmological assessment were included for evaluation of the pRNFL, mGCL and mINL thicknesses after resolution of optic disc swelling.

Results

The mean age of included patients was 56.5 (standard deviation (SD) ± 14.85) years with a mean follow-up duration of 216 days. Thinning of pRNFL (mean: 66.26, SD ± 31.80 µm) and mGCL (mean volume: 0.27 mm³, SD ± 0.09) were observed in either group during follow-up, the mINL volume showed no thinning with 0.39 ± 0.05 mm³. The mean decrease in visual acuity was 4.13 (SD ± 14.27) ETDRS letters with a strong correlation between mGCL thickness and visual acuity (rho 0.74, p < 0.001).

Conclusion

After resolution of acute optic disc swelling, atrophy of pRNFL and mGCL became apparent in all cases of diabetic papillopathy and diabetic NAION, with preservation of mINL volumes. Analysis of OCT did not provide a clear diagnostic distinction between both entities. We suggest a diagnostic overlay with the degree of pRNFL and mGCL atrophy of prognostic relevance for poor visual acuity independent of the semantics of terminology.