RetiNerveNet: using recursive deep learning to estimate pointwise 24-2 visual field data based on retinal structure

Datta, Siddhartha; Mariottoni, Eduardo B.; Dov, David; Jammal, Alessandro A.; Carin, Lawrence; Medeiros, Felipe A.

doi:10.1038/s41598-021-91493-9

Cited by 15 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With a similar approach, Datta et al developed a recurrent neural network (RNN) that analyzed the circumpapillary RNFL thickness, but used anatomic knowledge to improve predictions. 14 Their results were similar to those described by Mariottoni and colleagues.…”

Section: Prediction Of 24-2 Visual Field Sensitivity Threshold Valuessupporting

confidence: 89%

A review of deep learning in structure and function in glaucoma

Mariottoni¹,

Medeiros²,

Costa³

2022

MAIO

Self Cite

View full text Add to dashboard Cite

The relationship between structural damage and functional loss in glaucoma is of great importance for its diagnosis and management. The functional status is usually assessed through visual field examination, a subjective test that is burdensome and time-consuming. Moreover, it depends on patients’ answers and there is a learning curve until accurate and reliable measurements are possible. Structural assessment, on the other hand, has remarkably improved since the development of optical coherence tomography, a fast test that allows for objective and quantitative analysis of retinal layers. The relationship between the two tests, however, is complex and nonlinear, and is influenced by interindividual variability. Thus, qualitative evaluation or the use of conventional statistics might not be appropriate. In recent years, we have seen a remarkable evolution of artificial intelligence algorithms and deep learning models. These techniques have proved adequate to model such complicated relationships. In this review, we summarize studies that investigate the structure and function relationship in glaucoma making use of artificial intelligence and deep learning, the challenges associated with predicting visual field information from structural measurements, and the strategies adopted to improve their accuracy.

show abstract

Section: Prediction Of 24-2 Visual Field Sensitivity Threshold Valuessupporting

confidence: 89%

A review of deep learning in structure and function in glaucoma

Mariottoni¹,

Medeiros²,

Costa³

2022

MAIO

Self Cite

View full text Add to dashboard Cite

show abstract

“…Other lines of research may study different layers and may also have different devices and methods to measure RT and VF. For example, Park et al 13 studied GCIPL and RNFL that were measured in a different way from our data, and Datta et al 14 only studied RNFL.…”

Section: Glaucoma Estimationmentioning

confidence: 57%

A Two-Phase Deep Learning Approach to Glaucoma Estimation Using Retinal Layers Thickness

Xu,

Asaoka,

Cao

et al. 2023

Preprint

View full text Add to dashboard Cite

Glaucoma, which makes progressive and irreversible sight damage to human eyes, is the second leading cause of blindness worldwide. The damage is principally estimated by visual field (VF) sensitivity through costly visual field tests. To achieve a less costly estimation, a promising method is to first measure retinal layers thickness (RT) by optical coherence tomography and then map RT into VF. There are some recent studies showing that the mapping can be effectively learned by convolutional neural networks (CNNs). However, different stages of glaucoma, e.g., early stage and severe stage, that result in different relations between VF and RT, have not yet been distinguished in the learning. Consequently, the learned mapping may be a poor fit for each of these stages especially when training data are small-scale. In this paper we propose for the first time an approach that distinguishes two different glaucoma stages in learning the mapping. In particular, our approach is two-phase, the first of which is to estimate the probability of an RT being a particular stage and the second is to learn stage-specific mappings. For the probability estimation, we employ a CNN based classification model. We design a set of data augmentation techniques for training as well as a test-time augmentation technique for testing by referring to the domain knowledge of the data. To learn stage-specific mappings, we design a regression model with a novel two-branch architecture. Moreover, we introduce a novel attention mechanism to CNN for both the classification and the regression. Through experiments on a real-world dataset, we demonstrate that our proposed method outperforms existing methods by 6.82% in terms of mean of the root mean square error for early stage of glaucoma.

show abstract

“…Various studies focused on predicting threshold sensitivity values in 24-2 standard automated perimetry from segmented OCT images. [24][25][26][27][28] While the majority of these studies used SD-OCT imaging, Park et al used swept-source OCT images, and the root mean squared error of the global prediction error for their model was 4.44 dB. 25 images to predict each of the 52 sensitivity points on the 24-2 VF, and their model had an MAE of 0.485 (0.438-0.533).…”

Section: Glaucomamentioning

confidence: 99%

Predictive Deep Learning Applications in Ophthalmology

Koseoglu,

Liu

2023

touchREVIEWS in Ophthalmology

View full text Add to dashboard Cite

Therapies for retinal diseases have witnessed remarkable progress, particularly in the domain of drug delivery routes. Intravitreal injections have especially revolutionized the treatment landscape for various proliferative retinopathies. The emergence of suprachoroidal space as a novel drug delivery site is poised to overcome challenges seen in optimizing drug penetration, targeting, and bioavailability. This approach holds promise for delivering various agents, including steroids, anti-vascular endothelial growth factor molecules, gene vectors, and small molecules, with multiple clinical trials demonstrating this delivery method's potential to reduce treatment burdens and optimize visual outcomes. Particularly, small molecule therapy for retinal diseases finds ease of application through the suprachoroidal route in that small molecule agents have a wide range of targets in disease pathway, greater tissue penetration, and cost-effectiveness in production. The suprachoroidal administration of axitinib, a tyrosine kinase inhibitor, has shown promise in the OASIS trial for treating neovascular age-related macular degeneration. Integrin antagonists and plasma Kallikrein-kinin system inhibitors, also delivered suprachoroidally, exhibit targeted distribution and potential for long-acting therapy.

show abstract

RetiNerveNet: using recursive deep learning to estimate pointwise 24-2 visual field data based on retinal structure

Cited by 15 publications

References 28 publications

A review of deep learning in structure and function in glaucoma

A review of deep learning in structure and function in glaucoma

A Two-Phase Deep Learning Approach to Glaucoma Estimation Using Retinal Layers Thickness

Predictive Deep Learning Applications in Ophthalmology

Contact Info

Product

Resources

About