Predicting 10-2 Visual Field From Optical Coherence Tomography in Glaucoma Using Deep Learning Corrected With 24-2/30-2 Visual Field

Hashimoto, Yukihiko; Kiwaki, Taichi; Sugiura, Hiroki; Asano, Shotaro; Murata, Hiroshi; Fujino, Yuri; Miki, Atsuya; Mori, Kazuhiko; Ikeda, Yoko; Kanamoto, Takashi; Yamagami, Junkichi; Inoue, Kenji; Tanito, Masaki; Yamanishi, Kenji; Asaoka, Ryo

doi:10.1167/tvst.10.13.28

Cited by 17 publications

(14 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a similar study, Hashimoto et al combined the pattern-based regularization and the 24-2 SAP correction. 24 They found a MAE of 5.3 dB with pattern-based regularization alone versus 4.2 dB when combined with 24-2 SAP correction.…”

Section: Prediction Of 10-2 Visual Field Sensitivity Threshold Valuesmentioning

confidence: 97%

A review of deep learning in structure and function in glaucoma

Mariottoni¹,

Medeiros²,

Costa³

2022

MAIO

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 10-2 Visual Field Sensitivity Threshold Valuesmentioning

confidence: 97%

A review of deep learning in structure and function in glaucoma

Mariottoni¹,

Medeiros²,

Costa³

2022

MAIO

View full text Add to dashboard Cite

show abstract

“…A very attractive approach is to use AI to combine structural information with functional outcomes. Glaucoma diagnosis is improved when using both ONH parameters and VF outcomes: algorithms capable of predicting 10-2 VF parameters from macular OCT scans, and 24-1 VF parameters from both macular and ONH OCT scans have been developed (362)(363)(364)(365). Also, various teams (Lazaridis et (366)(367)(368)(369)(370).…”

Section: Artificial Intelligence and Glaucomamentioning

confidence: 99%

The Development and Clinical Application of Innovative Optical Ophthalmic Imaging Techniques

et al. 2022

View full text Add to dashboard Cite

The field of ophthalmic imaging has grown substantially over the last years. Massive improvements in image processing and computer hardware have allowed the emergence of multiple imaging techniques of the eye that can transform patient care. The purpose of this review is to describe the most recent advances in eye imaging and explain how new technologies and imaging methods can be utilized in a clinical setting. The introduction of optical coherence tomography (OCT) was a revolution in eye imaging and has since become the standard of care for a plethora of conditions. Its most recent iterations, OCT angiography, and visible light OCT, as well as imaging modalities, such as fluorescent lifetime imaging ophthalmoscopy, would allow a more thorough evaluation of patients and provide additional information on disease processes. Toward that goal, the application of adaptive optics (AO) and full-field scanning to a variety of eye imaging techniques has further allowed the histologic study of single cells in the retina and anterior segment. Toward the goal of remote eye care and more accessible eye imaging, methods such as handheld OCT devices and imaging through smartphones, have emerged. Finally, incorporating artificial intelligence (AI) in eye images has the potential to become a new milestone for eye imaging while also contributing in social aspects of eye care.

show abstract

“…ResNet and VGG are two popular CNN models that have shown remarkable progress in medical image analysis for smart medicine [27]. In the literature, these two models are commonly used in VF image prediction for glaucoma detection and progression grading [28][29][30][31][32][33][34]. VGG16 and ResNet18 consist of 41 and 71 layers and use 224×224-pixel size images in the input layers.…”

Section: Deep Learning Modelsmentioning

confidence: 99%

Glaucoma Detection and Staging from Visual Field Images using Machine Learning Techniques

Akter

Jack

et al. 2022

Preprint

View full text Add to dashboard Cite

Purpose: In this study, we investigated the performance of deep learning (DL) models to differentiate between normal and glaucomatous visual fields (VFs) and classify glaucoma from early to the advanced stage to observe if the DL model can stage glaucoma as Mills criteria using only the pattern deviation (PD) plots. The DL model results were compared with a machine learning (ML) classifier trained on conventional VF parameters. Methods: A total of 265 PD plots and 265 numerical datasets of Humphrey 24-2 VF images were collected from 119 normal and 146 glaucomatous eyes to train the DL models to classify the images into four groups: normal, early glaucoma, moderate glaucoma, and advanced glaucoma. The two popular pre-trained DL models: ResNet18 and VGG16, were used to train the PD images using five-fold cross-validation (CV) and observed the performance using balanced, pre-augmented data (n= 476 images), imbalanced original data (n=265) and feature extraction. The trained images were further investigated using the Grad-CAM visualization technique. Moreover, four ML models were trained from the global indices: mean deviation (MD), pattern standard deviation (PSD) and visual field index (VFI), using five-fold CV to compare the classification performance with the DL models result. Results: The DL model, ResNet18 trained from balanced, pre-augmented PD images, achieved high accuracy in classifying the groups with an overall F1-score: 96.8%, precision: 97.0%, recall: 96.9%, and specificity: 99.0%. The highest F1 score was 87.8% for ResNet18 with the original dataset and 88.7% for VGG16 with feature extraction. The DL models successfully localized the affected VF loss in PD plots. Among the ML models, the random forest (RF) classifier performed best with an F1 score of 96%. Conclusion: The DL model trained from PD plots was promising in differentiating normal and glaucomatous groups and performed similarly to conventional global indices. Hence, the evidence-based DL model trained from PD images demonstrated that the DL model could stage glaucoma using only PD plots like Mills criteria. This automated DL model will assist clinicians in precision glaucoma detection and progression management during extensive glaucoma screening.

show abstract

Predicting 10-2 Visual Field From Optical Coherence Tomography in Glaucoma Using Deep Learning Corrected With 24-2/30-2 Visual Field

Cited by 17 publications

References 31 publications

A review of deep learning in structure and function in glaucoma

A review of deep learning in structure and function in glaucoma

The Development and Clinical Application of Innovative Optical Ophthalmic Imaging Techniques

Glaucoma Detection and Staging from Visual Field Images using Machine Learning Techniques

Contact Info

Product

Resources

About