A Deep Learning Approach for Semantic Segmentation of Gonioscopic Images to Support Glaucoma Categorization

Peroni, Andrea; Cutolo, Carlo Alberto; Pinto, Luís Abegão; Paviotti, Anna; Campigotto, Mauro; Cobb, Caroline; Gong, Jacintha; Patel, Sirjhun; Tatham, Andrew J.; Gillan, Stewart; Trucco, Emanuele

doi:10.1007/978-3-030-52791-4_29

Cited by 6 publications

(13 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The recent availability of new semi-automatic imaging devices 8 , 9 for gonioscopy has the potential to address current limitations and offers an unprecedented opportunity for the development of automated image analysis software, such as machine and deep learning algorithms, to support assisted diagnoses or to present augmented data to clinicians 10 (Cappellari L, et al IOVS . 2020;61:ARVO E-Abstract 1620).…”

Section: Introductionmentioning

confidence: 99%

On Clinical Agreement on the Visibility and Extent of Anatomical Layers in Digital Gonio Photographs

Peroni

Paviotti²,

Campigotto³

et al. 2021

Trans. Vis. Sci. Tech.

Self Cite

View full text Add to dashboard Cite

Purpose To quantitatively evaluate the inter-annotator variability of clinicians tracing the contours of anatomical layers of the iridocorneal angle on digital gonio photographs, thus providing a baseline for the validation of automated analysis algorithms. Methods Using a software annotation tool on a common set of 20 images, five experienced ophthalmologists highlighted the contours of five anatomical layers of interest: iris root (IR), ciliary body band (CBB), scleral spur (SS), trabecular meshwork (TM), and cornea (C). Inter-annotator variability was assessed by (1) comparing the number of times ophthalmologists delineated each layer in the dataset; (2) quantifying how the consensus area for each layer (i.e., the intersection area of observers’ delineations) varied with the consensus threshold; and (3) calculating agreement among annotators using average per-layer precision, sensitivity, and Dice score. Results The SS showed the largest difference in annotation frequency (31%) and the minimum overall agreement in terms of consensus size (∼28% of the labeled pixels). The average annotator's per-layer statistics showed consistent patterns, with lower agreement on the CBB and SS (average Dice score ranges of 0.61–0.7 and 0.73–0.78, respectively) and better agreement on the IR, TM, and C (average Dice score ranges of 0.97–0.98, 0.84–0.9, and 0.93–0.96, respectively). Conclusions There was considerable inter-annotator variation in identifying contours of some anatomical layers in digital gonio photographs. Our pilot indicates that agreement was best on IR, TM, and C but poorer for CBB and SS. Translational Relevance This study provides a comprehensive description of inter-annotator agreement on digital gonio photographs segmentation as a baseline for validating deep learning models for automated gonioscopy.

show abstract

Section: Introductionmentioning

confidence: 99%

On Clinical Agreement on the Visibility and Extent of Anatomical Layers in Digital Gonio Photographs

Peroni

Paviotti²,

Campigotto³

et al. 2021

Trans. Vis. Sci. Tech.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Digital gonioscopy has been shown useful in documenting post-operative and abnormal findings of the ACA and has a great potential for teaching gonioscopy. Deep learning algorithms for automatic classification of the angle are on the horizon and their validity in the clinical setting should be accurately evaluated [31][32][33]. Informed Consent Statement: Written informed consent has been obtained from the patients to publish images in this paper.…”

Section: Discussionmentioning

confidence: 99%

“…Using a software annotation tool that permitted tracing the contours of anatomical layers, Peroni et al found that scleral spur had the minimum agreement whereas it was best on iris root, trabecular meshwork and cornea [ 31 ]. Based on these studies, deep learning semantic segmentation algorithms for processing images acquired by GS-1 are currently in development [ 32 , 33 , 34 ].…”

Section: Methodsmentioning

confidence: 99%

Moving beyond the Slit-Lamp Gonioscopy: Challenges and Future Opportunities

et al. 2021

Self Cite

View full text Add to dashboard Cite

After almost a century from its introduction in clinical practice, slit-lamp gonioscopy is still considered the reference standard for evaluating the anterior chamber angle (ACA). Gonioscopy is essential for diagnosing angle closure disease, and ACA features are included in glaucoma’s diagnostics and treatments algorithms. However, shortcomings of slit-lamp gonioscopy include a steep learning curve, lack of agreement between examiners and poor documentation. Thanks to advances in miniaturization and computing, new instruments for digital gonioscopy have been developed and marketed. This narrative review focuses on the Gonioscope GS-1, which permits semi-automated circumferential documentation of the ACA in real-colour photographs. Advantages and disadvantages of GS-1 compared with slit-lamp gonioscopy and other ACA imaging technologies such as optical coherence tomography are discussed. Finally, potential opportunities offered by this device for telemedicine, virtual clinics, and automatic classification with deep learning are presented.

show abstract

“…Others have attempted to build segmentation algorithms of angle structures on goniophotographs. Peroni et al123 performed semantic segmentation of the anatomic structure of the ACA with a DL system,129 achieving ~88% of average pixel classification accuracy in a 5-fold cross-validation on a very limited size annotated image data set. Subsequently, in 2021, Peroni et al128 continued to develop and test a new DL model.…”

Section: Methodsmentioning

confidence: 99%

“…132 Second, these AI tools may not perform satisfactorily in complicated clinical scenarios. 116,117,119,123 In the study conducted by Peroni et al, the DL classification system often misclassified certain shadows as targets when the cornea was in direct contact with the iris, thus leading to segmentation failure. 123 In the study by Baskaran et al 117 , the automatic grading system incorrectly identified angle closure if most of the open angles in the gonio-images had very slight trabecular meshwork pigmentation or dense pigmentation.…”

Section: Goniophotographs: Challengesmentioning

confidence: 99%

Applications of Artificial Intelligence and Deep Learning in Glaucoma

Chen

Anran

Tan

et al. 2023

Asia-Pacific Journal of Ophthalmology

View full text Add to dashboard Cite

Diagnosis and detection of progression of glaucoma remains challenging. Artificial intelligence-based tools have the potential to improve and standardize the assessment of glaucoma but development of these algorithms is difficult given the multimodal and variable nature of the diagnosis. Currently, most algorithms are focused on a single imaging modality, specifically screening and diagnosis based on fundus photos or optical coherence tomography images. Use of anterior segment optical coherence tomography and goniophotographs is limited. The majority of algorithms designed for disease progression prediction are based on visual fields. No studies in our literature search assessed the use of artificial intelligence for treatment response prediction and no studies conducted prospective testing of their algorithms. Additional challenges to the development of artificial intelligence-based tools include scarcity of data and a lack of consensus in diagnostic criteria. Although research in the use of artificial intelligence for glaucoma is promising, additional work is needed to develop clinically usable tools.

show abstract

A Deep Learning Approach for Semantic Segmentation of Gonioscopic Images to Support Glaucoma Categorization

Cited by 6 publications

References 20 publications

On Clinical Agreement on the Visibility and Extent of Anatomical Layers in Digital Gonio Photographs

On Clinical Agreement on the Visibility and Extent of Anatomical Layers in Digital Gonio Photographs

Moving beyond the Slit-Lamp Gonioscopy: Challenges and Future Opportunities

Applications of Artificial Intelligence and Deep Learning in Glaucoma

Contact Info

Product

Resources

About