1
|
Diener R, Renz AW, Eckhard F, Segbert H, Eter N, Malcherek A, Biermann J. Automated Classification of Physiologic, Glaucomatous, and Glaucoma-Suspected Optic Discs Using Machine Learning. Diagnostics (Basel) 2024; 14:1073. [PMID: 38893600 PMCID: PMC11171940 DOI: 10.3390/diagnostics14111073] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/15/2024] [Revised: 05/09/2024] [Accepted: 05/17/2024] [Indexed: 06/21/2024] Open
Abstract
In order to generate a machine learning algorithm (MLA) that can support ophthalmologists with the diagnosis of glaucoma, a carefully selected dataset that is based on clinically confirmed glaucoma patients as well as borderline cases (e.g., patients with suspected glaucoma) is required. The clinical annotation of datasets is usually performed at the expense of the data volume, which results in poorer algorithm performance. This study aimed to evaluate the application of an MLA for the automated classification of physiological optic discs (PODs), glaucomatous optic discs (GODs), and glaucoma-suspected optic discs (GSODs). Annotation of the data to the three groups was based on the diagnosis made in clinical practice by a glaucoma specialist. Color fundus photographs and 14 types of metadata (including visual field testing, retinal nerve fiber layer thickness, and cup-disc ratio) of 1168 eyes from 584 patients (POD = 321, GOD = 336, GSOD = 310) were used for the study. Machine learning (ML) was performed in the first step with the color fundus photographs only and in the second step with the images and metadata. Sensitivity, specificity, and accuracy of the classification of GSOD vs. GOD and POD vs. GOD were evaluated. Classification of GOD vs. GSOD and GOD vs. POD performed in the first step had AUCs of 0.84 and 0.88, respectively. By combining the images and metadata, the AUCs increased to 0.92 and 0.99, respectively. By combining images and metadata, excellent performance of the MLA can be achieved despite having only a small amount of data, thus supporting ophthalmologists with glaucoma diagnosis.
Collapse
Affiliation(s)
- Raphael Diener
- Department of Ophthalmology, University of Muenster Medical Center, 48149 Muenster, Germany; (H.S.); (N.E.); (J.B.)
| | - Alexander W. Renz
- Department of Informatics, University of Applied Sciences Darmstadt, 64295 Darmstadt, Germany; (A.W.R.); (A.M.)
| | - Florian Eckhard
- Department of Informatics, Technical University of Munich, 80333 Munich, Germany;
| | - Helmar Segbert
- Department of Ophthalmology, University of Muenster Medical Center, 48149 Muenster, Germany; (H.S.); (N.E.); (J.B.)
| | - Nicole Eter
- Department of Ophthalmology, University of Muenster Medical Center, 48149 Muenster, Germany; (H.S.); (N.E.); (J.B.)
| | - Arnim Malcherek
- Department of Informatics, University of Applied Sciences Darmstadt, 64295 Darmstadt, Germany; (A.W.R.); (A.M.)
| | - Julia Biermann
- Department of Ophthalmology, University of Muenster Medical Center, 48149 Muenster, Germany; (H.S.); (N.E.); (J.B.)
| |
Collapse
|
2
|
Salvetat ML, Pellegrini F, Spadea L, Salati C, Zeppieri M. Pharmaceutical Approaches to Normal Tension Glaucoma. Pharmaceuticals (Basel) 2023; 16:1172. [PMID: 37631087 PMCID: PMC10458083 DOI: 10.3390/ph16081172] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/05/2023] [Revised: 08/14/2023] [Accepted: 08/15/2023] [Indexed: 08/27/2023] Open
Abstract
Normal tension glaucoma (NTG) is defined as a subtype of primary open-angle glaucoma (POAG) in which the intraocular pressure (IOP) values are constantly within the statistically normal range without treatment and represents approximately the 30-40% of all glaucomatous cases. The pathophysiology of this condition is multifactorial and is still not completely well known. Several theories have been proposed to explain the onset and progression of this disease, which can be divided into IOP-dependent and IOP-independent factors, suggesting different therapeutic strategies. The current literature strongly supports the fundamental role of IOP in NTG. The gold standard treatment for NTG tends to be based on the lowering IOP even if "statistically normal". Numerous studies have shown, however, that the IOP reduction alone is not enough to slow down or stop the disease progression in all cases, suggesting that other IOP-independent risk factors may contribute to the NTG pathogenesis. In addition to IOP-lowering strategies, several different therapeutic approaches for NTG have been proposed, based on vaso-active, antioxidant, anti-inflammatory and/or neuroprotective substances. To date, unfortunately, there are no standardized or proven treatment alternatives for NTG when compared to traditional IOP reduction treatment regimes. The efficacy of the IOP-independent strategies in decreasing the risk or treating NTG still remains inconclusive. The aim of this review is to highlight strategies reported in the current literature to treat NTG. The paper also describes the challenges in finding appropriate and pertinent treatments for this potentially vision-threatening disease. Further comprehension of NTG pathophysiology can help clinicians determine when to use IOP-lowering treatments alone and when to consider additional or alternatively individualized therapies focused on particular risk factors, on a case-by-case basis.
Collapse
Affiliation(s)
- Maria Letizia Salvetat
- Department of Ophthalmology, Azienda Sanitaria Friuli Occidentale, 33170 Pordenone, Italy
| | - Francesco Pellegrini
- Department of Ophthalmology, Azienda Sanitaria Friuli Occidentale, 33170 Pordenone, Italy
| | - Leopoldo Spadea
- Eye Clinic, Policlinico Umberto I, “Sapienza” University of Rome, 00142 Rome, Italy
| | - Carlo Salati
- Department of Ophthalmology, University Hospital of Udine, 33100 Udine, Italy
| | - Marco Zeppieri
- Department of Ophthalmology, University Hospital of Udine, 33100 Udine, Italy
| |
Collapse
|
3
|
Screening fundus photography predicts and reveals risk factors for glaucoma conversion in eyes with large optic disc cupping. Sci Rep 2023; 13:81. [PMID: 36596820 PMCID: PMC9810728 DOI: 10.1038/s41598-022-26798-4] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/15/2022] [Accepted: 12/20/2022] [Indexed: 01/04/2023] Open
Abstract
This study aimed to investigate the risk factors for glaucoma conversion and progression in eyes with large optic disc cupping without retinal nerve fiber layer defect (RNFLD). Five hundred forty-two eyes of 271 subjects who had a vertical cup-to-disc ratio (CDR) ≥ 0.6 without RNFLD were enrolled. Characteristics for optic disc configuration (including CDR, vertical cupping, ISNT rule, disc ovality, peripapillary atrophy [PPA]-to-disc area [DA] ratio, and lamina cribrosa pore visibility) and blood vessels (including central retinal vessel trunk [CRVT] nasalization, bayoneting of vessels, baring of circumlinear vessels, history of disc hemorrhage [DH] and vessel narrowing/sclerotic change) were evaluated. From a median follow-up of 11.3 years, 26.6% of eyes (n = 144) developed RNFLD within a median of 5.1 years. Baseline factors, including vertical CDR ≥ 0.7 (hazard ratio [HR] = 2.12), vertical cupping (HR = 1.93), ISNT rule violation (HR = 2.84), disc ovality ≥ 1.2 (HR = 1.61), PPA-to-DA ratio ≥ 0.4 (HR = 1.77), CRVT nasalization ≥ 60% (HR = 1.77), vessel narrowing/sclerotic change (HR = 2.13), DH history (HR = 5.60), and baseline intraocular pressure ≥ 14 mmHg (HR = 1.70) were significantly associated with glaucoma conversion (all Ps < 0.05). An HR-matched scoring system based on initial fundus photography predicted glaucoma conversion with specificity of 90.4%. Careful examination of the optic nerve head and vascular structures can help to predict the risk of glaucoma conversion in eyes with large optic disc cupping.
Collapse
|
4
|
Zhou L, Zhan W, Wei X. Clinical pharmacology and pharmacogenetics of prostaglandin analogues in glaucoma. Front Pharmacol 2022; 13:1015338. [PMID: 36313286 PMCID: PMC9596770 DOI: 10.3389/fphar.2022.1015338] [Citation(s) in RCA: 11] [Impact Index Per Article: 5.5] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/09/2022] [Accepted: 09/29/2022] [Indexed: 07/30/2023] Open
Abstract
Glaucoma is the main cause of irreversible visual loss worldwide, and comprises a group of progressive, age-related, and chronic optic neuropathies. Prostaglandin analogs are considered a first-line treatment in the management of glaucoma and have the best efficacy in reducing intraocular pressure. When comparing these therapeutic agents between them, long-term therapy with 0.03% bimatoprost is the most effective followed by treatment with 0.005% latanoprost and 0.004% travoprost. The prevalence of adverse events is lower for latanoprost than for other prostaglandin analogs. However, some patients do not respond to the treatment with prostaglandin analogs (non-responders). Intraocular pressure-lowering efficacy differs significantly between individuals partly owing to genetic factors. Rs1045642 in ABCB1, rs4241366 in SLCO2A1, rs9503012 in GMDS, rs10306114 in PTGS1, rs11568658 in MRP4, rs10786455 and rs6686438 in PTGFR were reported to be positive with the response to prostaglandin analogs in patients with glaucoma. A negative association was found between single nucleotide polymorphisms of PTGFR (rs11578155 and rs6672484) and the response to prostaglandin analogs in patients with glaucoma. The current review is an analysis of the information relevant to prostaglandin analog treatments based on previous literatures. It describes in detail the clinical pharmacology and pharmacogenetics of drugs belonging to this therapeutical class to provide a sound pharmacological basis for their proper use in ophthalmological clinical practice.
Collapse
Affiliation(s)
- Lin Zhou
- Department of Ophthalmology, West China Hospital, Sichuan University, Chengdu, China
| | - Wenyi Zhan
- West China School of Medicine, Sichuan University, Chengdu, China
| | - Xin Wei
- Department of Ophthalmology, West China Hospital, Sichuan University, Chengdu, China
| |
Collapse
|
5
|
Waisberg E, Micieli JA. Neuro-Ophthalmological Optic Nerve Cupping: An Overview. Eye Brain 2021; 13:255-268. [PMID: 34934377 PMCID: PMC8684388 DOI: 10.2147/eb.s272343] [Citation(s) in RCA: 19] [Impact Index Per Article: 6.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/20/2021] [Accepted: 12/02/2021] [Indexed: 11/23/2022] Open
Abstract
Optic nerve cupping or enlargement of the cup-to-disc ratio is widely recognized as a feature of glaucoma, however it may also occur in non-glaucomatous optic neuropathies. The most well-recognized non-glaucomatous optic neuropathies that cause cupping include compressive optic neuropathies, arteritic anterior ischemic optic neuropathies, hereditary optic neuropathies, and optic neuritis. Cupping is thought to consist of two main components: prelaminar and laminar thinning. The former is a shallow form of cupping and related to loss of retinal ganglion cells, whereas the latter involves damage to the lamina cribrosa and peripapillary scleral connective tissue. Differentiating glaucomatous and non-glaucomatous optic nerve cupping remains challenging even for experienced observers. Classically, the optic nerve in non-glaucomatous causes has pallor of the neuroretinal rim, but the optic nerve should not be examined in isolation. The patient’s medical history, history of presenting illness, visual function (visual acuity, color vision and visual field testing) and ocular examination also need to be considered. Ancillary testing such as optical coherence tomography of the retinal nerve fiber layer and ganglion cell layer-inner plexiform layer may also be helpful in localizing the disease. In this review, we review the non-glaucomatous causes of cupping and provide an approach to evaluating a patient that presents with an enlarged cup-to-disc ratio.
Collapse
Affiliation(s)
- Ethan Waisberg
- UCD School of Medicine, University College Dublin, Belfield, Dublin, 4, Ireland
| | - Jonathan A Micieli
- Department of Ophthalmology and Vision Sciences, University of Toronto, Toronto, Ontario, Canada.,Kensington Vision and Research Centre, Toronto, Ontario, Canada
| |
Collapse
|
6
|
Ganesh SS, Kannayeram G, Karthick A, Muhibbullah M. A Novel Context Aware Joint Segmentation and Classification Framework for Glaucoma Detection. COMPUTATIONAL AND MATHEMATICAL METHODS IN MEDICINE 2021; 2021:2921737. [PMID: 34777561 PMCID: PMC8589492 DOI: 10.1155/2021/2921737] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 09/19/2021] [Revised: 10/19/2021] [Accepted: 10/23/2021] [Indexed: 11/17/2022]
Abstract
Glaucoma is a chronic ocular disease characterized by damage to the optic nerve resulting in progressive and irreversible visual loss. Early detection and timely clinical interventions are critical in improving glaucoma-related outcomes. As a typical and complicated ocular disease, glaucoma detection presents a unique challenge due to its insidious onset and high intra- and interpatient variabilities. Recent studies have demonstrated that robust glaucoma detection systems can be realized with deep learning approaches. The optic disc (OD) is the most commonly studied retinal structure for screening and diagnosing glaucoma. This paper proposes a novel context aware deep learning framework called GD-YNet, for OD segmentation and glaucoma detection. It leverages the potential of aggregated transformations and the simplicity of the YNet architecture in context aware OD segmentation and binary classification for glaucoma detection. Trained with the RIGA and RIMOne-V2 datasets, this model achieves glaucoma detection accuracies of 99.72%, 98.02%, 99.50%, and 99.41% with the ACRIMA, Drishti-gs, REFUGE, and RIMOne-V1 datasets. Further, the proposed model can be extended to a multiclass segmentation and classification model for glaucoma staging and severity assessment.
Collapse
Affiliation(s)
- S. Sankar Ganesh
- Department of Artificial Intelligence and Data Science, KPR Institute of Engineering and Technology, Coimbatore, 641407 Tamil Nadu, India
| | - G. Kannayeram
- Department of Electrical and Electronics Engineering, National Engineering College, Kovilpatti, 628503 Tamil Nadu, India
| | - Alagar Karthick
- Renewable Energy Lab, Department of Electrical and Electronics Engineering, KPR Institute of Engineering and Technology, Coimbatore, 641407 Tamil Nadu, India
| | - M. Muhibbullah
- Department of Electrical and Electronic Engineering, Bangladesh University, Dhaka 1207, Bangladesh
| |
Collapse
|