Variability in Grading Diabetic Retinopathy Using Retinal Photography and Its Comparison with an Automated Deep Learning Diabetic Retinopathy Screening Software.

BACKGROUND: Diabetic retinopathy (DR) screening using colour retinal photographs is cost-effective and time-efficient. In real-world clinical settings, DR severity is frequently graded by individuals of different expertise levels. We aim to determine the agreement in DR severity grading between human graders of varying expertise and an automated deep learning DR screening software (ADLS). METHODS: Using the International Clinical DR Disease Severity Scale, two hundred macula-centred fundus photographs were graded by retinal specialists, ophthalmology residents, family medicine physicians, medical students, and the ADLS. Based on referral urgency, referral grading was divided into no referral, non-urgent referral, and urgent referral to an ophthalmologist. Inter-observer and intra-group variations were analysed using Gwet's agreement coefficient, and the performance of ADLS was evaluated using sensitivity and specificity. RESULTS: The agreement coefficient for inter-observer and intra-group variability ranged from fair to very good, and moderate to good, respectively. The ADLS showed a high area under curve of 0.879, 0.714, and 0.836 for non-referable DR, non-urgent referable DR, and urgent referable DR, respectively, with varying sensitivity and specificity values. CONCLUSION: Inter-observer and intra-group agreements among human graders vary widely, but ADLS is a reliable and reasonably sensitive tool for mass screening to detect referable DR and urgent referable DR.

Teleglaucoma: improving access and efficiency for glaucoma care.

Teleglaucoma is the application of telemedicine for glaucoma. We review and present the current literature on teleglaucoma; present our experience with teleglaucoma programs in Alberta, Canada and Western Australia; and discuss the challenges and opportunities in this emerging field. Teleglaucoma is a novel area that was first explored a little over a decade ago and early studies highlighted the technical challenges of delivering glaucoma care remotely. Advanced technologies have since emerged that show great promise in providing access to underserviced populations. Additionally, these technologies can improve the efficiency of healthcare systems burdened with an increasing number of patients with glaucoma, and a limited supply of ophthalmologists. Additional benefits of teleglaucoma systems include e-learning and e-research. Further work is needed to fully validate and study the cost and comparative effectiveness of this approach relative to traditional models of healthcare.

MEDLINE versus EMBASE and CINAHL for telemedicine searches.

INTRODUCTION: Researchers in the domain of telemedicine throughout the world tend to search multiple bibliographic databases to retrieve the highest possible number of publications when conducting review projects. Medical Literature Analysis and Retrieval System Online (MEDLINE), Excerpta Medica Database (EMBASE), and Cumulative Index to Nursing and Allied Health Literature (CINAHL) are three popular databases in the discipline of biomedicine that are used for conducting reviews. Access to the MEDLINE database is free and easy, whereas EMBASE and CINAHL are not free and sometimes not easy to access for researchers in small research centers. OBJECTIVE: This project sought to compare MEDLINE with EMBASE and CINAHL to estimate what proportion of potentially relevant publications would be missed when only MEDLINE is used in a review project, in comparison to when EMBASE and CINAHL are also used. METHODS: Twelve simple keywords relevant to 12 different telemedicine applications were searched using all three databases, and the results were compared. RESULTS: About 9%-18% of potentially relevant articles would have been missed if MEDLINE had been the only database used. CONCLUSIONS: It is preferable if all three or more databases are used when conducting a review in telemedicine. Researchers from developing countries or small research institutions could rely on only MEDLINE, but they would loose 9%-18% of the potentially relevant publications. Searching MEDLINE alone is not ideal, but in a resource-constrained situation, it is definitely better than nothing.

A survey of the state of telemedicine in Western Australia.

To provide a baseline for future planning we conducted a survey of the use of telemedicine in the state of Western Australia. A questionnaire was designed, validated and posted to all public and private health-care facilities in Western Australia (metropolitan hospitals, country hospitals and nursing posts). Of the 132 questionnaires sent out, 102 were returned (77% response rate). Seven of the responding hospitals (all public) were classified as mainly providers of telehealth services and 95 (both public and private) were mainly receivers. Of these 95 receivers of services, 58 facilities (61%) reported that they had access to videoconferencing for telehealth purposes. The most common purposes for which videoconferencing was used were reported to be education (76% of those using videoconferencing), wound care (55%) and psychiatry (53%). The most common store-and-forward application was tele-ECG, which was reported by more than half (54%) of respondents. Eighty-five percent of public health-care facilities reported the use of telehealth (either videoconferencing or store-and-forward) in comparison with 24% of those in the private sector. There was a tendency for facilities further away from Perth to be users of more telehealth services: the correlation between distance and the number of telehealth services received was significant (Spearman's rho = 0.54, P < 0.001). The survey showed that about 73% of responding health-care facilities in Western Australia were using telemedicine.

Automated determination of distance visual acuity: towards teleophthalmology services.

PURPOSE: The aim of this study was to evaluate the visual acuity scores of a computer-based visual function testing (CVFT) system. The automated chart uses logMAR notation (logarithm of the minimum angle of resolution) and randomised tilting E as optotype. The test was conducted using a laptop computer. METHODS: This prospective, clinic-based, test validation study was based on a model of repeatability in two observations. The right eyes of a total of 104 participants were randomly tested with both CVFT and the manual, gold standard Early Treatment of Diabetic Retinopathy Study (ETDRS) chart. RESULTS: The mean difference between CVFT and ETDRS was -0.01+/-0.11 (SD) logMAR units. The Spearman Rank correlation coefficient (r) between the absolute difference and the average was 0.35. The average time for testing one eye was 54.7+/-11.89 seconds. The mean difference between two consecutive, repeated CVFT tests in the same eye was 0.02+/-0.07 logMAR units (paired t-test p=0.046). CONCLUSIONS: A novel, automated visual acuity testing application is reported. This automated system should be a useful tool for mass screening of visual impairment in rural and remote regions with its ease of use and shorter testing time.

Glaucoma screening: analysis of conventional and telemedicine-friendly devices.

PURPOSE: Portable, telemedicine-friendly devices offer novel opportunity for screening and monitoring glaucoma in the remote and rural regions of the world. This study examines the effective combination of telemedicine-friendly screening devices for detection of glaucoma in relation with conventional, hospital-based devices. METHODS: A total of 399 eyes were screened with telemedicine-friendly devices and conventional, hospital-based devices such as ophthalmoscope, tonometer and perimeter. RESULTS: Combination of age and family history of glaucoma alone has a sensitivity of 35.6% (specificity 94.2%, area under the curve 0.81, correctly classified 81.1%) and an addition of telemedicine-friendly or conventional visual field tests optimized the sensitivity to 91.1% (specificity 93.6%, area under the curve 0.95, correctly classified 93%). Analysis indicates good agreement between vertical cup-to-disc ratio by ophthalmoscopy and digital image reading. An addition of intraocular pressure test does not change sensitivity (35.6%) and specificity (94.2%). CONCLUSION: This study indicates that evaluations of cup-to-disc ratio and visual field, using telemedicine-friendly devices, are most useful tools in screening for glaucoma. When used together these devices may be an alternative for conventional glaucoma screenings.

Telemedicine-friendly, portable tonometers: an evaluation for intraocular pressure screening.

PURPOSE: To evaluate the intraocular pressure (IOP) readings from two portable, telemedicine-friendly tonometers for suitability in glaucoma screening. METHODS: 213 eyes of 107 consenting patients attending an eye clinic were tested with an I-care tonometer and a Pulsair-Easy Eye puff-air tonometer. Gold standard IOP was measured with a Goldmann applanation tonometer (GAT). Effect of central corneal thickness, anterior chamber depth and refractive errors on IOP measurements were also analysed. RESULTS: The mean difference of IOP by GAT and both the portable tonometers was +/- 2.2 mmHg. The analysis indicates minimal difference between IOP readings of both the portable tonometers. The mean difference between two consecutive readings by I-care was 0.01 mmHg. Using 21 mmHg as a threshold for suspected glaucoma, both the portable digital tonometers reported a sensitivity of 38% and specificity of >95%. In the subjects studied, central corneal thickness had statistically significant influence on IOP measurements while refractive errors and anterior chamber depth had no significant influence on IOP measurements with any tonometry. CONCLUSION: The IOP readings by both portable tonometers are comparable and were within clinically acceptable range from GAT. These portable tonometers are useful tools for IOP screening.

Remote ophthalmology services: cost comparison of telemedicine and alternative service delivery options.

We calculated the cost of an established tele-ophthalmology service, from a health-provider's perspective, and compared this with the cost of three other existing eye-care service delivery options. During a 12-month study period, 118 persons took part in the tele-ophthalmology consultations between a rural clinic located approximately 900 km from the Lions Eye Institute in Perth. The variable costs of tele-ophthalmology were 166.89 dollars(Australian dollars) per patient, and the alternatives cost 445.96 dollars, 271.48 dollars and 665.44 dollars per patient. Tele-ophthalmology incurred a set-up cost of 13,340 dollars. The threshold at which tele-ophthalmology became cheaper than any of the alternative options occurred at a workload of 128 patients. Tele-ophthalmology offers a viable alternative to conventional eye-care service in rural and remote areas.

Retinal image analysis: concepts, applications and potential.

As digital imaging and computing power increasingly develop, so too does the potential to use these technologies in ophthalmology. Image processing, analysis and computer vision techniques are increasing in prominence in all fields of medical science, and are especially pertinent to modern ophthalmology, as it is heavily dependent on visually oriented signs. The retinal microvasculature is unique in that it is the only part of the human circulation that can be directly visualised non-invasively in vivo, readily photographed and subject to digital image analysis. Exciting developments in image processing relevant to ophthalmology over the past 15 years includes the progress being made towards developing automated diagnostic systems for conditions, such as diabetic retinopathy, age-related macular degeneration and retinopathy of prematurity. These diagnostic systems offer the potential to be used in large-scale screening programs, with the potential for significant resource savings, as well as being free from observer bias and fatigue. In addition, quantitative measurements of retinal vascular topography using digital image analysis from retinal photography have been used as research tools to better understand the relationship between the retinal microvasculature and cardiovascular disease. Furthermore, advances in electronic media transmission increase the relevance of using image processing in 'teleophthalmology' as an aid in clinical decision-making, with particular relevance to large rural-based communities. In this review, we outline the principles upon which retinal digital image analysis is based. We discuss current techniques used to automatically detect landmark features of the fundus, such as the optic disc, fovea and blood vessels. We review the use of image analysis in the automated diagnosis of pathology (with particular reference to diabetic retinopathy). We also review its role in defining and performing quantitative measurements of vascular topography, how these entities are based on 'optimisation' principles and how they have helped to describe the relationship between systemic cardiovascular disease and retinal vascular changes. We also review the potential future use of fundal image analysis in telemedicine.

The reengineering of a software system for glaucoma analysis.

Glaucoma is a destructive eye disease that causes blindness in individuals displaying little or no symptoms. There is no cure as yet though there are treatments that can arrest its effects or slow its development. The earlier the disease is detected, the more likely the treatment will be successful; however early detection of the disease can be difficult. This highlights the importance of ophthalmologists having access to tools that can assist in accurately diagnosing glaucoma and other retinal diseases as early as possible. The stereo optic disc analyser (SODA) software package is a tool intended to be used by ophthalmologists, to aid in the accurate detection of retinal diseases. SODA will use stereoscopy and three-dimensional image analysis to assist in accurately detecting changes in the retina, caused by diseases such as glaucoma. This paper will focus on the reengineering and redesign of the SODA software package to overcome the shortcomings inherent in its prototype implementation and develop a package that can be commercialised. Software Engineering principles and the software development lifecycle, along with principles of object-orientation and usability, have been used to establish a framework for SODA, improve its accuracy, enhance its usability and to redevelop the product into an implementation that can later be commercialised.