Dr. ABHISHEK SHEEMAR, Dr. Mahesh Shanmugam P, Dr. Mishra Divyansh Kailashcandra, Dr. Rajesh Ramanjulu
Abstract
Aim
Comparison of novel prototype device with portable non mydriatic (NM) devices in DR screening.
Purpose
To compare smartphone based NM wide field fundus imaging device (prototype device) with portable NM devices for identification of diabetic retinopathy (DR) in DR screening camps.
Methods
Trained optometrists screened 750 eyes with diabeteswith the above mentioned devices. Devices were utilized to capture NM images&the images were uploaded on to the cloud system and graded at the base hospital.
Results
The gradable images were comparable in both the devices. Prototype device was able to capture 120* of field which is 2.5 times wider than conventional NM cameras (50*). Prototype device has 30% more peripheral DR detection as compared to conventional device.
Conclusion
Prototype device is a better alternative for the present available NM portable devices in DR screening camps. The quality of images are good, providing widest field of view, identifying even peripheral DR. The device is less expensive, mobile & has a shorter learning curve.
Keywords – smartphone imaging, diabetic retinopathy, screening, non mydriatic, widefield.
Introduction
The first available fundus camera was produced by Carl Zeiss in 1926, following which significant improvements have been made in the field of view (fov).1 technological advancement has made it possible to capture 200 degrees fovof fundus. The introduction of fundus photography has impacted retinal imaging and retinal screening programs significantly. In 2017 there are estimated 451 million (age 18–99 years) people with diabetes worldwide. These figures were expected to increase to 693 million) by 2045.2 Diabetic retinopathy (DR), an eye disease secondary to diabetes, is a major cause of visual loss and blindness worldwide.3 Recently, 35% of diabetic patients have been estimated to be living with DR worldwide.4
Early detection and management of DR require an effective screening program. Access to screening is a major obstacle in developing countries where lack of resources and ophthalmologists poses a significant challenge. A common alternative is remote fundus photography also known as telemedicine or telescreening.5 which can be a cost effective alternative in countries like India. Screening programs using these remote techniques have been successful in various high- and low-resource settings, including India.6,7-11Fundus imaging with a smartphone based device offers a low cost, fast and non-invasive tool for screening DR.12Currently available smartphone based fundus imaging devices require pupil dilatation which increases the effective time for screening as well as poses a risk for eyes with suspected or definite angle closure. We have developed a novel prototype device which not only is a non-mydriatic device but also is a wide field of view as compared to other devices presently available. In this study we compared our device with otherportable NM fundus imaging device.
Methods
The study was done from Jan 2018 to May 2018. Written informed consent was taken from the patient. The patients were recruited from screening camps held by the hospital. The screening was done by trained optometrists with prototype device and other NM device on the same day. 750 eyes (375 patients) with history of diabetes mellitus were screened.
Inclusion criteria included age group 18-65 years, diagnosis of diabetes mellitus and consent to undergo fundus examination with all devices. Exclusion criteria included non-gradable images, posterior capsular opacity, very small pupil, uncooperative patient.
Patients were imaged with both devices on the same day. A relevant diabetic history including duration of DM and systemic and ocular history was noted before imaging including fasting blood sugar FBS and BP. Vison was recorded. Then fundus image was taken with both the devices centred on fovea and optic disc. The image were evaluated for clarity, if unsatisfactory a repeat image was taken.
All the images were coded and stored digitally on a cloud. The images were then reviewed at the base hospital. The images were graded according to modified ETDRS classification.The patients were classified as no DR, mild NPDR, moderate NPDR, severe NPDR and PDR. Sight threatening DR (STDR) was also classified. DME was labelled according to ETDRS criteria. The sensitivity and specificity for detecting any type of DR was calculated and compared. Statistical analysis was performed using Microsoft excel.
Results
A total of 750 (1500 eyes) patients were screened. The number of males were n=348 and females n=402. The mean age was 65.67 years (range 60-75 years).The mean duration of diabetes was 10.56 years. All patients had type 2 diabetes.
10.6% images were ungradable due to various causes like vitreous haemorrhage, dense cataract, ocular surface disease etc. the prevalence of DR based on fundus imaging by prototype device and other NM device was 64.3% and 62.7% respectively. The various types of DR captured by both the devices are depicted in table 1.
Table 1
| Prototype device% | Other NM device | |
| No DR | 35.7 | 34.3 |
| Mild DR | 5.2 | 8.1 |
| Moderate DR | 36.4 | 41.1 |
| Severe DR | 16.6 | 10.2 |
| PDR | 6.1 | 6.3 |
| DME | 12.1 | 14.2 |
| STDR | 59.1 | 57.6 |
The sensitivity and specificity of both the devices are depicted in table 2.
Table 2
| Prototype device% (sensitivity,specificity) | Other NM device(sensitivity,specificity) | |
| Any DR | 95.3,95.6 | 96.1,95.8 |
| Mild DR | 86.1,92.6 | 88.2,94.6 |
| Moderate DR | 89.3,94.2 | 93.0,93.1 |
| Severe DR | 88.2,92.8 | 90.8,92.3 |
| PDR | 90.1,94.9 | 90.0,93.3 |
| DME | 89.2,93.1 | 92.1,93.6 |
| STDR | 90.1,92.7 | 93.1,94.1 |
The field of view in prototype device is 120 degrees which is about 2.5 times more than the other device. Due to wider field of view peripheral detection rate was 30% more in prototype device.
The non-gradable images were more in prototype device n=30 as compared to n=20 in other NM device.
Discussion
Currently many portable NM devices are available for DR screening. It serves as a useful tool in peripheral areas where service of a trained ophthalmologist may not be available. It serves to bridge this gap. Images from a peripheral centre can be directly viewed by an ophthalmologist at a base hospital and the patient can be referred for further management.
Our prototype device had sensitivity and specificity comparable to other portable NM devices currently available. The sensitivity and specificity for detecting STDR and DME was in good agreement with other NM devices 90.1,92.7%. The sensitivity and specificity for screening any type of DR was 95.3 and 95.6% which is comparable to other NM devices.PDR was detected in 6.1% of eyes as compared to 6.3% in other NM devices with similar sensitivity and specificity. Early detection of PDR and DME is important as prompt referral and management significantly decreases visual morbidity.
Most of the devices have field of view in range of 40- 50 degrees. The prototype device in the study has a field of view 120 degree about 2.5 times larger than conventional devices. Wider field of view has translated to better detection of peripheral DR which can be missed with conventional devices. The detection of peripheral DR was 30% more than other NM devices.
The image quality is also a significant factor impacting the outcome of screening. In case of unsatisfactory image quality a repeat image was taken. The images were graded for quality at the base hospital. Non gradable images were higher n=30 as compared to other NM devices n=20. Non-gradable images were excluded. Media opacity significantly decreases image quality in NM devices as conventional fundus cameras. Those with non-gradable images were referred to base hospital to be evaluated by ophthalmologist.
The disadvantage with prototype device is a short learning curve in handling and operating the device. The wider field of view has an advantage but it also causes image minimisation which can lead to smaller lesion being missed.
In conclusion the prototype device is a cost effective, portable and comparable to other NM devices in detecting DR with an acceptable sensitivity and specificity and can serve as a useful tool in DR screening camps.
References
- YannuzziLA The Retinal Atlas. New York: Elsevier,2010.
- Cho, N. H., Shaw, J. E., Karuranga, S., Huang, Y., da Rocha Fernandes, J. D., Ohlrogge, A. W., & Malanda, B. (2018). IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Research and Clinical Practice, 138, 271–281.
- H. O., Action plan for the prevention of avoidable blindness and visual impairment, 2009-2013. World Health Organization, 2010.
- W. Yau, S. L. Rogers, R. Kawasaki, E. L. Lamoureux, J. W. Kowalski, T. Bek, S.-J. Chen, J. M. Dekker, A. Fletcher, J. Grauslund et al., Global prevalence and major risk factors of diabetic retinopathy,” Diabetes care, p. DC 111909, 2012.
- Early detection and management of DR require an effective screening program.
- Rachapelle S, Legood R, Alavi Y, et al. The cost-utility of telemedicine to screen for diabetic retinopathy in India. Ophthalmology 2013;120:566–73.
- Cuadros J, Bresnick G. EyePACS: an adaptable telemedicine system for diabetic retinopathy screening. J Diabetes SciTechnol 2009;3:509–16.
- Ogunyemi O, Terrien E, Eccles A, et al. Teleretinal screening for diabetic retinopathy in six Los Angeles urban safety-net clinics: initial findings. AMIA AnnuSymp Proc 2011;2011:1027–35.
- Hautala N, Aikkila R, Korpelainen J, et al. Marked reductions in visual impairment due to diabetic retinopathy achieved by efficient screening and timely treatment. Acta Ophthalmol2013;92:582–7.
- Mansberger SL, Gleitsmann K, Gardiner S, et al. Comparing the effectiveness of telemedicine and traditional surveillance inproviding diabetic retinopathy screening examinations: a ran-domized controlled trial. Telemed J E Health 2013;19:942–8.
- Mohan V, Prathiba V, Pradeepa R. Tele-diabetology to screen for diabetes and associated complications in rural India: The Chunampet Rural Diabetes Prevention Project J Diabetes SciTechnol 2014;8:256–61.
- Cavallerano JD, Aiello LP, Cavallerano AA, et al. Non-mydriatic digital imaging alternative for annual retinal exam-ination in persons with previously documented no or mild diabetic retinopathy. Am J Ophthalmol 2005;140:667–73.
Leave a Comment