• Users Online: 180
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 


 
 Table of Contents  
ORIGINAL ARTICLE
Year : 2022  |  Volume : 34  |  Issue : 3  |  Page : 216-226

Evaluation of quality of non-mydriatic fundus images obtained with non-contact TrueColor, confocal scanner during third phase of nationwide lockdown


1 Medical Officer, Department of Glaucoma and Research, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
2 Medical Officer, Department of Cataract and Refractive Surgery, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
3 Medical Officer, Department of Vitreo-Retinal Surgery, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
4 Bio-Statistician, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
5 Optometrist, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
6 Head of the Department of Cataract and Refractive Surgery, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India
7 Chief Medical Officer, Mahathma Eye Hospital Private Limited, Tiruchirappalli, Tamil Nadu, India

Date of Submission05-Mar-2021
Date of Decision01-Apr-2021
Date of Acceptance30-Apr-2021
Date of Web Publication22-Dec-2022

Correspondence Address:
Dr. Prasanna Venkatesh Ramesh
Department of Glaucoma and Research, Mahathma Eye Hospital Private Limited, No. 6, Tennur, Seshapuram, Tiruchirappalli - 620 017, Tamil Nadu
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/kjo.kjo_58_21

Rights and Permissions
  Abstract 


Aims: This study aims to evaluate the non-mydriatic fundus image quality obtained with a confocal fundus device. Also, to evaluate the influence of non-mydriatic pupil size on image quality. Settings and Design: A cross-sectional study was conducted in the outpatient department of a high-volume tertiary eye care centre in South India during the third phase of the COVID-19 lockdown. Subjects and Methods: 831 consenting, consecutive patients (1539 eyes) were photographed from May 5 to May 17, 2020, and were graded excellent, fair, adequate, or inadequate; based on white noise, black noise, and image acquisition signals. Pupil diameters were obtained with light-emitting diode flash technology. Statistical Analysis: The collected data were analyzed using SPSS software. Descriptive statistics in the form of frequencies and percentages were calculated. Simple (univariate) linear regression analysis and adjusted regression analysis were used to establish the relationship between pupil size and fundus photo quality. Results: The quality of the fundus photo was excellent in 70.12% (n = 1079), fair in 15.72% (n = 242), adequate in 5.78% (n = 89), and inadequate in 8.38% (n = 129). Of the 1410 obtained images (subtracting the inadequate quality images), 87.66% (n = 1236) were normal and 12.34% (n = 174) were pathological. There was a positive linear relationship between pupil size and quality of retinal photograph; OS (R2 = 0.935) had a slightly better association than OD (R2 = 0.901). Conclusions: The majority of the confocal images were excellent in quality. Our four-point grading system serves as a reliable measure of non-mydriatic photograph quality. Pupil size is a significant predictor of image quality for non-mydriatic photographs in surveillance programs.

Keywords: Confocal fundus scanning, fundus imaging quality, non-mydriatic images, pupil size


How to cite this article:
Ramesh PV, Ramesh SV, Balamurugan A, Ansar SM, Devadas AK, Ramesh MK, Rajasekaran R. Evaluation of quality of non-mydriatic fundus images obtained with non-contact TrueColor, confocal scanner during third phase of nationwide lockdown. Kerala J Ophthalmol 2022;34:216-26

How to cite this URL:
Ramesh PV, Ramesh SV, Balamurugan A, Ansar SM, Devadas AK, Ramesh MK, Rajasekaran R. Evaluation of quality of non-mydriatic fundus images obtained with non-contact TrueColor, confocal scanner during third phase of nationwide lockdown. Kerala J Ophthalmol [serial online] 2022 [cited 2023 Feb 2];34:216-26. Available from: http://www.kjophthal.com/text.asp?2022/34/3/216/364706




  Introduction Top


The ophthalmic services at our centre resumed from May 5, 2020, when the third phase of the COVID-19 lockdown began in India with considerable relaxations, and continuation of curbs only in containment zones.[1] A coronavirus safety pod; coronicle (corona + cubicle) was created for investigating and examining all outpatient department (OPD) patients [Figure 1].[2] The cubicle was predominantly constructed with transparent acrylic material, and openings were made in front of the chin rest of all ophthalmic gadgets, for the patients to place their chin during testing. In addition, there were inward angulated dents, designed and created for the placement of the patient's legs during investigations and slit-lamp examination, for extra co-operation and comfort.
Figure 1: (a) The Eidon confocal capture device inside the coronicle (corona + cubicle) for examining the fundus, (b) The Coronicle (corona + cubicle)

Click here to view


While re-starting our ophthalmic outpatient services, there were few challenges [Figure 2] regarding the 90 diopter or 78 diopter examination for the evaluation of fundus; especially the fear of inadvertent patient's face touch or mask touch by the examining doctor, and proximity of examination which can cater to aerosol mediated infection. Also, due to the usage of face shields by the ophthalmologists, the standard stereoscopic examination with 90 diopter or 78 diopter examination was proving very difficult. The similar challenges were encountered with indirect ophthalmoscopy while performing fundus examination. Hence, fundus imaging was done for all; with non-contact, non-mydriatic, automated, confocal, light-emitting diode (LED) fundus capture device (Eidon, iCare) which gave a good 60° exposure of the fundus in a single field with a TrueColor confocal image.
Figure 2: (a) Ninety diopter fundus examination performed wearing face shield. (b) Inadvertent face-mask touch during 90 diopter examination

Click here to view


A review of literature has shown that there is improved image quality, enhanced contrast, more finely detailed images, suppression of scattered light, and better imaging of patients with poor dilation with the confocal technology over traditional flash-based systems.[3],[4],[5],[6],[7],[8],[9] This area of confocal imaging and its subsequent image quality has been poorly reported in the literature before, especially in Indian settings, as this is a new imaging modality. As the posterior segment examination was done only with the confocal fundus capture device at our centre, during the third phase of COVID-19 nationwide lockdown; and the fundus diagnosis were made only with the images obtained from it, the main aim of this study was to evaluate the quality of non-mydriatic fundus images obtained with this fundus capture device. Our secondary objective was to evaluate the influence of non-mydriatic pupil size on image quality.


  Subjects and Methods Top


Population

This is an observational cross-sectional study conducted between May 5 and May 17, 2020, during our emergency ophthalmic outpatient services which was resumed during the third phase of nationwide lockdown during COVID-19 crises. Written, informed consent was obtained from all participants entering the study. There were no specific inclusion or exclusion criteria. All the consenting, consecutive patients presenting to the ophthalmic OPD from May 5 to May 17, 2020 (i.e.,) during the third phase of the nationwide lockdown were included in the study. The idea behind having no exclusion criteria such as lid edema, severe dry eye or wheelchair patient was that we wanted to evaluate the impact of this confocal imaging device as a whole among all the patients presenting to the hospital during that period. Totally 831 consenting, consecutive patients were enrolled in the study during this interval. Subject selection was done from all the patients entering the hospital, not just from a particular department or unit of the hospital. All the patients included in this study were incidentally binocular.

Out of the 831 patients (1662 eyes), 1539 eyes underwent non-mydriatic, non-stereoscopic automated confocal fundus imaging with the Eidon device after a complete ophthalmic examination, including best corrected visual acuity and slit-lamp examination. Based on the nature and progress of disease pathology, certain diseases (such as glaucoma and proliferative diabetic retinopathy [PDR]) which required intervention during the third phase of the lockdown period, were classified as two separate groups, newly diagnosed and already diagnosed/on treatment. 123 eyes were excluded from 1662 eyes due to certain logistic reasons such as shown in [Table 1]. The study protocol designed followed the tenets of the Declaration of Helsinki.
Table 1: Situations where fundus capture were not performed

Click here to view


Non-mydriatic fundus photograph acquisition and pupil diameter measurement

The photographs were taken between 10 am and 1 pm during the third phase of the nationwide lockdown, in our ophthalmology department set up at the coronicle (corona + cubicle) for OPD patients [Figure 1].[2] The right eye (OD) was imaged first followed by the left eye (OS) after giving a few seconds gap for the pupil to recover. The time gap interval between imaging both eyes (OU) was approximately 15-20 seconds. The time interval set up and movement of the capture machine from OD to OS is an automated process, which is in-built with the machine's technology.

A minimum gap of 3 minutes was then allowed, between two patient examinations at the device, after sanitizing the head and chin rest. The fundus examination was repeated at least two times for one eye. The best photograph with the highest quality for each eye was selected for review, after decision involving two ophthalmologists. The two ophthalmologists were one retina specialist and one glaucoma specialist. They both were Eidon users with experience of more than four years with this technology. Their qualifications were MS in ophthalmology with 2-year fellowship training in their respective sub-specialities. In the case of a tie regarding quality between the two ophthalmologists, the first image of that eye with the highest quality was selected for further analysis. During the process, the image readers were masked from the pupil size to avoid bias.

Non-stereoscopic, non-mydriatic, 60° single field exposure, confocal fundus images of the posterior pole were obtained using the LED capture device (Eidon, iCare). Images were obtained by a single experienced paramedical practitioner for all patients, under the supervision of an on-site member (corresponding author) of the study team. Photographs were taken in a semi-dark environment created especially for the capture of the non-mydriatic image, to reduce the ambient light exposure preventing constriction of the pupil. Pupil diameters of the patients were also measured with the LED flash technology of the confocal device.

Analysis of the photographs

The 1539 fundus image reviews were done on a 19-inch liquid crystal display monitor (Dell 1907FPV, resolution: 1280 × 1024 for BB; and HP W1907, resolution: 1440 × 900 for CL) using a local area network connection.[2] The photographs were reviewed in the order in which they were obtained. All color images were analyzed exactly as they were outputted from the device. No image processing (e.g., tone and contrast-enhancing/adjustment or color normalization) was performed. The general quality of each ocular fundus photograph was graded on a four-point scale: (1) Excellent, (2) fair, (3) adequate, or (4) inadequate based on illumination, white noise, black noise, and image acquisition signals [Figure 3]. This quality assessment score was created after a focused group discussion with five Eidon ophthalmic users, keeping in mind, common factors influencing imaging such as mentioned above, along with details to major anatomical sites of exposure such as macula and optic disc.
Figure 3: (a) Excellent scored image (No white noise or black noise) (b) Fair scored image with black noise (red arrow) (c) Adequate scored image with black noise (red arrow) and white noise (black arrow) (d) Inadequate scored image with total image showing black noise

Click here to view


Operational diagnosis-score definition

Excellent scored image

An image was scored as excellent, if white noise or black noise was less than or equal to 25% of the 60° non-mydriatic fundus image area and/or did not involve the optic disc or macula.

Fair scored image

An image was scored as a fair, if the white noise or black noise was between 25% and 75% of the non-mydriatic image area and/or did not involve the optic disc or macula.

Adequate scored image

An image was scored as adequate, if the white noise or black noise was >75% of the non-mydriatic area and/or involve the optic disc or macula but not preventing excluding emergent findings for diagnostic purposes.

Inadequate scored image

An image was scored as inadequate, if the white noise or black noise was 100% throughout the whole non-mydriatic image area, and/or involve the optic disc or macula preventing the exclusion of emergent findings, thus inadequate for diagnostic purposes.

Statistical analysis

The collected data was entered and coded in MS Excel and analyzed using SPSS (Statistical Package for Social Sciences; version 20, IBM USA). Descriptive statistics in the form of frequencies and percentages were then calculated. A Shapiro–Wilk test was performed to assess the normality of distribution. Mann–Whitney U-test and Kruskal-Wallis test (non-parametric test) were utilized for the distribution that did not follow normality. The quality of the images obtained was tested for association with the pupil size. Simple (univariate) linear regression analysis and adjusted regression analysis were used to establish the relationship between fundus photo quality and pupil size. A P value less than 0.05 was considered statistically significant.


  Results Top


Eight hundred and thirty-one consecutive patients (403 males; 428 females) were photographed from May 5 to May 17, 2020. Out of the 1662 eyes (831 patients), a total of 1539 patients' eyes were reviewed after 123 eyes were excluded due to various logistic reasons in whom fundus capture were not performed [Table 1]. The quality of the fundus photo was then evaluated in the remaining 1539 eyes; the quality was excellent in 70.12% (n = 1079), fair in 15.72% (n = 242), adequate in 5.78% (n = 89), and inadequate in 8.38% (n = 129). The reason for inadequate quality was mature cataract in 75.97% (n = 98) and pupil size less than 1.8 mm in 24.03% (n = 31). Out of the 1410 obtained images (excluding the inadequate quality images of 129 images), 87.66% (n = 1236) were normal, and 12.34% (n = 174) were pathological fundus. Out of the 174 pathological eyes, 28.73% (n = 50) were non-PDR (NPDR), 12.07% (n = 21) were previously diagnosed with glaucoma, 10.92% (n = 19) were age-related macular degeneration (ARMD), 9.20% (n = 16) were previously diagnosed with PDR, 6.32% (n = 11) were newly diagnosed with glaucoma, and 4.60% (n = 8) were newly diagnosed with PDR. The mean pupil size was 3.26 ± 0.84 mm in the excellent category, 2.81 ± 1.94 mm in the fair category, and 2.73 ± 0.70 mm in the adequate category [Table 2] and [Table 3].
Table 2: Sex, age and pupil size distribution of the study population

Click here to view
Table 3: Pupil size distribution for different quality of images in both eyes

Click here to view


Test of normality [Table 4] revealed that the pupil size of adequate quality images in OD male and OD female are the only two groups to follow normality. The pupil size of adequate quality images in OS male and OS female, along with the pupil size of excellent category and fair category images of both sexes separately in OU follows non-normal distribution. When they were evaluated as a whole without gender discrimination, test of normality [Table 5] revealed that the pupil size of adequate quality images in OD follows normality. The pupil size of adequate quality images in OS, along with the pupil size of excellent category and fair category images follows non-normal distribution. Mann–Whitney U-test [Table 6] revealed that there is a significant difference between OD and OS pupil size in the excellent, fair, and overall categories. As the pupil size of adequate quality images category in OD follows normality and the pupil size of adequate quality images category in OS follows non-normality, the above Mann–Whitney U-test cannot be performed between them.
Table 4: Test of normality revealed that the pupil size of adequate quality images in right eye male and left eye female were the only two groups to follow normality

Click here to view
Table 5: Test of normality revealed that the pupil size of adequate quality images in right eye follows normality

Click here to view
Table 6: Mann-Whitney U-test revealed that there was a significant difference between right eye and left eye pupil size in the excellent, fair and overall categories

Click here to view


Descriptive analysis of the pupil size with age and gender is shown in [Table 7]. Evaluating the pupil size with age and gender using Kruskal–Wallis test [Table 8] revealed that there is a significant difference between OD pupil size, OS pupil size, and OU pupil size (together) when compared with age. Whereas, there was no significant difference between OD pupil size, OS pupil size, and OU pupil size (together) when compared with gender.
Table 7: Descriptive analysis of the pupil size with age and gender

Click here to view
Table 8: Evaluating the pupil size with age and gender using Kruskal - Wallis test revealed that there was a significant difference between right eye pupil size, left eye pupil size and both eyes pupil size (together) when compared with age

Click here to view


There was a positive linear relationship between the pupil size and quality of retinal photograph; OS (R2 = 0.935) had a slightly better linear relationship than OD (R2 = 0.901). Model summary of adjusted regression analysis of the pupil size OU with age and gender is shown in [Table 9]. Coefficients of adjusted regression analysis of the pupil size OU with age and gender depicted [Table 10] that, if age decreased then the pupil size increased and it was statistically significant (beta coefficient −0.064, P = 0.04); and females had slightly larger pupil size compared to male, but was not statistically significant (beta coefficient 0.043, P = 0.052). [Table 11] depicts the model fitness information (multinomial logistic regression) of image quality with the pupil size that was assessed using the Chi-square test. The table showed that there was a significant relationship between the image quality and the pupil size. [Table 12] depicts the Pseudo R-Square (multinomial logistic regression) of image quality with the pupil size. The model accounted for 3.4%–6.2% of the variance and represented relatively decent-sized effects. [Table 13] depicts the likelihood ratio test (multinomial logistic regression) of image quality with the pupil size. This proved that the independent variables such as OD pupil size and OS pupil size of the patients were significant. [Table 14] depicts the parameter estimates (multinomial logistic regression) of image quality with the pupil size. From [Table 14], in that instance, SPSS was treating the “excellent category” images as the referent group. A one-unit increase in the variable (OD pupil size) was associated with a 0.019 increase in the relative log odds of being in fair image quality versus overall image quality; which was not statistically significant. A one-unit increase in the variable (OS pupil size) was associated with a 0.258 decrease in the relative log odds of being in fair image quality versus overall image quality; which was statistically significant. A one-unit increase in the variable (OD pupil size) was associated with a 0.787 decrease in the relative log odds of being in adequate image quality versus overall image quality, which was statistically significant. A one-unit increase in the variable (OS pupil size) was associated with a 0.472 increase in the relative log odds of being in fair image quality versus overall image quality, which was statistically significant.
Table 9: Model summary of adjusted regression analysis of the pupil size both eyes with age and gender

Click here to view
Table 10: Coefficients of adjusted regression analysis of the pupil size both eyes with age and gender depicted that if age decreased, then the pupil size increased and it was statistically significant (beta coefficient=-0.064, P=0.04); and females had slightly larger pupil size as compared to males and it was not statistically significant (beta coefficient 0.043, P=0.052)

Click here to view
Table 11: Depicts the model fitness information (multinomial logistic regression) of image quality with the pupil size, which was assessed using the Chi-square test

Click here to view
Table 12: Depicts the pseudo R2 (multinomial logistic regression) of image quality with the pupil size

Click here to view
Table 13: Depicts the likelihood ratio test (multinomial logistic regression) of image quality with the pupil size

Click here to view
Table 14: Depicts the parameter estimates (multinomial logistic regression) of image quality with the pupil size

Click here to view



  Discussion Top


The rationale for the choice of this particular instrumentation in this study for retinal imaging was that; it is a non-contact modality which is fully automated and does not require any skill from the examiner, who is performing the test, thus eliminating the human interface during performing the test; which is ironically the need during the pandemic. Another perceived advantage of this fundus capture device was its quick acquisition time of less than 1 min, for both eyes cumulatively, which in-turn reduces patient waiting time inside the hospital premises, during this COVID-19 crises. This instrument is, otherwise also regularly present at the study institute and all the authors were experienced with the usage and interpretation of the TrueColor confocal images obtained with them.

TrueColor confocal imaging - evolution

Color fundus photography is an important tool in the diagnosis and monitoring of various retinal pathologies. Clear and detailed photographs allow an accurate evaluation of the ocular fundus and provide precise documentation of the retinal findings; that can be archived or shared for telemedicine applications as well, or be used as a valuable educational tool.[3] In digital fundus cameras, a bright flash is used to illuminate the ocular fundus; and the light reflected is then captured on the pixel array of a charge-coupled device, thus subsequently generating a digital image. Conventional fundus cameras illuminate large areas of the retina, typically with a flash lamp, and capture 35°-45° field of view, and provide high-resolution digital images. Currently, color images acquired with traditional fundus cameras play a pivotal role in the documentation, diagnosis, and monitoring of retinal disorders.[4],[5] However, conventional flash devices frequently capture color images that are oversaturated in the red channel and are washed out in the green and blue channels, yielding a retinal picture that often looks flat and reddish.[6]

The Eidon device is a slit confocal system that captures 60°, 14-megapixel retinal images in an automated fashion through a non-mydriatic pupil (as small as 1.8 mm). The light source is a broad spectrum white-light LED with a wavelength of 440–650 nm. Sharp, high contrast, and TrueColor image of the retina are obtained. The color of the Eidon images are more balanced (i.e., not saturated by the red component) and “richer” (i.e. greater color discrimination and broader gamut) compared with those obtained with a traditional flash color fundus camera; thus yielding greater discriminative power and diagnostic accuracy, providing accurate documentation of the appearance of retina.[10]

Rationale for nil exclusion criteria

The idea behind having no exclusion criteria such as media opacities, pupil size less than 1.8 mm, lid edema, severe dry eye or wheelchair patient were that; we wanted to evaluate the impact of this technology as a whole, among all the patients presenting to the hospital, during the third phase of the nationwide lockdown. Although the association of the lockdown, with the use of this instrument, is a pragmatic reality; if such a situation arises in the future (such as second wave of pandemic as happening currently), then users would benefit from the data of patients, for whom fundus examination was not performed on a statistical scale due to various reasons listed in the manuscript [Table 1]. On the hindsight, if exclusion criteria were used and these conditions had been removed, then the statistical data of these patients would fall in our blind spot.

Quality of images

Out of the 1662 eyes (831 patients), a total of 1539 patient's eyes were reviewed after 123 eyes were excluded in whom fundus capture were not performed [Table 1]. The quality of fundus photo [Figure 4] was excellent in 70.12% (n = 1079), fair in 15.72% (n = 242), adequate in 5.78% (n = 89), and inadequate in 8.38% (n = 129). The majority of the images were excellent in quality. The reason is as follows: confocal systems allow the capture of reflected light through a small confocal aperture, which suppresses any scattered or reflected light outside the focal plane that could potentially blur the image. This results in a sharp, high-contrast image of an object layer located within the focal plane.[11] The rate of occurrence of low-quality images has been reported at 3.7%–19.7% in clinical studies.[12] In our study, the percentage of poor (inadequate quality) images was 8.38%. The reason for inadequate quality was mature cataract in 75.97% (n = 98) and pupil size <1.8 mm in 24.03% (n = 31). Due to the phase 1 and phase 2 lockdown, we experienced a surge in mature cataract OPD patients, once we resumed eye care services. And in all mature cataracts, fundus imaging was not possible resulting in inadequate image quality. Similarly, when the pupil size was <1.8 mm, the white light LED was not able to generate fundus images due to a very small confocal aperture.[11]
Figure 4: (a) Percentage of various quality of images of all participants. (b) Percentage of various quality of images in male and female. (c) Percentage of various quality of images in right eye and left eye

Click here to view


Pathological fundus

Out of the 1410 obtained fundus images (after excluding the 129 inadequate quality images), 174 images were pathological. Out of the 174 pathological eyes, we were able to pick up and diagnose many new cases, which warranted immediate intervention. The pathological fundi list diagnosed is shown in [Table 15]. Diseases such as NPDR, glaucomatous optic disc damage, PDR, ARMD, and retinitis pigmentosa were commonly diagnosed. Relatively, rare entities such as retinoschisis were also diagnosed. Pathological retina was identified, dilated and a mosaic view revealing up to the equator were imaged, and treated accordingly.
Table 15: 174 pathological fundi diagnosed with Eidon fundus capture device

Click here to view


Pupil size and image quality

Incidentally, it was found that the pupil size had a major role in influencing the quality of the fundus images [Figure 5]. The mean pupil size was 3.26 ± 0.84 mm in the excellent category, 2.81 ± 1.94 mm in the fair category, and 2.73 ± 0.70 mm in the adequate category [Table 3]. When separately studied, the OD mean pupil size was 3.54 ± 0.82 mm, 3.13 ± 0.63 mm, and 2.79 ± 0.71 mm, respectively and the OS mean pupil size was 2.97 ± 0.76 mm, 2.48 ± 0.61 mm, 2.68 ± 0.70 mm, respectively. There was a positive association between the pupil size and quality of retinal photograph in both eyes; OS (R2 = 0.935) had a slightly better association than OD (R2= 0.901). The mild disparity between eyes (OS > OD) was likely due to a greater flash exposure in OS, as OD was studied first followed by OS.
Figure 5: (a) Mean pupil size for various quality of images of all participants. (b) Mean pupil size for various quality of images of right eye and left eye. (c) Mean pupil size for various quality of images of left eye for male and female. (d) Mean pupil size for various quality of images of right eye for male and female

Click here to view


R-Squared (R2 or the coefficient of determination) is a statistical measure in a regression model that determines the proportion of variance in the dependent variable, that can be explained by the independent variable. The R2 shows how well the data fit the regression model (the goodness of fit). However, it does not provide information about a causal relationship between the independent and dependent variables. In addition, it does not indicate the correctness of the regression model. Therefore, conclusions about the model were drawn in this study by analyzing R2 together with other variables in a statistical model. The regression analysis allows to predict an outcome, which can be a good quality image or a poor quality image as seen in this study based on unit changes in the pupil size.

A review of comparison with similar market and post-market studies

Many market and post-market studies have been performed and inferred that confocal imaging has an excellent standard of fundus image quality. Additionally, Eidon TrueColor confocal technology applied with wide-field imaging is particularly crucial as it helps to improve the detection, analysis, and monitoring of pathologies that could arise in the periphery of the retina. EIDON's wide-field features help to preserve the sharpness and details even in the periphery, facilitating improved and early diagnosis.[10] Although the analysis of fundus photography utilizing non-mydriatic fundus camera has been studied in several studies across the globe, no study has so far reported the image quality analysis with confocal fundus camera.[13],[14],[15],[16],[17]

A positive correlation between pupil size and quality of fundus photo was noted in a study by Sangave et al.[18] Along with it, they noted a correlation disparity between eyes, which was attributed to greater flash exposure in OS, similar to our study. In non-mydriatic photography, it is reported that there is a faster pupil recovery time with lower flash intensities, thus improving image quality in the fellow eye.[15] In previously conducted studies, the percentage of ungradable images was found to be between 4% and 34%.[19] In our study, the inadequate ungradable images were minimal (about 8%) which falls within the category.

Our study has found a negative correlation between age and pupil size stating that, as age increases pupil size decreases. Higgs et al. suggested the same in an indirect way, stating that as age increases the percentage of ungradable fundus photo taken through non-mydriatic fundus camera increases, which corresponds to the decreasing pupil size with age.[20] A comparison between a white LED confocal imaging system and a conventional flash fundus camera using chromaticity analysis revealed that Eidon provides more balanced color images, with a wider richness of color content, compared to a conventional flash fundus camera. The overall higher chromaticity of Eidon may provide benefits in terms of discriminative power and diagnostic accuracy.[10]

Limitations

  1. Confocal image quality may reduce in cases with pupil size less than 1.8 mm
  2. Cost of this instrumentation is very high and may not be feasible for all ophthalmic centres
  3. The two image readers were not masked of the findings from one another. The intra-observer and inter-observer agreement between the image quality between the two readers was not assessed.


Suggestions

  1. Pupil diameters of all the patients were measured with the LED flash technology of the confocal device. To the best of our knowledge, this is the first study to measure pupil diameter with LED technology. Further studies are needed to throw more light on the potential of LED flash technology, for pupil size measurements
  2. To draw a definitive conclusion on the clinical utility of this new confocal device, a prospective comparative study comparing the performance of human graders and artificial intelligence-based algorithms of this device with conventional device fundus cameras are warranted.



  Conclusions Top


This study demonstrated that the new confocal, white-light LED system, Eidon, produces good quality color images, providing accurate documentation of retinal appearance and diagnosis. This new fundus imaging system has the potential to enhance patient care, especially in high-volume centres, as the total examination time is only one minute cumulatively for both eyes together, thus indirectly reducing waiting time inside hospital premises, minimizing hospital-acquired infection during the pandemic. [21, 22, 23]

Our four-point grading system serves as a reliable measuring tool for evaluating the quality of non-mydriatic fundus photographs. Pupil size is a significant predictor of image quality for non-mydriatic fundus photographs in high-volume tertiary eye care surveillance programs.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Lockdown 3.0: India Enters Third Phase Today as COVID-19 Tally Reaches 40,263 After Highest Single-Day Spike in Cases. [Internet]. [cited 2021 April 15] Available from: https://www.india.com/news/india/lockdown-3-0-india-enters-third-phase-today-as-covid-19-tally-reaches-40263-highest-single-day-spike-in-cases-4018712.  Back to cited text no. 1
    
2.
Ramesh PV, Parthasarthi S, Ramesh SV, Rajasekaran R, Ramesh MK. Interconnecting ophthalmic gadgets (infinity stones) at finger tips (personal computer desktop) with local area network for safe and effective practice during COVID-19 crises. Indian J Ophthalmol 2021;69:449-52.  Back to cited text no. 2
[PUBMED]  [Full text]  
3.
Keane PA, Sadda SR. Retinal imaging in the twenty-first century: State of the art and future directions. Ophthalmology 2014;121:2489-500.  Back to cited text no. 3
    
4.
Bernardes R, Serranho P, Lobo C. Digital ocular fundus imaging: A review. Ophthalmologica 2011;226:161-81.  Back to cited text no. 4
    
5.
Jackman WT, Webster JD. On photographing the retina of the living eye. Phila Photogr 1886;23:340-1.  Back to cited text no. 5
    
6.
Graham KW, Chakravarthy U, Hogg RE, Muldrew KA, Young IS, Kee F. Identifying features of early and late age-related macular degeneration: A comparison of multicolor versus traditional color fundus photography. Retina 2018;38:1751-8.  Back to cited text no. 6
    
7.
Ben Moussa N, Georges A, Capuano V, Merle B, Souied EH, Querques G. MultiColor imaging in the evaluation of geographic atrophy due to age-related macular degeneration. Br J Ophthalmol 2015;99:842-7.  Back to cited text no. 7
    
8.
Borrelli E, Nittala MG, Abdelfattah NS, Lei J, Hariri AH, Shi Y, et al. Comparison of short-wavelength blue-light autofluorescence and conventional blue-light autofluorescence in geographic atrophy. Br J Ophthalmol 2019; 103:610–16.  Back to cited text no. 8
    
9.
Borrelli E, Lei J, Balasubramanian S, Uji A, Cozzi M, Sarao V, et al. Green emission fluorophores in eyes with atrophic age-related macular degeneration: A colour fundus autofluorescence pilot study. Br J Ophthalmol 2018;102:827-32.  Back to cited text no. 9
    
10.
Sarao V, Veritti D, Borrelli E, Sadda SV, Poletti E, Lanzetta P. A comparison between a white LED confocal imaging system and a conventional flash fundus camera using chromaticity analysis. BMC Ophthalmol 2019;19:231.  Back to cited text no. 10
    
11.
Sharp PF, Manivannan A. The scanning laser ophthalmoscope. Phys Med Biol 1997;42:951-66.  Back to cited text no. 11
    
12.
Bartling H, Wanger P, Martin L. Automated quality evaluation of digital fundus photographs. Acta Ophthalmol 2009;87:643-7.  Back to cited text no. 12
    
13.
Harper CA, Livingston PM, Wood C, Jin C, Lee SJ, Keeffe JE, et al. Screening for diabetic retinopathy using a non-mydriatic retinal camera in rural Victoria. Aust N Z J Ophthalmol 1998;26:117-21.  Back to cited text no. 13
    
14.
Pugh JA, Jacobson JM, Van Heuven WA, Watters JA, Tuley MR, Lairson DR, et al. Screening for diabetic retinopathy. The wide-angle retinal camera. Diabetes Care 1993;16:889-95.  Back to cited text no. 14
    
15.
Scanlon PH, Foy C, Malhotra R, Aldington SJ. The influence of age, duration of diabetes, cataract, and pupil size on image quality in digital photographic retinal screening. Diabetes Care 2005;28:2448-53.  Back to cited text no. 15
    
16.
Lim JI, LaBree L, Nichols T, Cardenas I. Comparison of digital nonmydriatic fundus imaging with standard 35-millimeter slides for diabetic retinopathy. Ophthalmology 2000;107:866-70.  Back to cited text no. 16
    
17.
Hunter JJ, Cookson CJ, Kisilak ML, Bueno JM, Campbell MC. Characterizing image quality in a scanning laser ophthalmoscope with differing pinholes and induced scattered light. J Opt Soc Am A Opt Image Sci Vis 2007;24:1284-95.  Back to cited text no. 17
    
18.
Sangave A, Desmore V, SIsson C, Fischer W, Ramchandran RS. Pupil size and quality of retinal photos in Tele-I-Care: A diabetic retinopathy surveillance project. Invest Ophthalmol Vis Sci 2014;55:2288.  Back to cited text no. 18
    
19.
Leese GP, Ahmed S, Newton RW, Jung RT, Ellingford A, Baines P, et al. Use of mobile screening unit for diabetic retinopathy in rural and urban areas. BMJ 1993;306:187-9.  Back to cited text no. 19
    
20.
Higgs ER, Harney BA, Kelleher A, Reckless JP. Detection of diabetic retinopathy in the community using a non-mydriatic camera. Diabet Med 1991;8:551-5.  Back to cited text no. 20
    
21.
Ramesh PV, Ramesh SV, Ray P, Devadas AK, Ansar SM, Raj PM, et al. The Do-It-Yourself (DIY) novel, safe and cost-effective ophthalmic cubicle (coronicle) in COVID-19 era. TNOA J Ophthalmic Sci Res 2022;60:42-7.  Back to cited text no. 21
  [Full text]  
22.
Ramesh PV, Ramesh SV, Ramesh MK, Rajasekaran R. Utilization of hospital car parking garage for COVID-19 triage and screening in a high-volume tertiary eye care center. TNOA J Ophthalmic Sci Res 2021;59:114-6.  Back to cited text no. 22
  [Full text]  
23.
Ramesh PV, Parthasarathi S, Ramesh SV, Rajasekaran R, Ramesh MK. Innovative utilization of patient's triage with integrated information system during COVID crisis. Kerala J Ophthalmol 2021;33:87-90.  Back to cited text no. 23
  [Full text]  


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
 
 
    Tables

  [Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6], [Table 7], [Table 8], [Table 9], [Table 10], [Table 11], [Table 12], [Table 13], [Table 14], [Table 15]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
Abstract
Introduction
Subjects and Methods
Results
Discussion
Conclusions
References
Article Figures
Article Tables

 Article Access Statistics
    Viewed176    
    Printed8    
    Emailed0    
    PDF Downloaded22    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]