The Use of Innovative Digital Mobile Health Technology for Community Eye Care in Poor Resource Settings

Abstract

The practice of ophthalmology depends largely on ophthalmoscopy and photographic documentation of examination findings from patients. These images are used for patient education, monitoring, storage, expert reviews and treatment. In Africa, due to poor management and poor resources in some health systems, there is a dearth of infrastructure and equipment like fundus cameras. In such resource-poor settings, eye care personnel manage patients with available resources, or improvise with low-cost measures. Smartphone fundoscopy is one of such improvisations. This study describes a novel adapter for smartphone fundoscopy, used in conjunction with a tripod. It was built on existing research on smartphone fundoscopy. Fundus images were captured with a smartphone and a +20D lens, both coupled by an improvised adapter; the Chizaram-Condensing Lens Adapter for Smartphones (C-CLAS) on a tripod. This system works as an indirect ophthalmoscope. Videos of the fundus were recorded, and high-quality still-images were taken from segments of the videos. A total of 54 patients were used in the study, aged between 8 and 74 years. The C-CLAS captured high-quality fundus images from 39 patients (72%), in a variety of normal and pathological conditions. The use of a tripod rendered the procedure hands-free, introducing another dimension to smartphone fundoscopy.

1. Introduction

Fundus image capture is an integral part of ophthalmic practice where they serve the purposes of learning, documentation, patient management and education [1]. The gold standard of retinal image acquisition is the fundus camera, which is an expensive, bulky, stationary table-top equipment that is mostly used in the eye clinic setting [2]. In some parts of Africa, including Nigeria, due to poor health systems, some health institutions lack the capacity and resources necessary for basic health service delivery [3]. This has spurred discussions in several quarters on how to harness innovative technologies to bridge the digital divide and make services efficient, accessible, and affordable [4, 5].

Africa is currently the second most populous continent in the world, with an estimated population of 1 billion; and is predicted to rise to 4 billion by the end of the century [6]. This population growth comes with an exponential rise in the penetration of innovative technologies. Internet penetration, including the use of mHealth technologies for health is on the increase among African economies [6, 7]. mHealth is the application of mobile phone technologies for facilitating health service delivery [7]. The evolution of smartphones with novel designs, and functionalities is revolutionizing the field of digital eye health, especially among poor-resource settings where expensive equipment like fundus cameras are not affordable. In such settings, the simplest and cheapest approach is using a smartphone and an indirect condensing lens for patients with dilated pupils [8, 9, 10, 11, 12]. This chapter aims to contribute to the ongoing discussions on smartphone fundoscopy, with a particular focus on the unique features of the hardware (adapter) used, and its applications among impoverished contexts. It describes the Chizaram-Condensing Lens Adapter for Smartphones (C-CLAS), used in conjunction with a tripod-stand for retinal imaging; and also highlights key findings, challenges, ongoing progress and future directions in-view of adopting, implementing and scaling the adapter.

2. The C-CLAS study

2.1 Methodology

This study was conducted in the eye clinic of the Federal Teaching Hospital Owerri (FTHO), located in the capital of Imo State, Nigeria. FTHO is a tertiary-level healthcare center for training interns, resident doctors, and for research. Thus, this study was part of the requirements for postgraduate medical training in Nigeria and was focused on the application of mHealth technologies in the advancement of retinal imaging in this facility. All patients used in this study signed a written consent for their eyes to be imaged, and for their pictures to be used. Ethical clearance was obtained from the Ethics Committee of FTHO, Nigeria. This study targeted to sample 10% of the total fundoscopy procedures done each day at the FTH, Owerri across different age ranges, gender and pathologies. A total of 54 randomly selected patients aged 8–74 years, without significant media opacities who attended the eye clinic were used for this study. The project team assumed this to be a good representation of the cases in the eye clinic, and suitable for this study. We chose this facility because of large population of patients in need of retinal examination.

The C-CLAS adapter was designed and fabricated using locally-sourced polyvinylchloride PVC materials used in household plumbing. This material is lightweight, malleable, and was cut and molded to the design of the adapter using heat. An Infinix X650B- Hot 8 smartphone was used in this study. However, the adapter can also be used with other smartphone brands, making it versatile and customizable. The tripod-stand was purchased separately from an online retail store, and subsequently modified by adding a clamp for the adapter. The adapter was designed in two parts: A smartphone holder that clamps the smartphone in place, with a window for the camera and flashlight of the smartphone, and a tubular arm of fixed length (15 centimeters) with a lens mount at its end. The +20D Volk condensing lens has about 40°′ field of view, and an angular magnification of 3. The tubular arm length of 15 centimeters was chosen for reasons of weight balancing, and to enable the auto-focus of the smartphone camera easily acquire clear images. The fixed nature of the tubular arm meant the working distance between the lens and smartphone cannot be altered; and this worked well with different patients, irrespective of different ocular axial lengths and refractive errors due to the autofocus capability of the smartphone’s camera.

The tubular arm was attached to the smartphone holder such that the window was aligned with the +20D lens on the lens mount; the unit functioning as a coaxial system, similar to an indirect ophthalmoscope. The picture of the C-CLAS is shown in Figure 1. The condensing lens mount was designed with a slight undersized fit to make it clamp the +20D lens firmly. An adhesive strap was added as an extra measure to secure the lens and prevent it from accidentally falling-off the mount. The device is hand-held (with one hand); but can also be coupled to a tripod. The use of a tripod reinforced the stability, and uniquely enabled the adapter to be used hands-free. In addition, the stability conferred by the tripod-stand also enabled the auto-focus feature of the smartphone to easily acquire a sharp focus of the fundus, thereby improving image quality. The tripod can be used with the legs shortened or extended (Figure 2A and B). The adapter can be used with the patient in standing, sitting or supine positions. A Ray tracing of the optics of the C-CLAS is shown in Figure 3.

Figure 1.

The C-CLAS adapter.

Figure 2.

Tripod shortened (A) and extended (B).

Figure 3.

Optics of the C-CLAS; based on the optics of indirect ophthalmoscopy.

It is important to note that good pupillary dilation prior to filming, and patient cooperation were necessary for easier imaging of the fundus. A minimum pupil aperture of 4 mm was required for imaging. The examination room was dimly lit, and the C-CLAS was placed about 5 centimeters in front of the eye, with the native video recorder and flashlight of the Infinix X650B turned on. The working distance between the lens and the eye was adjusted until an image of the area of interest in the retina filled the field of view (Figure 4). The auto-focus feature of the smartphone’s 13-megapixel triple rear camera and quad flashlight adjusts for clarity automatically. A pre-recording zoom setting of 1.5 to 2× was found to be ideal for image capture, and video recording of the desired area of the fundus was commenced. Before filming, the smartphone’s screen was cast on a large screen television or monitor, for a larger image and for teaching purposes. Live screen casting while filming also promoted social distancing among the team members, in-line with COVID-19 infection prevention protocols. A closer view of the C-CLAS is shown in Figure 5.

Figure 4.

Hands-free fundus recording with the C-CLAS. A: view from patient’s side; B: view from doctor’s side.

Figure 5.

A closer view of the C-CLAS system.

Some patients were intolerant to the intensity of the smartphone’s flashlight, which was not dimmable. To address this, a double-layer of yellow adhesive paper-tape was placed over the smartphone’s flashlight. This significantly reduced the intensity of the flashlight to a tolerable and comfortable level, while still retaining good image quality. However, the adapter can still be easily modified to incorporate either polarizing filters, or neutral density filters to reduce the light intensity. Once the areas of interest were filmed, the video gallery of the smartphone was accessed. The video is played and paused at the desired image, and screenshots were taken. No additional application was needed. Editing was not necessary as the pictures were good enough for interpretation. Lastly, the high-quality images from the patients were ready to be archived in the medical record or used for telemedicine purposes as necessary.

2.2 Results

The C-CLAS was initially validated with randomly selected cohorts and supervised by the Head of Department of the Ophthalmology, FTH Owerri, Nigeria. After validation, the prototype was tested to see how it performed before the commencement of this study. Of the 54 patients used in this study, 39 were co-operative, while 15 were uncooperative (mainly children).

It was observed that the C-CLAS was able to record good quality videos from which quality images were gotten in a variety of normal fundus (Figure 6A–C), and pathologic fundus (Figure 7A–K) among 39 cooperative patients. Of the 39 cooperative patients 24 patients (44%) had different pathologies, while 15 patients (28%) had normal findings (Table 1). It was also observed that the C-CLAS was able to uniquely record hands-free, when coupled to the tripod-stand among cooperative patients.

Figure 6.

(A-C) Normal findings.

Figure 7.

(A–K) Pathologic fundus conditions; A—hypertensive retinopathy with hemorrhages (black arrow) and exudates (white arrow); B—macular involvement in same eye of the patient in A; C—florid drusen in a patient with dry age-related macular disease (ARMD); D—glaucomatous optic atrophy with vitreous floater; E—central retinal vein occlusion in the right eye of a woman with eclampsia; F—macular edema in left eye of a patient with diabetes mellitus; G—central retinal vein occlusion in right eye of a hypertensive patient; H—macular involvement in the same eye of the patient in G; I—Grade 1 hypertensive retinopathy with generalized arteriolar constriction (black arrows); J—large soft confluent drusen in right eye of a patient with dry ARMD; K—geographic atrophy in the left eye of a patient with dry ARMD (same patient in J).

Variable	Frequency (N = 54)	Percentage (%)
Co-operation
Co-operative	39	72
Unco-operative	15	28
Fundus findings
Normal	15	28
Abnormal	24	44

Table 1.

Distribution of participants according to co-operation and fundus findings.

Another key finding of this study was that the researchers were able to self-acquire images of their own retina without needing assistance from anyone, by simple positioning in front of the adapter, and putting a mirror on the opposite site. The mirror aided and guided the process.

2.3 Discussion

The use of the video camera and flashlight of a smartphone, manually and meticulously-aligned with a condensing lens (usually a +20D lens), has made it possible to capture fundus images at no or low-cost. Recently, this technique of using smartphones in conjunction with a +20D indirect ophthalmoscopy condensing lenses has gained traction among eye care providers— locally and internationally. This approach has become a standard examination technique reported by various ophthalmologists and has evolved over time [6, 7, 8, 13, 14, 15, 16]. The first reported use of smartphones for fundus image capture was in 2010 by Lord et al. [15]. They demonstrated smartphone fundoscopy use for clinical and educational purposes, including capturing images of the eye, such as external photographs, fundus photographs, and slit-lamp photographs [15]. Since then, newer techniques and modifications have evolved, and resulted in the acquisition of high-quality images but with more expensive ophthalmic imaging devices such as the fundus camera.

Myung et al. [1], reported a three-dimensional printed lens holder attachment for an iPhone smartphone, as an improvement of the older technique by Lord et al. [15]. They reported that high-quality fundus images from normal and pathological findings were captured using their technology. Their adapter made the technique easier to master and reduced image acquisition time (Figure 5). Other 3D-printed adapters similar to that of Myung et al. [1], include: The ODocs (Figure 6) and Fundus explorer (Figure 7). These devices coupled a condensing lens to the smartphone for fundus imaging; and also employed the use of a third-party apps to acquire better images including: ProCamera [Cocologics GmbH] for iOS only; and FiLMiC Pro [FiLMiCInc] for Android and iOS). Third-party apps regulate the flashlight during the procedure for adequate illumination of the fundus. This also helps to improve patients’ comfort [17]. The FiLMiC Pro App was found to provide control of the light intensity and independent control of focus and exposure. The light levels of the iPhone 4 were tested by Kim et al. [16] and were found to be well within the safety standards for human eyes and 10 times less than the levels produced by the commercially available Keeler Vantage Plus LED indirect ophthalmoscope [17]. Other third-party apps, such as Movie-To-Image or Video-2-Photo enable image extraction from the recorded videos (Figures 8–10) [16, 17].

Figure 8.

3D-printed adapter by Myung et al. [1].

Figure 9.

The ODocs adapter.

Figure 10.

The fundus explorer adapter.

Conceptually, different techniques have been reported for smartphone fundoscopy. Haddock et al. [17] noted the following functionalities and capabilities for smartphone fundoscopy. First, as a fundus camera without hardware attachments (smartphone indirect ophthalmoscopy). This approach was the earliest technique used and relies on the smartphone’s native video recording app, flashlight app, and a condensing lens (+20D or +28D) to obtain fundus photographs. Fundus images are captured in the video mode, and still-images subsequently acquired from the videos [17]. Secondly, as an image transfer device, i.e., the image from a conventional fundus camera is transferred to the smartphone and sent to another site. This technique remains useful for imaging of the anterior segment of the eye. However, for fundus imaging, newer techniques have made imageacquisition simpler and of higher quality [18]. Thirdly, as a digital camera attached to an existing fundus-examination device, such as an ophthalmoscope or a slit lamp.

Following this, in January 2013, the United States Food and Drug Administration approved the iExaminer, which is a smartphone adapter system that attaches and aligns an iPhoneto Welch Allyn’s PanOptic Ophthalmoscope (Figure 11). It uses the iPhone’s camera to capture images of the retina and optic nerve and can capture fundus images without dilation of the pupil with 25° field of view [19]. Although good images of the optic nerve can be acquired with this device, yet, fundus imaging was limited by the narrow field of view. Various slit-lamp adapters have also been invented for ocular imaging with smartphones. Most of these adapters simplify attachment of the phone to available slit lamps and align the smartphone camera with the ocular of the slit lamp. Most of these adapters are designed for anterior-segment imaging including EyePhotoDoc (Fullerton, CA), Zarf’s (Spokane, WA) iPhone and slit-lamp adapters, and the Keeler’s (Broomall, PA) Portable Slit Lamp iPhone 4 image adapter (Figure 12A–C) [20, 21].

Figure 11.

Panoptic ophthalmoscope.

Figure 12.

(A) Eyephotodoc; (B) Zarf’s adapter; (C) portable keeler adapter.

Fourthly, as a fundus camera with hardware attachments that are smartphone specific (often referred to as hardware apps). Giardini et al. [22] documented an adapter for the Samsung Galaxy S III smartphone, which is a low-cost alternative to a direct ophthalmoscope and can capture high-quality fundus images, including the optic nerve [22, 23]. Russo et al. [23] also reported on the use of the D-Eye adapter (Padova, Italy), which magnetically attaches to the smartphone and can captures fundus images of an approximately 20° of field of view (Figure 13).

Figure 13.

D-eye smartphone adapter.

Recent publications have shown that more complex fundus imaging can be accomplished with smartphones. Suto et al. reported on a technique for performing fluorescein fundus angiography using a smartphone [9]. Smartphone fundus angiography could be beneficial for bedridden patients and infants in poor resource settings. Suto et al. were able to capture adequate images for diagnosing certain diseases, although they cautioned that subtle findings might be missed due to the quality of the images with the current technology [9]. In addition, Maamari et al. (described a prototype wide-field lens attachment for the smartphone called Ocular CellScope, which can slide into the phone and capture wide-field images using the phone’s built-in camera [12]. Their technique could capture approximately 55° of the fundus in a single image. Additional software could be used to merge multiple images to create a montage.

Technically, the C-CLAS adapter is similar to the adapters invented in previous studies, such as the ODocs, Fundus explorer, and the 3D adapter by Myung et al. [1]. However, compared to the C-CLAS, these adapters are 3D-printed, relatively more expensive and require both hands for stabilization. Functionally, they share the same optical principles with the C-CLAS. Compared to the Welch Allyn Panoptic ophthalmoscope which has a 25° field of view, the C-CLAS presents a 45° field of view with the +20D lens. Both systems are also portable. While the Panoptic ophthalmoscope is more complex, the C-CLAS offers greater working distance and simplicity. Though much more portable than the C-CLAS, D-Eye smartphone adapter has an approximately 20° field of view in a dilated eye, and 5–8° field of view in an un-dilated eye. This is lower, compared to the C-CLAS’ 45° field of view.

The Keeler, Eyephotodoc, and Zarf’s adapter are slit lamp-based platforms. They require a seated patient, and a stationary slit lamp biomicroscope for fundoscopy, which compromises portability of the entire system. In addition, the adapters are relatively expensive. The Ocular Cellscope comes with a 55° field of view; 10° higher than the C-CLAS, with a mobile application to capture retinal images and record patient information. The C-CLAS offers greater working distance than these adapters, and can be used with the patient seated, standing or supine. The C-CLAS is novel as it rendered the technique of coupling a lens to a smartphone hands-free, with the addition of the tripod; in stark contrast and refinement of the initial method by Lord et al. [15]. In addition, the initial method of placing a +20 D lens over the eye, and manually aligning it with the smartphone’s light and camera as reported by Lord et al. [15] and other researchers had a steep learning curve [13, 24]. The 3D-printed lens adapters that directly and reversibly coupled an iPhone to a +20D lens aimed to address this challenge. It was a step forward, but had problems of weight balancing at increased working distances, which made the authors resort to a fixed working distance. Some of the adapters (ODocs, Fundus explorer) require the system to be stabilized with one hand, leaving the other hand to operate the smartphone. However, the design of the C-CLAS addresses this challenge, making it very easy to use with one hand.

Other systemic challenges noted with the technique of smartphone fundoscopy include the high intensity of the smartphone’s light which was reported to be discomforting for some patients. Haddock et al. [24] reported an iPhone application that allows control of the native flash as an adjustable light-source. This reduced the light intensity to safe levels for indirect ophthalmoscopy [24, 25]. Myung et al. [1] used neutral density filters or polarizing filters to overcome discomforting high flashlight intensity. However, the C-CLAS used a readily available double-layer of yellow paper-tape, and was able to achieve comparable results in terms of comfort and image quality. Yellow filters are thought to improve patient comfort and are very affordable for resource-poor contexts [26].

The use of third-party apps has improved ease of image acquisition and processing. Haddock et al. [24] described the use of either the App Movie-To-Image or Video-2-Photo that enabled the still-images of interesting areas to be taken from the video clips [24, 25]. These applications were not used with the C-CLAS system as images were acquired directly as screenshots of the paused videos, making it simple and efficient to acquire images. Also, screen casting of the fundoscopy procedure on a large screen by the C-CLAS system made fundoscopy a fun-filled and interesting procedure, enabling patient relatives and eye team members to watch comfortably at spaced-out distance; in compliance to COVID-19 social distancing protocols.

The C-CLAS system addresses challenges related to cost and sophistication documented by other studies. The polyvinylchloride material used in designing the adapter is cheap, lightweight, readily available and does not require 3D printing. This makes it easy for mass production. Despite the incorporation of a tripod-stand, its portability was not compromised, as the tripod-stand can be extended or retracted. This means that it can be carried to rural settings where fundus cameras are lacking, and images captured can be transmitted with the smartphone for telemedicine purposes. In addition, it requires no special training to use, making it easily usable by other members of the eye team. The simplicity in design of the C-CLAS, and the ability for self-image acquisition with the aid of a mirror means that patients can actually use it for self-screening and monitoring of select fundus pathologies with minimal training; thus making it a veritable tool of prevention in eye care.

While quality of smartphone-captured media cannot compare to higher-end fundus cameras [24, 27]. They are of sufficient quality to be interpreted, and recent studies have validated their potential for screening and telemedicine. Studies showed that smartphone fundoscopy images could be successfully used to screen for glaucoma and diabetic retinopathy. The results were comparable to standard images or examinations, and they could largely expand its utility in the community setting and in isolated rural areas with poor resources [22, 23].

2.4 Challenges

Despite interesting capabilities of the C-CLAS innovation, yet this study noted some limitations. The authors noted the presence of a white translucent ring artifact at the center of the images captured with the +20D Volks lens. However, when another Volks lens was used, the same artifact appeared on the images, though reduced. These artifacts came from the condensing lenses, but did not hamper the ability to interpret the images, as the original video showed different angles and views. We recommend future studies to explore other lens brands that can be used to address this challenge. In addition, high-fidelity artificial intelligence, can be incorporated to translate captured pathological conditions into the local language, for better understanding. The +20D Volk lenses used in this study had about 40°′ field of view, implying that peripheral view of the retina is limited. Future studies should explore the use of wide-field lenses for more peripheral views. The flashlight of the Infinix X650B smartphone is not dimmable, and luxmeter measurements of the flashlight intensity was not taken with a standard luxmeter. However, to ensure patient’s safety, and eliminate risk of thermal and photochemical hazard, in-keeping with the limit set by the International Organization for Standardization (ISO 15004-2.2) [28], a double-layer of adhesive yellow paper-tape was placed over the flashlight of the smartphone. This significantly reduced the flashlight intensity, and improved patient’s comfort without altering image quality. We recommended that future models of smartphones have adjustable control of the flashlight. A total of 54 patients were used in this study, and based on the projections from the findings, the utility of this hardware will remain unchanged, even when the patient number is called up significantly in the hundreds or thousands. Therefore, future studies can explore this factor. Lastly, the C-CLAS is not yet commercially available, however, efforts are in top gear towards making it available and affordable.

2.5 Conclusion

The C-CLAS system introduced another dimension to smartphone fundoscopy with the coupling of a tripod-stand to a smartphone adapter, and wireless screen-casting of the fundoscopy procedure, while retaining simplicity. This innovation can leverage on the expanding mobile phone networks to advance digital health in ophthalmology among resource-poor settings. African governments should identify turn-around investment strategies for adopting and implementing simple mHealth innovations to improve population health, strengthen service delivery and advance ophthalmic health outcomes. Moving forward, the authors are hopeful that as the smartphone camera technology continues to evolve, better images with better resolution will be obtained.

Acknowledgments

We acknowledge the support from the staff and management of the Federal University Teaching Hospital, Owerri, Nigeria for their support during the Postgraduate training period. This research was completed during the Postgraduate training year as part of the requirements for completing the Medical Postgraduate Training in Nigeria. We also acknowledge the contributions and support from Cynthia Alozie during the Training year and in getting this manuscript ready. We are solely responsible for the views expressed in this manuscript, as it does not reflect the views and policies of authors’ affiliated institutions.

Funding

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Disclosure

The authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf and have nothing to declare.