TY - JOUR
AU - Garg, Sunir J.
AB - As a profession, we are fortunate to be able to obtain high-quality photographs of essentially all ocular structures and view them in high resolution nearly instantly. Rather than being luxuries, they are essential to the function of our office. The downside remains that each office has several hundred thousand dollars of imaging equipment that is constantly in need of maintenance, upgrading, and repair. We also spend a lot of time and effort recruiting and training our photographers. Getting good-quality images is only part of the equation. Analyzing these images both to diagnose disease as well as to monitor progression is a fundamental skill of our profession, and much of residency, fellowship, and postgraduate training are spent on image analysis. In our current model of health care, both acquiring images as well as interpreting them remains resource intensive in terms of personnel and capital expense. This model has traditionally worked in the United States but for the several hundred million people with visual impairment who live in developing and/or low-income countries this model does not work.1 Africa is disproportionally burdened by a high rate of blindness, with glaucoma as the leading cause of irreversible blindness.2 Developing delivery systems that are able to efficiently screen people in areas lacking reliable transportation or access to health care facilities is critical to prevent treatable causes of blindness. Over the past few years, several groups described using the ubiquitous smartphone as a high-resolution camera to obtain both anterior and posterior segment images.3 Most commonly, the phones are either mounted to the ocular on a slitlamp, or are simply held in front of the eye and used to record an image; when imaging the optic disc or the retina, the examiner usually obtains a short video then uses image capture software to extract a representative still image. In this issue of JAMA Ophthalmology, Bastawrous and colleagues4 used smartphone-based ophthalmoscopy in Kenya. They developed a user-friendly imaging device using an inexpensive smartphone that they modified with a plastic housing that uses a prism to change the path of the camera’s LED light (flash) to make it more in line with the camera’s field of view similar to a direct ophthalmoscope. Optic disc images obtained with a reference mydriatic fundus camera operated by an experienced ophthalmic assistant were compared with mydriatic images acquired using the smartphone. Importantly, in the smartphone arm, images obtained from an experienced ophthalmic clinical officer were compared with images acquired by a layperson with no health care background who was given a short education prior to being sent into the field. Finally, a masked reader several thousand miles away graded the images. Several findings from this article stand out. Similar to earlier trials, the authors found that smartphone-based ophthalmoscopy obtained high-resolution images comparable in quality with a standard reference camera.3 With the rapid progress in camera optics, computational processing power, and image processing software, smartphone camera image quality approaches that from an SLR camera mounted to the back of a traditional fundus camera. Much of the image quality difference is due to variations in camera operability; as useful as autofocus can be, it can sometimes hamper the final image. Table-mounted cameras can be manually focused, and have fixation and focusing aids, while the photographer relies on the appearance of blur on the display as the only focusing tool when using a smartphone. To get images of sufficient quality, eyes must be dilated, as nonmydriatic posterior segment images from a smartphone are not very good. A traditional table-mounted fundus camera provides a more stable platform for image acquisition, whereas most smartphone-based ophthalmoscopy is performed freehand. However, this study suggests that even with these disadvantages, there was a high degree of correlation when comparing cup-disc ratios between modalities. In many remote areas, transporting expensive equipment along poorly developed roads is also not feasible, some remote areas may not have reliable sources of electricity, and equipment maintenance and repairs can be costly and time-consuming. In comparison, a smartphone can be repaired or replaced at relatively low cost given the widespread availability of the devices. Similar to models of eye screening used elsewhere, in the current study, nonophthalmic personnel were rapidly trained to use the smartphone camera, obtain high-quality images of the optic disc, and transmit the images to a central reading center. This model provides a rapidly deployable system for screening patients for glaucoma near their homes, using laypeople with minimal training, in areas with few ophthalmologists. Such screening remains an important step to develop a system capable of facilitating referrals of high-risk patients to eye care professionals. This model frees up valuable physician time for both diagnosing pathology as well as concentrating therapeutic efforts on patients with a high degree of pathology. This is particularly important in Africa, many parts of which have few ophthalmologists, each of whom may need to serve 1 million people or more.1 The study also looked at the ability of nonophthalmic personnel to grade cup-disc ratio. There was less agreement between expert grader and the lay grader. This may be due to inadequate education of the lay trainer. Some investigators have demonstrated that using a crowdsourcing platform, nonophthalmic graders achieve very good reproducibility and accurate assessment of optic neuropathy.5 Combined with an easily deployable mobile screening platform, crowdsourcing may also enable efficient and cost-effective health care to patients who would otherwise receive none.6 Over time, automated determination of cup-disc ratio may augment or supplant human grading of images, which would further allow ophthalmologists to concentrate their efforts on patients most in need.7 A sustainable health care delivery model requires more than just screening one part of the eye. For glaucoma, visual field testing, optical coherence tomography, and intraocular pressure obtained in the field would increase the sensitivity and specificity of the diagnosis made by a smartphone image, and images with wider field of view could permit more comprehensive screening. The infrastructure required to treat patients once identified remains resource intensive. Nonetheless, programs to improve community eye health require appropriate screening of the population to be effective. Mobile camera platforms also can provide methods to record procedures from the surgeon point of view,8 expanding the ability to train ophthalmologists, and ophthalmic personnel even in remote areas can use their smartphone to access the many high-quality educational resources available. With the availability of smartphones, someday it may be possible for patients to take photographs of their own eyes as a screening device for glaucoma, macular degeneration, or diabetic retinopathy. Others are working on optical coherence tomography on a chip that could allow mobile optical coherence tomography to be performed. As demonstrated in this issue of JAMA Ophthalmology, screening programs with mobile devices will enable communities to more effectively and affordably screen their population to identify patients who are at risk of losing vision from treatable conditions. Although this study was conducted in a low-income countries, there are still millions who do not get regular eye examinations in the United States and other western countries who may benefit from this type of screening technique. Ultimately, regardless of locale, mobile platforms may prove the best way to reach more individuals at risk for vision loss. Back to top Article Information Corresponding Author: Sunir Garg, MD, The Retina Service of Wills Eye Hospital, Thomas Jefferson University, MidAtlantic Retina, 840 Walnut St, Philadelphia, PA 19107 (sgarg@midatlanticretina.com). Published Online: November 25, 2015. doi:10.1001/jamaophthalmol.2015.4823. Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. References 1. World Health Organization. Action plan for the prevention of avoidable blindness and visual impairment 2009-2013. http://www.who.int/blindness/ACTION_PLAN_WHA62-1-English.pdf, page 11. Accessed September 9, 2015. 2. Resnikoff S, Pascolini D, Etya’ale D, et al. Global data on visual impairment in the year 2002. Bull World Health Organ. 2004;82(11):844-851.PubMedGoogle Scholar 3. Adam MK, Brady CJ, Flowers AM, et al. Quality and diagnostic utility of mydriatic smartphone photography: the Smartphone Ophthalmoscopy Reliability Trial. Ophthalmic Surg Lasers Imaging Retina. 2015;46(6):631-637.PubMedGoogle ScholarCrossref 4. Bastawrous A, Giardini ME, Bolster NM, et al. Clinical validation of a smartphone-based adapter for optic disc imaging in Kenya [published online November 25, 2015]. JAMA Ophthalmol. doi:10.1001/jamaophthalmol.2015.4625.Google Scholar 5. Mitry D, Peto T, Hayat S, et al. Crowdsourcing as a screening tool to detect clinical features of glaucomatous optic neuropathy from digital photography. PLoS One. 2015;10(2):e0117401.PubMedGoogle ScholarCrossref 6. Brady CJ, Villanti AC, Pearson JL, Kirchner TR, Gupta OP, Shah CP. Rapid grading of fundus photographs for diabetic retinopathy using crowdsourcing. J Med Internet Res. 2014;16(10):e233.PubMedGoogle ScholarCrossref 7. Hatanaka Y, Nagahata Y, Muramatsu C, et al. Improved automated optic cup segmentation based on detection of blood vessel bends in retinal fundus images. Conf Proc IEEE Eng Med Biol Soc. 2014;2014:126-129.PubMedGoogle Scholar 8. Rahimy E, Garg SJ. Google Glass for recording scleral buckling surgery. JAMA Ophthalmol. 2015;133(6):710-711.PubMedGoogle ScholarCrossref
TI - Applicability of Smartphone-Based Screening Programs
JF - JAMA Ophthalmology
DO - 10.1001/jamaophthalmol.2015.4823
DA - 2016-02-01
UR - https://www.deepdyve.com/lp/american-medical-association/applicability-of-smartphone-based-screening-programs-xa69F50nnM
SP - 158
EP - 159
VL - 134
IS - 2
DP - DeepDyve
ER -