Evaluation leukocoria app

Evaluation of a smarthphone photoscreening app to detect leukokoria in children aged 1-5Aldo Vagge1,2, Nutsuchar Wangtiraumnuay3, Michele Iester1,2, Riccardo Scotto1, Carlo Enrico Traverso 1,21 Eye Clinic of Genoa, Policlinico San Martino, Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DiNOGMI), University of Genova, 16132 Genova, Italy2 IRCCS Ospedale Policlinico San Martino, 16132 Genova, Italy3 Department of Ophthalmology, Queen Sirikit National Institute of Child Health, Bangkok, ThailandConflict of Interest: No conflicting relationship exists for any authorsFinancial support: NoneRunning head: evaluation of leukocoriaCorresponding Authors:Aldo Vagge, MD, PhDUniversity Eye Clinic of GenoaDiNOGMI – University of GenoaIRCCS Ospedale Policlinico San MartinoViale Benedetto XV, 516132 – Genova (GE) – ItalyPhone: +39 010.353.8491Email: [email protected] IntroductionLeukocoria is the medical term for the white-eye reflex common in several childhood eye diseases affecting any of the structures that form the visual axis, including the cornea, the lens, the vitreous, and the retina (1). Common causes of leukocoria in a child are retinoblastoma (Rb), persistent hyperplastic primary vitreous retinopathy of prematurity, Coats disease, toxocariasis, congenital cataract, phakomatoses, Norrie disease, and retinal dysplasia and detachment (2). Hence, leukocoria requires urgent evaluation by an ophthalmologist, since the majority of these conditions are significant for visual impairment and life-threatening. Leukocoria is the presenting sign in at least 50% of patients with Rb (3). Leukocoria can be detectable by pediatrics and ophthalmologists during screening visits, eliciting the red reflex, a non-invasive and child-friendly screening tool (4). However, parents often notice a white pupil or relatives on flash photography (5) and this is the first reason of consultation (6). The popularization of smartphone cameras enables parents to look for early diagnosis (6). In addition, eye-related smartphone application, or ‘app’, are nowadays available. In particular, The Computer Assisted Detection of Leukocoria (CRADLE) smartphone app was developed with the purpose to increase early diagnosis of RB by detecting leukocoria (7). The aim of this study is to determine whether a white eye detector smartphone app (CRADLE – ComputeR Assisted Detector of LEukocoria) can be used as a screening tool to detect early diagnosis of leukokoria in clinical practice. MethodsFor this prospective, single visit study, we enrolled patients children aged 1-5 years old presenting for complete pediatric ophthalmological examination to the University Eye Clinic of Genova, Department of Neuroscience, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), IRCCS Ospedale Policlinico San Martino and Queen Sirikit National Institute of Child Health, Bangkok, Thailand. A parents or legal guardian gave informed consent before undergoing the study, and the study itself was conducted in accordance to the tenets of the Declaration of Helsinki. Exclusion criteria included any history of ocular surgery or any conditions that would preclude the ability of the investigators to obtain a reliable photo image. All children who met the enrollment criteria were screening using the smarhphone app (CRADLE – ComputeR Assisted Detector of LEukocoria) on the IPhone 7 by an ophthalmic technician. The app is able to detect the eyes on a face in a subject upright and facing the camera with the face fully visible, providing a quadrate red or green colored template at the correct focal distance around each single eye. The red template is associated with the legend “WHITE” meaning a leukokoric eye whereas the green template is associated with the legend “normal” meaning a normal eye. Following photoscreening, a comprehensive examination according to the AAPOS guidelines was performed (8). Cycloplegic retinoscopy and fundus examination were performed by the examining pediatric ophthalmologist 30 to 40 minutes following the instillation of proparacaine hydrochloride ophthalmic solution USP 0.5% followed by 1-2 drops of a pediatric “combo drop” of tropicamide1%, phenylephrine2.5%, and cyclopentolate 1%. The two pediatric ophthalmologists (A.V. and N.W.) participated in the study were masked to photoscreener results.Statistical analysisContinuous variables are given as means and standard deviations (SD), whereas categorical variables as number and/or percentage of subjects. Comparison of the two methods yielded sensitivity, specificity, positive predictive value and negative predictive value. ResultsA total of 244 eyes of 122 children, 64 (52%) male and 58 female (48%), were enrolled for evaluation. The majority of patients were Caucasian (67%), followed by Asian (16%) and Hispanic (17%). The mean age was 19.7± 26.4 months (range 2-60). Nine eyes of 244 (3.6%) had amblyogenic cataract, 1 patient (0.4%) retinopathy of prematurity V stage and 3 (1.2%) Rb at the eye visit by the pediatric ophthalmologist. All these eye conditions were able to determine leukokoria evaluable by penlight or ophthalmoscope. The smartphone app was able to detect only two leukokoric eyes of the same patient caused by bilateral Rb. None of the nine eyes with amblyogenic cataract was detected by the smartphone app (Figure 1 and 2). Therefore, there were 11 false negative and none false positive. The sensitivity of white-eye detector app was 15.38% (95% CI: 1.92% to 45.45%) whereas the specificity was 100% (95% CI: 98.48% to 100.00%). The results are shown in Table 1. DiscussionThe red reflex test is an effective screening test in the early diagnosis of various illnesses and disorders of eyes. The red reflex test uses transmission of light from an ophthalmoscope through all the normally transparent parts of a subject’s eye, including the tear film, cornea, aqueous humor, crystalline lens, and vitreous humor. This light reflects off the ocular fundus, is transmitted back through the optical media and through the aperture of the ophthalmoscope, and it is imaged in the eye of the examiner. The term “red reflex”, is the reflection of the fundus color (the color combination of vascular area and choroid pigments). Any resentment placed within this path in the eye, partially or completely prevents the reflection, and would appear as a black mark or a shadow. An abnormal red reflex can result from several eye conditions can cause white pupil as a first sign, such as congenital cataract, Rb, retinopathy of prematurity, nematode endophthalmitis, persistent hyperplastic primary vitreous, Coats’ disease, ocular toxocariasis, retinal hamartomas, severe refractive error, organized vitreous hemorrhage, ocular albinism, myelinated fibers, and the presenting sign of abusive head trauma (9). As recommended by American Academy of Pediatrics (AAP) (10), all neonates, infants, and children should have an examination of the red reflex of the eyes performed by a pediatrician or other primary care clinician trained in this examination technique before discharge from the neonatal nursery and during all subsequent routine health supervision visits. Leukokoria is an abnormal pupillary light reflection that usually results from an intraocular abnormality and is seen most often in children. One-half of the cases of childhood leukokoria are caused by Rb (11). Retinoblastoma is the most common intraocular tumor of childhood and can be life threatening. For this reasons, several analysis to detect Rb evaluating leukokoria by amateur photography have been developed. CRADLE app was developed by an American engineer whose son’s life-threatening case of Rb was not diagnosed until he was 3 months old, even though photographs taken of him beginning when he was 12 days old showed the telltale white. By activating, the smartphone’s LED and video camera, CRADLE converts a smartphone into a computer-assisted ophthalmoscope; the app then analyzes each video frame as it is recorded in real time, highlighting images of eyes that show signs of leukokoria with red boxes. CRADLE was been considered as an ophthalmoscope alternative in underdeveloped countries (12). Researcher that developed this app pointed out that the purpose of the app is not to diagnose disease e it does not replace eye visit by a physician. CRADLE app is producing algorithms, which can learn from examples to automatically detect when a photo or video contains a white-eye. In this study, we observed that although the specificity of this app was 100%, the sensitivity was 15.38%. In fact, the app was not able to recognize 11 eyes with a leukokoria evaluable by penlight or ophthalmoscope. Probably the reason is to be found in the lack of similar sample images in the experimental database. Although we recognize that this study has limitation as a small cohort, we feel that this smarthphone photoscreening app cannot be considered as an ophthalmoscope alternative in children aged 1-5 cannot be considered. Volunteer on community health projects in underdeveloped countries should be trained to perform the red reflex test of the eyes. However, the ability of the red reflex test to detect Rb is difficult to evaluate because of the low incidence of this disease (13) Of course, smarthphone photoscreening app to detect leukokoria can be a valuable support for children’s parents and relative unless it does not induce to do not perform the recommended eye check-up by a pediatric ophthalmologist when the result of the screening is “normal”. Smartphone medical apps may be extremely helpful for patients, practitioners, and trainees, especially for life threatening disease. Further studies with a larger simple size are needed to confirm the role of this tool in detect leukokoria. Table 1: Sensitivity, specificity and Negative Likelihood Ratio of the white detector smartphone app (CRADLE – ComputeR Assisted Detector of LEukocoria). Statistic Value 95% CISensitivity 15.38% 1.92% to 45.45%Specificity 100.00% 98.49% to 100.00%Negative Likelihood ratio 0.85 0.67 to 1.07ReferencesDamasco VC, Dire DJ. A child with leukocoria. Pediatr Emerg Care. 2011;27(12):1170–1174Kembhavi SA, Sable N, Vora T, Arora B. Leukokoria: All That’s White Is Not Retinoblastoma. J Clin Oncol. 2011 Jul 1;29(19):e586-7Abramson DH, Frank CM, Susman M, Whalen MP, Dunkel IJ, Boyd NW. Presenting signs of retinoblastoma. J Pediatr 1998;132:505–508McLaughlin C, Levin AV. The red reflex. Pediatr Emerg Care. 2006 Feb;22(2):137-40Christian LW, Bobier WR. Case 2: A three-year-old boy with photographic leukocoria. Paediatr Child Health. 2015 Oct;20(7):345-6.Asensio-Sánchez VM, Díaz-Cabanas L, Martín-Prieto A. Photoleukocoria with smartphone photographs. Int Med Case Rep J. 2018 May 16;11:117-119.https://www.umassmed.edu/news/news-archives/2016/06/umms-student-testing-smartphone-app-to-detect-childhood-cancer-in-guatemala/Donahue SP, Arthur B, Neely DE, Arnold RW, Silbert D, Ruben JB. AAPOS Vision Screening Committee. Guidelines for automated preschool vision screening: a 10-year, evidence-based update. J AAPOS. 2013;17:4–8Balmer A, Munier F. Differential diagnosis of leukocoria and strabismus, first presenting signs of retinoblastoma. Clin Ophthalmol. 2007 Dec; 1(4): 431–439American Academy of Pediatrics; Section on Ophthalmology; American Association for Pediatric Ophthalmology And Strabismus; American Academy of Ophthalmology; American Association of Certified Orthoptists. Red reflex examination in neonates, infants, and children. Pediatrics. 2008 Dec;122(6):1401-4. doi: 10.1542/peds.2008-2624.Smirniotopoulos JG, Bargallo N, Mafee MF. Differential diagnosis of leukokoria: radiologic-pathologic correlation. Radiographics. 1994 Sep;14(5):1059-79. Sandra Gray. UMMS student testing smartphone app to detect childhood cancer in Guatemala. https://www.umassmed.edu/news/news-archives/2016/06/umms-student-testing-smartphone-app-to-detect-childhood-cancer-in-guatemala/Sun M, Ma A, Li F, et al. Sensitivity and Specificity of Red Reflex Test in Newborn Eye Screening. J Pediatr. 2016 Dec;179:192-196.e4.Figures28003541465500187071049085500308610812800018707109207500Figure1: 2 years old baby with bilateral amblyogenic cataract, evaluated as “normal” by the smarthphone photoscreening app2390140380365002366010760095006515107981960066548043180000 Figure 2: 4 years old baby with bilateral amblyogenic cataract under general anesthesia, evaluated as “normal” by the smarthphone photoscreening app