Development and international validation of custom-engineered and code-free deep-learning models for detection of plus disease in retinopathy of prematurity: a retrospective study.

BACKGROUND: Retinopathy of prematurity (ROP), a leading cause of childhood blindness, is diagnosed through interval screening by paediatric ophthalmologists. However, improved survival of premature neonates coupled with a scarcity of available experts has raised concerns about the sustainability of this approach. We aimed to develop bespoke and code-free deep learning-based classifiers for plus disease, a hallmark of ROP, in an ethnically diverse population in London, UK, and externally validate them in ethnically, geographically, and socioeconomically diverse populations in four countries and three continents. Code-free deep learning is not reliant on the availability of expertly trained data scientists, thus being of particular potential benefit for low resource health-care settings. METHODS: This retrospective cohort study used retinal images from 1370 neonates admitted to a neonatal unit at Homerton University Hospital NHS Foundation Trust, London, UK, between 2008 and 2018. Images were acquired using a Retcam Version 2 device (Natus Medical, Pleasanton, CA, USA) on all babies who were either born at less than 32 weeks gestational age or had a birthweight of less than 1501 g. Each images was graded by two junior ophthalmologists with disagreements adjudicated by a senior paediatric ophthalmologist. Bespoke and code-free deep learning models (CFDL) were developed for the discrimination of healthy, pre-plus disease, and plus disease. Performance was assessed internally on 200 images with the majority vote of three senior paediatric ophthalmologists as the reference standard. External validation was on 338 retinal images from four separate datasets from the USA, Brazil, and Egypt with images derived from Retcam and the 3nethra neo device (Forus Health, Bengaluru, India). FINDINGS: Of the 7414 retinal images in the original dataset, 6141 images were used in the final development dataset. For the discrimination of healthy versus pre-plus or plus disease, the bespoke model had an area under the curve (AUC) of 0·986 (95% CI 0·973-0·996) and the CFDL model had an AUC of 0·989 (0·979-0·997) on the internal test set. Both models generalised well to external validation test sets acquired using the Retcam for discriminating healthy from pre-plus or plus disease (bespoke range was 0·975-1·000 and CFDL range was 0·969-0·995). The CFDL model was inferior to the bespoke model on discriminating pre-plus disease from healthy or plus disease in the USA dataset (CFDL 0·808 [95% CI 0·671-0·909, bespoke 0·942 [0·892-0·982]], p=0·0070). Performance also reduced when tested on the 3nethra neo imaging device (CFDL 0·865 [0·742-0·965] and bespoke 0·891 [0·783-0·977]). INTERPRETATION: Both bespoke and CFDL models conferred similar performance to senior paediatric ophthalmologists for discriminating healthy retinal images from ones with features of pre-plus or plus disease; however, CFDL models might generalise less well when considering minority classes. Care should be taken when testing on data acquired using alternative imaging devices from that used for the development dataset. Our study justifies further validation of plus disease classifiers in ROP screening and supports a potential role for code-free approaches to help prevent blindness in vulnerable neonates. FUNDING: National Institute for Health Research Biomedical Research Centre based at Moorfields Eye Hospital NHS Foundation Trust and the University College London Institute of Ophthalmology. TRANSLATIONS: For the Portuguese and Arabic translations of the abstract see Supplementary Materials section.

The effect of ethnicity on anatomic success following macular hole surgery: a multicentre cohort study.

PURPOSE: The purpose is to assess the effect of ethnicity on surgical macular hole closure. METHODS: A retrospective cohort study was undertaken in five UK National Health Service Hospitals. We included all patients with known ethnicity undergoing vitrectomy, internal limiting membrane peel, and gas/oil tamponade for all stages of primary full-thickness macular hole (FTMH). The primary outcome was anatomic success, defined as FTMH closure with one operation. The secondary outcome was mean change in best-corrected visual acuity (BCVA) comparing baseline with final review. RESULTS: Of 334 operations, the ethnicity profile comprised 78.7% White patients, 11.7% Black patients, 8.1% Asian patients, and 1.5% in mixed/other ethnicities. Mean age was 69.7 years with 68.5% females. Overall, 280 (83.8%) had anatomic success. Anatomic failure occurred in 38.5% of Black patients versus 12.6% of White patients (relative risk: 1.788; 95% CI: 1.012 to 3.159; P = 0.045). Overall, baseline logarithm of the minimum angle of resolution BCVA improved by 0.34, from 0.95 (95% CI: 0.894 to 1.008) to 0.62 (95% CI: 0.556 to 0.676). Mean BCVA improved by 0.35 in White patients, 0.37 in Black patients, 0.23 in Asian patients, and 0.38 in mixed/other ethnicity (P = 0.689). Greater FTMH minimum linear diameter was associated with an increased risk of anatomic failure (relative risk: 1.004; 95% CI: 1.002 to 1.005; P < 0.0001), whereas better pre-operative BCVA (F [1,19] = 162.90; P < 0.0001) and anatomic success (F [1,19] = 97.69; P < 0.0001) were associated with greater BCVA improvement. Socio-economic status did not significantly influence anatomic success or BCVA change. CONCLUSIONS: Black ethnicity is associated with an approximately twofold greater risk of failed FTMH surgery. The reasons for this difference warrant further study.

Recurrent presumed herpes simplex keratitis and episcleritis in keratosis follicularis (Darier's disease).

Keratosis follicularis (Darier's disease) is an autosomal dominant dermatological disorder characterised by abnormal epidermal differentiation and loss of normal cell-to-cell adhesion. Cardinal features include diffuse hyperkeratotic warty papules with scaly plaques in seborrhoeic regions with associated mucous membrane changes. Darier's disease is rare (prevalence 2.7 in 100,000), with few ocular sequelae reported: commonly dry eye with or without Sjögren's syndrome. This is the first report, to the best of our knowledge, to describe a case of recurrent herpes simplex virus (HSV) keratitis and episcleritis in a 47-year-old man suffering from Darier's disease. The patient's condition predisposed him towards developing ocular complications due to several factors: impaired desmosome function leading to poor cell-to-cell adhesion in the corneal epithelium, dry eye and HSV invasion of inflamed periocular skin presumably combining to allow viral colonisation of a poorly protected cornea.