PANDA-ASOCT: Paediatric Anterior Uveitis Detection With Automated Image Analysis Using ASOCT

Uveitis

ASOCT, Uveitis, Ophthalmology

Uveitis, an inflammation inside the eye. Although rare, it is an important auto-inflammatory eye condition that affect children and young people (CYP). If not identified and treated promptly, it can lead to permanent vision loss. Uveitis is estimated to affect around 4-5 in every 100,000 children, and up to 26% of those with childhood-onset disease may develop severe visual impairment. Because uveitis often causes no symptoms, even when it is advanced, it is sometimes described as a "silent blinding disease" Uveitis can occur on its own or together with other immune-related conditions. It is the most common non-joint (extra-articular) complication of Juvenile Idiopathic Arthritis (JIA), the most frequent childhood auto-inflammatory joint disorder, which affects about 5.6 per 100,000 children in the UK. National guidelines from the Royal Colleges recommend regular uveitis screening for all children and young people with JIA. This study aims to assess whether a high-resolution, non-irradiating eye imaging system can accurately detect inflammatory cells inside the eye objectively. The scan is non-invasive, non-contact, painless, and well tolerated by children. It can take 50,000 images per second, allowing the front of the eye to be captured in seconds. These images will be analysed using validated automated algorithms to identify and measure signals from inflammatory cells objectively, without influence of human factor. The qualitative and quantitative information may help predict disease activity and monitor treatment response. We aim to compare the imaging results with the current standard slit-lamp eye examination in 300 children and young people (600 eyes). Both the imaging operators and the senior clinical examiners will be not know what each other's findings. The primary aim is to determine how AS-OCT imaging system is in detecting uveitis in children and if the results is the same as the standard eye test. The measurement data will also support early-phase deep-learning development to create a cost-effective, automated tool for diagnosing, grading, and crucially monitoring uveitis over time, particularly during treatments that involve complex immunosuppressive medications.

Uveitis, an inflammation inside the eye, is a rare auto-inflammatory eye condition in children and young people (CYP). It is a potentially blinding disorder if left untreated. Uveitis is a preventable blindness in CYP, estimated 4-5 per 100,000 children developing it. Blindness was reported in 26% of childhood onset uveitis. Children with uveitis do not always have symptoms, even in the most severe end-stage, and often labelled as 'silent blinding disease'. Uveitis can occur in association with systemic auto-immune disorder and is the main extra-articular manifestation of juvenile idiopathic arthritis (JIA). JIA is the most common auto-inflammatory joints disease in children, affecting 5.6 per 100,000 children in the United Kingdom. Guidelines published by the Royal Colleges recommend uveitis screening for all CYP with JIA. This study aims to evaluate the clinical application of a high-resolution non-irradiating imaging system to detect inflammatory cells inside the eye. The camera is non-invasive, non-contact and well tolerated in children. It has a scan-speed of 50,000 images per second. The entire front of the eye can be scanned in 1.2 seconds. The images will be analysed using an appropriately calibrated automated algorithms to detect signals from inflammatory cells objectively. The qualitative and quantitative measurements can be used to predict disease and monitor treatment response. The imaging data will be compared with current standard eye examination in 300 children (600 eyes), masking both the scanner operators and the senior examiners in clinics. The primary objective is to assess sensitivity of the automated imaging system in detecting and predicting uveitis in children. The quantification data will contribute to early phase of deep learning in developing a cost-effective automated diagnostic tool in detecting, grading and most importantly monitoring of uveitis objectively during treatment(s) which often involve complex immunosuppressive agents.

No related papers found yet.

United Kingdom