The benefits of INSIGHT
INSIGHT’s main aim is to benefit patients and the NHS. We do this by making it easier for trusted organisations to undertake research using large-scale, anonymous patient data.
Here we highlight some of the areas of healthcare where researchers are already making discoveries using eye data, and where INSIGHT could make a difference.
Why do researchers need patient data?
If you are a patient at one of our two NHS Trust partners, University Hospitals Birmingham and Moorfields Eye Hospital, it’s likely that your eye data (including routine scans and tests, details of eye conditions and associated health records) has been collected as standard practice. This is done both to improve NHS services and for research purposes. By combining data from thousands of patients over many years, the NHS has uniquely rich datasets that are an extremely valuable resource for medical research.
In recent years, researchers have made a number of breakthroughs using these types of dataset. For example, in a process known as ‘machine learning’, computers can analyse thousands of eye scans in minute detail, detecting tiny differences that might be missed by the human eye. By feeding in the patient records associated with each scan, the computer can begin to identify links between the image and a particular condition or characteristic. Once machines have been ‘trained’ using these datasets, they can analyse a new scan in the clinic, diagnosing or predicting a condition at the touch of a button. The implications for speeding up treatment are huge, because at the moment there can be a significant wait to get an accurate diagnosis from a specialist doctor.
Why do researchers need INSIGHT?
At the moment, it can be complicated and time-consuming to search for and get access to the kind of high quality datasets required for machine learning. INSIGHT will make it easier for carefully vetted researchers to access the data they need through an efficient, safe and ethical process.
INSIGHT applies the highest standards of security and transparency at every stage, from the anonymisation and storage of patient data through to scrutinising applications and publishing research outcomes. You can find out more about how INSIGHT works in About us.
Areas of research where INSIGHT could help
Diagnosing age-related macular degeneration (AMD)
AMD is the leading cause of blindness in the UK. Can we save sight through quicker diagnosis and treatment?
Researchers should be able to use the millions of images available through INSIGHT to improve the diagnosis and treatment of common retinal disorders like AMD. One example is a project led by INSIGHT researcher Dr Pearse Keane. It uses large datasets from Moorfields Eye Hospital, and has already proved that we can use computers to detect the early signs of AMD and help save people's sight.
Working with artificial intelligence company DeepMind, Moorfields has developed a new tool that can detect as many as 50 retinal diseases from a simple OCT (optical coherence tomography) scan in seconds. Once clinical approval has been granted, this tool will ensure that patients with most serious cases of AMD are seen urgently by a consultant, lessening the risk of sight loss.
By 2030, 5.5 million people in the UK are expected to have diabetes. Can we help prevent the harm caused by diabetic retinopathy (a complication of diabetes), which can lead to sight loss?
Researchers are working hard to reduce the harm caused by diabetic retinopathy through improved targeting of treatments and better diagnosis. They are using eye datasets to investigate the effectiveness of existing medication and the potential for developing new screening techniques using artificial intelligence. INSIGHT's large datasets could play an important role in this research.
Vascular dementia and Alzheimer’s disease
By 2025, almost a million people in the UK will suffer from dementia (most commonly Alzheimer’s disease). Can we use 3D scans of the eye to detect the early signs and treat patients more quickly?
Researchers have known for some time that dementia is associated with retinal change. This led to a project called AlzEye, set up by Moorfields Eye Hospital in 2017 and involving INSIGHT researcher Dr Pearse Keane. Its aim is to investigate whether dementia can be detected in the retina before signs appear elsewhere. It uses machine learning and a huge dataset (over 250,000 people) linking retinal photographs and OCT scans with NHS hospital records on cardiovascular and neurogenerative disease, as well as other related conditions.
Inherited retinal diseases
Inherited retinal diseases (IRDs) are a group of rare,
blinding conditions caused by one or more malfunctioning genes. Can we use the large datasets held by INSIGHT to help us diagnose IRDs at an earlier stage, when treatment
is more likely to be effective?
Moorfields Eye Charity is currently funding a project to investigate the use of artificial intelligence (A.I.) in diagnosing Stargadt disease, the commonest inherited disease of the macula. It applies deep learning to electroretinography (ERG), which measures electrical
activity of the eye.
Glaucoma – where the optic nerve is damaged by excess pressure – is responsible for about 10% of blind registrations in the UK. Can INSIGHT datasets help researchers looking for new and improved ways of diagnosing and managing Glaucoma?
A number of studies have shown that artificial intelligence algorithms can successfully screen for glaucoma and the risk of glaucoma progression through the analysis of OCT scans, fundus photography and visual fields tests. You can see some of the academic papers related to these studies on our Resources page.
How can the NHS monitor the impact on the delivery of eye care services caused by the global coronavirus pandemic? INSIGHT data is already being used to carry out a clinical audit and evaluation relating to patients with neovascular (wet) age-related macular degeneration (nAMD) in England.
Not only can personalised medicine offer patients faster, more precise diagnoses based on their unique situation and more effective treatments with fewer side effects, it could even help prevent conditions from developing at all. To make this possible, researchers need to analyse genetic and clinical information on a huge scale, and this is where INSIGHT’s datasets could play a vital role, given their unprecedented scale and quality.
Health data poverty
How do we reduce global health inequalities? Technology has an important role to play, but is often based on data that is not representative of diverse global populations.
Through the scale and quality of its datasets, representing the diverse populations of London and Birmingham, INSIGHT has the potential to address some of the imbalances in traditional research. By providing a boost for eye health data research in general, INSIGHT can help to identify areas where data is lacking, for example for conditions like cataracts, trachoma and refractive error that are rare in the UK, but in common in low- and middle-income countries.
Artificial Intelligence and the role of doctors
How will the arrival of artificial intelligence (A.I.) in the clinic affect the role of the doctor? A.I. is already helping to bring about quicker, more accurate diagnosis and improved treatments through an unbiased evaluation of all available data. But this does not mean that A.I. is about to replace doctors. Instead, it should allow them to focus more on the things humans are much better at than machines. This includes things like delivering news to patients in a relatable, sympathetic way, understanding how a condition affects daily life, gaining a patient’s trust, and discovering where the problem really lies. So, by relieving strain on health systems, A.I. will not only help doctors make better decisions, it will also make healthcare more human by giving doctors more time with patients when it matters.
Images (top to bottom): 1) Dr Pearse Keane with a patient at Moorfields, 2019. Credit: Moorfields Eye Hospital (library shot); 2) Image of fundus showing scatter laser surgery for diabetic retinopathy. Credit: National Eye Institute, National Institutes of Health, Public domain, via Wikimedia Commons; 3) Histopathology of Alzheimer's disease in the CA3 area of the hippocampus (coronal section). Credit: Mikael Häggström and brainmaps.org, CC BY 3.0, via Wikimedia Commons; 4) OCT retinal scan, 2019. Credit: Moorfields Eye Hospital (library shot); 5) Patient being examined at Moorfields, 2018. Credit: Anton Webb / UCL Institute of Ophthalmology; 6) Woman wearing a facemask, gown and surgical gloves. Credit: Photo by Jakayla Toney on Unsplash; 7) Man taking a visual acuity test at Moorfields, 2018. Credit: Anton Webb / UCL Institute of Ophthalmology; 8) Hands of four different people laid out on a table. Credit: Photo by Clay Banks on Unsplash; 9) A patient undergoing an eye test at Moorfields, 2018. Credit: Anton Webb / UCL Institute of Ophthalmology
Please note, photos of people in shared spaces without facemaks pre-date the Coronavirus pandemic.