OCTAHEDRON DATABASE
Research type
Research Database
IRAS ID
293696
Contact name
Anya Hurlbert
Contact email
Research summary
Annotated, anonymised OCT retinal image database for improved diagnosis of ophthalmological and neurological disease
REC name
North East - Newcastle & North Tyneside 1 Research Ethics Committee
REC reference
21/NE/0159
Date of REC Opinion
10 Nov 2021
REC opinion
Further Information Favourable Opinion
Data collection arrangements
The data are specialist images of the inside of the eye from thousands of individuals, each with accompanying information about disease diagnoses that the individual carries, but with no personal identifying information. These eye images are retinal Optical Coherence Tomography (OCT) scans. OCT is a non-invasive, rapid technique that ophthalmologists as well as high-street opticians use to inspect the microscopic structure of the retina, the light-sensitive tissue that lines the back of the eye.
The database will contain existing OCT scans that have already been collected and stored in ophthalmology patients’ healthcare records. No new data will be collected. Instead, the existing image data will be linked to existing information about the neurological disease status of the patient. This data linkage will be done within the NHS computer systems that store patients’ electronic heath care records. After the linkage, all data in the database will be fully pseudonymised, containing no information such as names, addresses or hospital patient IDs that might identify individual patients. Individual health care records will not be stored within the database or accessible from it. Before being exported to researchers, the entries in the database will be completely anonymised.Research programme
Neurodegenerative diseases such as Parkinson’s and Alzheimer’s may cause changes in the retina, the light-sensitive tissue which lines the back of the eye. The retina is effectively an outgrowth of the brain. OCT scans provide rich detail about the three-dimensional structure of the retina, but the changes due to neurodegenerative disease are too subtle to be detected directly in these images, even by expert ophthalmologists. Researchers are working to develop Artificial Intelligence (AI)-based techniques to help identify these changes. To learn to recognise subtle patterns of disease, the AI software must study thousands or even millions of OCT scans, including some from patients who went on to be diagnosed with neurodegenerative disease. The key to creating the AI software is therefore a large dataset of OCT scans which have been labelled by disease status. The research teams developing these AI tools include computer scientists, neuroscientists, ophthalmologists and neurologists. Ultimately, this database of anonymised, annotated OCT scans will enable research to be carried out that will make diagnosis of neurodegenerative diseases more efficient. Neurological disorders will be detected sooner, so treatment can begin earlier. The reduced disability and improved outcomes will benefit patients and reduce costs to wider society.
Research database title
Annotated, anonymised OCT retinal image database for improved diagnosis of ophthalmological and neurological disease
Establishment organisation
Newcastle upon Tyne Hospitals NHS Trust
Establishment organisation address
Regent Point
Regent Farm Road
Gosforth
NE3 3HD