Related content

Streamlining Clinical Research and AI Development using Junior Doctors in Medical Image Analysis


Newcastle University, an NHS trust, and have collaborated to advance medical image analysis in ophthalmology, at the forefront of research in AI development.

ophthalmologist using an OCT scan machine with patient

The clinical research project, OCTAHEDRON, will use machine learning to detect early signs of neurodegenerative diseases (e.g. Parkinson’s and Alzheimer’s) in Optical Coherence Tomography (OCT) scans of the retina. Globally, nearly 1 in 6 people suffer from neurodegenerative diseases.


The OCTAHEDRON project is led by Anya Hurlbert (Professor of Visual Neuroscience) at Newcastle University’s Biosciences Institute. Dr Rahele Kafieh is developing the algorithm for the AI application helped by Jaume Bacardit, Reader in Machine Learning and Jenny Read, Professor of Vision Science.

Notably, the project represents an ambitious collaboration between academia, the health service, and industry. With no fewer than 12 senior contributors from academia and the NHS, the project capitalises upon the knowledge of senior consultants in ophthalmology including Dr Jeffry Hogg, Mr David Steel, and Dr Will Innes, from the Newcastle Hospitals NHS Foundation Trust.

The OCTAHEDRON project is funded by the National Institute for Health Research (NIHR).


In 2021, provided the OCTAHEDRON team with a bespoke, on-premises deployment of an early version of the platform.’s software enabled the project leaders to manage their annotation team which includes nearly 50 junior doctors as well as senior ophthalmology consultants.

Using’s platform, the annotators used drawing tools to annotate the boundaries between different tissue layers in OCT scans of retinas. Easy to use tools are crucial for accurately annotating the different layers of the retina – a challenging task even for experienced ophthalmologists.

Driven by the project’s bespoke needs, also developed a feature to allow the team to integrate their own plug-ins which is now being implemented on’s main platform. By customising the platform in this way, the team could ensure that the software fully fits their requirements, for example by embedding their own scoring plug-in to enable the junior doctors’ analyses to be assessed.


By using for medical image analysis, the OCTAHEDRON team has saved time in their AI development in four ways, by:

  • Making it easier for senior consultants, who are time poor, to participate in the collective annotation process
  • Enabling a large team of junior doctors to annotate images, thus reducing the time required of the senior consultants
  • Removing the need to develop their own annotation tools and secure data storage, instead benefiting from’s expertise in this area
  • Ensuring the project had the necessary technical support, and continuity in software provision.’s software enabled OCTAHEDRON to use data in keeping with the tight security controls within the NHS and their ethics approvals. As the platform is end-to-end encrypted, the team can now incorporate NHS-derived OCT scans to improve and enlarge their dataset. All whilst knowing that the data is highly secure.

The OCTAHEDRON team worked very closely with to inform the development of the software platform. As well as providing critical user feedback and testing, the project has helped shape’s platform by highlighting some specific needs for facilitating effective image analysis in clinical research.


“Clinical research is very specialised, so it was great to find a company that specialises in tools for annotating medical images. has been instrumental to our research by providing the critical software that we needed.”

Prof Jenny Read, Newcastle University


OCT scans (readily available in the UK’s public health sector, and increasingly in high-street opticians) are able to scan a patient’s eye within a few seconds. Crucially, these scans could potentially indicate whether a patient may have a neurodegenerative disorder. 

OCTAHEDRON is capitalising on its leading edge research, expert knowledge, and machine learning techniques to develop an AI application that could more efficiently identify patients who might have neurodegenerative diseases, based on OCT scans. This would help detect disorders sooner, improve patient outcomes and reduce healthcare costs.

The project takes place against the backdrop of the UK Government plan to increase the country’s capacity and capability to deliver cutting-edge research.