Ward developed an AI-based eye disease detection service to prevent irreversible vision loss.
To prevent permanent vision loss from eye diseases such as glaucoma and retinopathy, it is important to diagnose them in time. AI is a promising technique for detecting these asymptomatic eye diseases and has the potential to make diagnosing these diseases easier and more affordable. While tests in labs with AI systems often seem promising, adoption in the medical world often fails. The aim of this thesis was to design a service around the AI system to adapt it to the needs of different medical specialists.
This project started with the discovery phase, where knowledge was gathered through literature research, observation and interviews with experts. In the develop phase, the findings from the discover phase are translated into design goals. In the deliver phase, various methods are used to validate the concept and the final design is created in a subsequent iteration.
This resulted in the design of a new AI- based eye disease detection service that uses opticians for pre-diagnosis. The opticians take a fundus image and run it through an AI system that categorises patients based on the severity of their condition. Ksyos then sends a letter to the patient explaining the next steps in the diagnosis. The optometrists interpret the AI system’s decisions and decide whether they agree with them. Eventually the AI system will learn and will provide more accurate diagnosis.
