A virtual VI rehabilitation assistant

Professor Chris Dickinson, University of Manchester - £81,449

September 2020 - December 2021

Research summary

Many people with sight loss or macular disease require extra support from eye clinics, charities and low vision services. However, often those who need it most are not aware of the resources available. Professor Dickinson, from the University of Manchester, looked to create a chat-bot assistant to help answer these questions and signpost patients to useful resources.

What’s the problem?

When people with macular disease start losing their central vision, this can have a serious impact on their ability to do day-to-day tasks and can affect their mental wellbeing. There are many services available to those with a visual impairment. However, these are not always available, promoted or discussed with patients. Therefore, by creating a virtual assistant, Professor Dickinson hopes that all those with visual impairments will be able to access the help they need.

What did this project achieve?

This project surveyed over 100 people with a visual impairment, those that support someone with a VI, and rehabilitation officers, to create a list of questions that the chat-bot system should be able to answer. They also asked how people would feel about a virtual assistant and what issues needed to be taken into account when designing it. 
With all this information, the researchers designed and tested a prototype with a list of question and answers which was trialled by seven participants with visual impairments. Overall, the participants found the system easy to use and understand, and found the interaction to be a positive experience.

What’s next?

The next step for this research is to fine-tune the chat-bot to improve the inclusivity of speech recognition. The researchers will also aim to incorporate mental health support and cognitive behavioural therapy (CBT) interventions and improve the database of patient information. The ultimate goal is that the chat-bot would be able to automatically understand patient information needs and be able to provide suitable resources.  

Publications

Taylor, J., Subramanian, A., Freitas, A., Mendes Ferreira, D., & Dickinson, C. (2022). Qualitative investigation of information behaviour in individuals with visual impairment to inform virtual digital-assistant design. Invest. Ophthalmol. Vis. Sci.