The study


The rate and accuracy of dementia diagnosis varies greatly. According to ARUK figures from 2016 only 59% of people currently living with dementia receive a formal diagnosis. Current methods for diagnosis are expensive and intrusive, including brain scans and expensive spinal fluid tests. It is important to seek complementary cost-effective non-invasive methods to support timely and accurate diagnosis.


To assist dementia diagnosis and monitoring, we aim to use computational methods to create a method for detecting changes in linguistic ability that is cost effective and can be embedded in user-friendly mobile technology in the future.


We propose to develop new approaches for tracking cognitive decline based on the analysis of longitudinal spoken and written language, collected using a tablet application that encourages users to reminisce in speech and writing. We plan to develop automated computational methods for measuring topic transition, syntactic and semantic coherence, emotional fluctuation as well as social interaction based on language data by participants with dementia and healthy controls. We can then use the patterns of change over time as predictors for presence or progression of dementia.

A tablet application for recording conversations and written thoughs based on images from the past has already been developed at the University of Warwick, in collaboration with a University spin-out company, Clinvivo. Recruitment of participants with dementia and age matched controls is expected to take place in Summer 2017 and the first round of data collection is expected to start shortly after.

The study has received ethics approval from the Research Ethics Committee (REC) and the Health Research Authority (HRA), with reference number 16/WS/0226.

For more details, please see the participant information leaflet.