Zhao Liu, a PhD student affiliated with Kent and Medway Medical School (KMMS), has had his research selected for a poster presentation at a major international conference, Advancing Health in AI, in China last week.
The research is entitled ‘Analysing Behavioural Motion Data Across Virtual and Physical Environments: Enhancing Understanding of the Impact of Cognition on Behaviour with LSTM Networks’.
The study investigates the effectiveness of the Long Short-Term Memory (LSTM) network (a kind of deep learning model) in analysing motor data collected from healthy participants.
Traditional cognitive testing usually uses specific tasks with a particular purpose, such as drawing a clock or recognising shapes. However, such test components are not commonly found in real life, so may not truly reflect the cognitive ability of a person in real-life situations.
Zhao’s study collected data from healthy participants as they engaged in a recognizable and familiar task – making sushi – in three different environments: a) fully immersive Virtual Reality (VR), b) the physical world, and c) a 2D digital tablet setup. By collecting and analysing behavioural data in these settings, the study seeks to refine and optimise the LSTM model to improve the accuracy of predicting and classifying behavioural data, laying the groundwork for future analysis of patient data, particularly those with cognitive decline such as dementia.
Zhao’s project is a collaboration between Dr Daniel Soria and Professor Chee Siang Ang at the University of Kent’s School of Computing, and Professor Sukhi Shergill, Co-Director of Research at KMMS and Director of Research at Kent and Medway Social Care Partnership Trust.
The Advancing Health in AI conference was organized by the Eye Hospital, Wenzhou Medical University.
More details about the conference are available on the Nature Conference website.