Up to 20% of people aged over 65 experience strong declines in memory, problem solving abilities and speed of processing – clinically defined as mild cognitive impairment (MCI). With an ageing population, there is an urgent need to identify who is at risk of MCI and how associated decline can be stemmed. The game of chess with its historical records of games, represent a unique opportunity to track temporal signatures of cognitive performance over many years. The focus of the project is to provide proof-of concept that chess performance data can be analysed to obtain longitudinal measures of cognitive function, measures which can predict individuals’ trajectories of cognitive decline.
The progression from normal ageing to clinical outcomes, such as dementia-related diseases, is defined by a transitional period termed mild cognitive impairment (MCI). Approximately 12% of MCI patients develop dementia, yet the MCI period is often psychometrically indistinguishable from the normal ageing, as both are described by gradual declines in memory, problem-solving abilities, and reaction time measures. Previous studies showed that performance in the domains of expertise is sensitive to the gradual age-related declines seen in the case of MCI patients. There is evidence of chess players noticing declines in their play three years before the clinical diagnosis of MCI. In this work, we propose to investigate how changes in complex cognitive processing can be used as a screening tool for MCI development. In particular, we will utilise longitudinal move-by-move decisions made by chess practitioners to develop statistical models that can identify types of errors made by players who experience strongest declines in their performance over their lifetime.
Unlike other domains, chess boasts organised and structured records of the activity and gameplay of millions of people going back several decades freely available in chess repositories. In this study, the longitudinal move-by-move decisions in the games will be used as a representation of a cognitive data stream, where each played game represents the player's journey over the lifetime in the space of possible good decisions, errors and blunders. We plan to use already existing datasets of expert chess players, but also collect new data using online chess rooms and online questionnaires.
In the first part of the project, we plan to use artificial neural networks to analyse a collection of chess moves with the primary goal to predict game outcomes, as well as the changes in the player’s rating scores across their career. In general terms, our models will identify players who change their decision-making process in the later stages of their career and unexpectedly start making chess moves that result in frequent losses or strongest age-related declines in rating.
In the second phase of the project, we plan to collaborate with experts in chess with the two-fold aim: 1) to interpret predictive patterns of errors indicative of steep declines in rating and 2) to collect an additional set of measures to validate the developed model. Second part will consist of collecting information on observed difficulties in playing chess, keeping up with hobbies, but also recognizing problems in other daily behaviours, such as remembering appointments.
The scope of the complete project is to develop an early identification tools for mild cognitive impairment that utilises rich longitudinal data on performance measures in the domain of peoples’ expertise.
Applications are open to students from both the UK and overseas, though we note that due to funding constraints the availability of positions for students with overseas fee status will be more limited. We anticipate competition for these studentships to be very intense. We would expect applicants to have an excellent undergraduate degree in a relevant discipline. We would also expect applicants to have completed or be undertaking a relevant master’s degree to a similar very high standard (or have equivalent research experience).
--
Please complete a University Postgraduate Research Application form available here: https://www.sheffield.ac.uk/postgradapplication/
Please clearly state the prospective main supervisor in the respective box and select ‘Neuroscience’ as the department.
After the application closing date, we will shortlist applicants for an online interview. We expect to carry out interviews (each lasting approximately 30 minutes) on Tuesday 27th April (am, GMT) and Tuesday 4th May (pm, GMT). If you are shortlisted for interview, we will aim to inform you of this no later than the end of Friday 23rd April. If you are unable to attend at the specified times, please let us know if we confirm that we would like to interview you.
Continue with Facebook
