Close
The Computational Platform for the Assessment of Cognition In Dementia
C-PLACID is a three-year Engineering & Physical Sciences Research Council (EPSRC) funded programme led by Sebastian Crutch, and involves an exciting new collaboration with the UCL Department of Computer Science. The programme aims to improve upon the assessments of cognition that are currently available to people with or at-risk of dementia.
Assessment of cognitive impairment
Cognitive impairment (for example, problems with memory, language and reasoning) is the most debilitating and frustrating aspect of most neurodegenerative diseases. As a result, assessing a person’s cognition is often a central feature of neuroscientific research and neurological diagnostic services, as well as being the most typical component by which the efficacy of pharmaceutical and non-pharmaceutical therapies is judged.
Despite the importance of quantifying cognition in both research and clinical settings, current cognitive assessments are often limited to a pen-and-paper format with cut-off scores, yielding a number of limitations, namely:
- A failure to capture critical aspects of performance (e.g. response time)
- A limit to the tests’ dynamic range
- An inability to adequately assess some domains of cognition
Whilst pragmatic, the traditional assessment of cognition remains in stark contrast to the approach applied to neuroimaging data, the analyses of which routinely and readily comprise sophisticated forms of statistical measurement and computational principles.
The potential of applying the latter approach to the assessment of cognition remains to be realised by both researchers and clinicians.
Aims of C-PLACID
This project aims to improve current cognitive assessments through the application of engineering, mathematical and computational principles, to achieve an accurate and sensitive measure of cognitive status in individuals with or at-risk of dementia, and healthy controls. C-PLACID’s goals have been engineered to address longstanding problems of cognitive assessment:
- Develop machine-learning algorithms and event –based models based on existing datasets to generate multidimensional cognitive profiles, and describe and track longitudinal change for different forms of dementia.
- Improve upon current cognitive assessments by including additional metrics of cognitive ability, with a view to yield a more informative assessment of cognitive status. This approach involves devising methods by which to measure voice reaction times automatically, applying psychophysical principles to test measurements, and establishing details about eye movements during the assessments.
- Innovate techniques to assess cognition, through two main avenues:
- The development of an ‘instruction-less’ paradigm that will evaluate cognitive status from the information taken from individuals’ eye movements whilst watching a screen.
- Investigating the potential of virtual reality to adequately assess individuals’ social cognition in a realistic and clinically relevant manner.
Impact
C-PLACID aims to benefit patients, carers, researchers and clinicians through the development of shorter and more personalised assessments that will yield more information than ever before about an individual’s cognition.
The adapted and novel tests will be developed to act as a powerful tool for clinicians, with the aim to not only provide a more accurate and sensitive diagnosis of dementia, but to estimate future cognitive capacity at a particular stage of disease progression based on the their own previous data.
This capacity of the adapted and novel tests will also be of particular benefit to patients’ families, in preparing emotionally, practically and financially for the effects of such a trajectory.
In terms of impact on research, the outcome of C-PLACID’s collaboration between the Dementia Research Centre and the UCL Department of Computer Science will promote novel interdisciplinary collaborations, with the aim to harness engineering and computational skills and technology not previously applied to clinical evaluations of cognition.
