Enhancing Understanding of Long COVID Using Novel Mathematical Clustering Techniques

The COVID pandemic continues to have a detrimental impact on lives. As of January 2022, the ONS estimated that there are 1.5 million people in the UK experiencing Long COVID. Around 65% of these have symptoms which lead to an adverse effect on day-to-day activities.
Understanding of Long COVID is however still in its early stages and there are many unanswered questions. Can we group people by severity of their symptoms and how long they persist? Are different variants of the COVID-19 virus related to different symptoms? Can we identify factors associated with more debilitating forms of Long COVID?
This intradisciplinary project will bring together techniques from pure mathematics, statistics and data science to address these questions, in full collaboration with clinicians and patients. It will make use of data from cohort studies and electronic records.
The project will develop ideas from graph theory and topological data analysis to produce groups of similar patients called clusters. These will allow the identification of Long COVID sufferers and of factors indicating an increased Long COVID risk. The project will also provide easy-to-use computer tools to implement and visualize the methodology. Appropriate training in pure mathematics, statistics or computing will be provided.
The project's results will benefit the NHS, allowing treatments and rehabilitation to be targeted at the most affected people.