Dublin, 23rd March, 2022: Ischemic stroke occurs when a blood clot blocks the blood flow in an artery in the brain and it is one of the leading causes of mortality worldwide. New research by Dr Elizabeth Hunter and Professor John D Kelleher from the ADAPT SFI Research Centre for AI-Driven Digital Content Technology at TU Dublin has led to novel prediction models that are more accurate at identifying people under 60 stroke with a high risk of stroke who may not be identified using a traditional stroke risk prediction model.
The research forms part of the Horizon 2020 research project Precise4Q and was recently published in the Frontiers of Neurology Journal. The novel platform has the potential to impact millions of people at high-risk of ischemic stroke and will pave the way for the prediction and treatment of ischemic stroke in unprecedented ways for younger populations.
Speaking about the research, Professor Kelleher said: “Age is a critical risk factor in stroke prediction. However, unless we are careful in terms of how age is used to predict stroke risk then age can dominate the information that comes from other risk factors. Another aspect that we need to consider in predicting stroke risk is that not all risk factors contribute proportionally to stroke risk by age. We created a set of models that can predict individual risk of stroke by age group. This research shows that by creating different risk models for different age groups we are better able to model the contribution from multiple factors to stroke risk that is appropriate for each age category. Using this separate model by age group we are better able to predict and identify younger individuals with a high risk of stroke as compared with using traditional stroke risk prediction models.”
Stroke is one of the leading causes of mortality worldwide. Research shows that the annual direct medical costs for stroke in 2012 was 71.55 billion US dollars. This does not include the indirect cost to life and loss of time to work for those suffering from it. This research will directly impact society by helping identify people at highest risk of ischemic stroke, thus striving to mitigate those risks in a timely and efficient manner. Additionally, it will also impact patient care positively.
“This study is one of our many efforts in combining the available health data in stroke patients to create models that better aid medical professionals advising and treating patients at highest risk of stroke. Predictive models will enable us to be preventative in a way, and if data can help achieve this reality, it would be a giant stride for both the medical as well as the data disciplines,” commented Professor Kelleher.
This research examines whether the contribution of stroke risk factors to an individual’s 5-year stroke risk is non-proportional by age. Using real health data collected by a longitudinal study that began in 1948 in Massachusetts, among other relevant data sets collected for the purpose of research, the researchers can better understand the other factors involved and weigh them accordingly to better predict stroke.
More informatioin about the research work of John D Kelleher in this area is available on the ADAPT Radio podcast.
The full study is available online.