Predicting Stroke Outcomes From Physiological Patterns
Stroke is one of the major diseases that can cause human deaths. However, despite the frequency and importance of stroke, there are only a limited number of evidence-based acute treatment options currently available. Recent research has indicated that the early changes in common physiological variables represent a potential therapeutic target, thus the manipulation of these variables may eventually yield an effective way to optimise stroke recovery. However, the effects and function domain of those physiological determinants are still unclear. Therefore, developing a relatively accurate prediction method of stroke outcome based on justifiable determinants becomes more and more important to the decision of the medical treatment at the very beginning of the stroke. Although there exist some initial statistical analyses on typical physiological variables, such as blood pressure, glucose, the accuracy is still far from satisfactory. In this work, we take a novel data mining based method to find correlations between physiological parameters of stroke patients, during 48 hours after stroke, and their 3-months stroke outcomes. Our approach not only consider statistical parameters of physiological data, but also include trend analyses of physiological data changes. We test our methods on a real data set of stroke patients registered at Royal Brisbane and Womens Hospital, Australia. Experiment results demonstrated that compared against methods only considering statistical variables, our methods can reach a high precision accuracy, 93.8％, and also a high recall accuracy, 87.4％.
Stroke Outcome Prediction Temporal Pattern
Pengjie Ye Yang Xie Qing Zhang Chaoyi Pang
The Australian e-Health Research Centre/CSIRO ICT Centre, Royal Brisbane and Womens Hospital, QLD, The University of New South Wales,Sydney, NSW, Australia The Australian e-Health Research Centre/CSIRO ICT Centre, Royal Brisbane and Womens Hospital, QLD, The Australian e-Health Research Centre/CSIRO ICT Centre, Royal Brisbane and Womens Hospital, QLD,
国际会议
南京
英文
1-8
2012-11-23（万方平台首次上网日期，不代表论文的发表时间）