PD3-2 Predicting Cardiovascular Disease from Clinical and HRV Analysis using Data Mining

• Published in: Journal of PHYSIOLOGICAL ANTHROPOLOGY 2007, Vol.26 (2), p.317-317
• Main Authors: Sang-Tae LEE, Ki-yong NOH, Heon-Gyu LEE, Keun-Ho RYU
• Format: Article
• Language: Japanese
• ISSN: 1880-6791

The most widely used signal in clinical practice is ECG, which is frequently recorded and widely used for the assessment of cardiac function. ECG is also a test that measures a heart's electrical activity, which provides valuable clinical information about the heart's status. Many researches have been pursued for heart disease diagnosis using ECG so far. However, electrocardio-graph has used foreign diagnosis algorithm due to inaccuracy of diagnosis results for a cardiovascular disease. In this paper, we propose a prediction method of data mining technique for extracting multi-parametric features by analyzing HRV from ECG, data preprocessing and cardiovascular disease pattern. The proposed method is a BCFP (Bayesian Classifier based on Frequent Patterns) based on the efficient FP-growth method. In theory, Bayesian classifiers have the minimum error rate compared to all other classifiers. However, in practice this is not always the case owing to inaccuracies in the unrealistic assumptions made for its use. The proposed BCFP algorithm addresses the problem of attribute dependence by discovering frequent patterns. Since the volume of patterns produced can be large in training phase, we offer a rule cohesion measure that allows a strong push of pruning patterns in the pattern-generating process. We conduct an experiment for the BCFP classifier, which utilizes multiple rules and pruning, and biased confidence (or cohesion measure) and dataset consisting of 670 participants distributed into two groups, namely normal people and patients with CAD (Coronary Artery Disease). Finally, our experiments show that the proposed prediction model achieves higher accuracy and is more efficient than other classifiers.