A Novel J wave Detection Method Based on Massive ECG Data and MapReduce

摘要

J wave is an ECG sign of many clinical syndrome and the accurate detection about it is conducive to the clinical diagnosis of J wave syndrome. Under the background of ECG big data, a novel J wave detection method based on massive ECG data and MapReduce is proposed, which use data mining technology to detect abnormal ECG signal, especially J wave. Firstly, the characteristic of ECG time and frequency domain signal are extracted, and the information gain of every feature is extracted; then, the decision tree is used to classify and recognize ECG signal; lastly, above process are implemented under the parallel programming model MapReduce so that the massive ECG data can be handled, to detect J wave accurately. In order to test and verify the validity of this method, all the ECG data of MIT-BIH are used to do the experiment, and the results demonstrated that, the accuracy and specificity of the proposed method are satisfactory, which provides a new research mentality for the detection of many clinical syndrome.