Automatic detection of ST depression on ECG

摘要

Automatic and quick detection of abnormal signals in electroencephalogram (ECG) could help cardiovascular patients. The optimal threshold value of correlation coefficients was explored to judge ST depression from the abnormal ECG signals. The optimal threshold was determined by the cross validation analysis based on a correlation coefficient between the ECG data on the template in ST depression and other diseases. As the results of this analysis, the optimal threshold of the correlation coefficient was around 0.8 in both the linear and spline interpolation. Moreover, the calculated threshold was little affected by the type of linear or spline interpolation and data length (100, 200, and 300 points for the normalization). These results could be useful for setting the application of smartphones or tablets to reduce the computation time in online analysis.