Detecting Ventricular Fibrillation and Ventricular Tachycardia for Small Samples Based on EMD and Symbol Entropy
In this paper, we proposed a new method based on Symbol Entropy and Empirical Mode Decomposition (EMD) to detect ventricular fibrillation (VF) and ventricular tachycardia (VT). Initially, we applied the EMD to decompose VF and VT signals into five sub-bands respectively. And then, we calculated the Symbol Entropy of each sub-bans as the feature to detect VT and VF. We employed the public data set to assess the proposed method. Experimental results showed that, using classification of support vector machine (SVM), the proposed method can successfully distinguish VF from VT with the classification accuracy up to 100 % based on small samples. The duration of each sample was 2 s. Moreover, the classification accuracy of the proposed method is far higher than the classification accuracy of the original signals using Symbol Entropy directly.
KeywordsSymbol entropy EMD VF VT SVM Small samples
This work was supported by the National Natural Science Foundation of China (Grant Nos. 61201428, 61302128), the Natural Science Foundation of Shandong Province, China (Grant Nos. ZR2010FQ020, ZR2013FL002), the Shandong Distinguished Middle-aged and Young Scientist Encourage and Reward Foundation, China (Grant Nos. BS2009SW003, BS2014DX015).
- 16.Xia, D., Meng, Q., Chen, Y., Zhang, Z.: Classification of ventricular tachycardia and fibrillation based on the Lempel-Ziv complexity and EMD. In: Huang, D.-S., Han, K., Gromiha, M. (eds.) ICIC 2014. LNCS, vol. 8590, pp. 322–329. Springer, Heidelberg (2014)Google Scholar