Recurrence Plot Was Detected And Compared With Other Methods To Inhibit The Outbreak

With the development of EEG signal research, more and more experts pay closeattention to the burst-suppression pattern (BSP). The detection of burst suppression is notjust for administration of anesthetic or sedative agents, but for the cerebral-protectivetreatment of various neurosurgical diseases. But the EEG data length is short, its smoothdegree is poor and its anti-property is weak. So it’s hard to detect the BSP and so far thereisn’t a relatively perfect method. Therefore, this paper will work on this topic.First, carefully study the recurrent plot method and grasp its characteristics. On thisbasis, choose the appropriate parameters including embedding dimension m, delay timeτand the critical distance r. And draw three recurrent plots of burst, suppression and normalstates to test the effectiveness of this method.Then, recurrence quantification analysis (RQA) can better describe the dynamicalcharacteristics. Three basic RQA indexes including the recurrence rate (RR), determinism(DET) and recurrence entropy (ENTR) are calculated. On this basis, the boxplot, theone-way analysis of variance (ANOVA) and multiple comparison test are employed toexamine the performance of above three indexes in order to select the most appropriateindex.Last, so far there are some methods to detect the BSP including spectral analysis,bispectral analysis, approximate entropy and the Nonlinear Energy Operator (NLEO). Theperformance of RR analysis is investigated in comparison with above four methods tolook for a better method to detect BSP in the EEG series.