Sinus Heart Rate Turbulence Trend Detection Based On Piecewise Linear Fitting

Heart rate turbulence (HRT), an important indicator of cardiovascular disease riskstratification,is a phenomenon that sinus heart rates following a ventricular prematurebeat accelerate before slowing down. Due to there are no uniformly standard methodsand parameters in HRT detection currently, the RR interval sequences of heart reatturbulence are used to detect HRT in most cases. Therefore it is really necessary todevelop the method furtherly. In this paper, we propose a piecewise linear fittingmethod to approximate the curve of HRT, which is able to describe the trend of HRTaccurately. Our work in this paper mainly contains three aspects as follows:The first step is to collect HRT samples. In this paper, we regard the MIT/BIHdatabase as the data source, and preprocess the ECG signals. Then we collect HRTsamples for further research.The second step is to research time series model epresentation method. In view ofthe time series data processing ideas, the piecewise linear method is introduced todetect sinus rhythm turbulence trend in this paper. Experimental results show that theproposed method is with fast computation speed, easy implementation, low complexityand good linear results.The last step is to achieve the trend representation based on the piecewise linearfitting method. On the basis of the segmentation by piecewise point, the least squaremethod error is combined to help the piecewise linear fitting. The feature that the slopis the liner is combined with TO, TS to analyze the turbulence trend. Through theMATLAB simulation, the algorithm can achieve good results.