Application Research On Visualization Of ECG Data Sequences Based On Variant Measurements

With the rapid development of modern science and technology, people’s the quality of life has improved generally, the individual’s health awareness has increased gradually. Every year, many enterprises and department asking the stuff to do a physical examination is a normal In the current era of massive data, digging and processing the useful information of diagnosis cardiovascular diseases hidden in the mass of data has become a research focus. And with along producing massive ECG data sequences of every time a medical examination, how processing a large sample of massive ECG data sequences is an application problem worth of studying. Because the description of the human body function needs to use a large number of linear and nonlinear relationships to form a complex model and system, and describing the state of the heart requires the help for complex nonlinear system about dynamics. So, it is a worth exploring and studying frontier subject how to make use of the nonlinear model and method to deal with large samples of massive ECG data and to visualize the results in high dimensional space. Currently, in the direction of this study, the main processing model is based on modern chaos theory and French scientists Poincare 100 years ago proposed classical graphic method, obtained result by these models is a series of ECG scatter diagrams.In this paper, we propose a visualization method based on the variant logic system, which is used to process a large sample of massive ECG data and display its processing results in high dimensional space. Variant model and visualization method is used to detect binary variant, DNA sequences and discrete random sequences test aspects, and obtain a series of research findings, this paper will extend its application range from binary variant random sequences to research aspect of multi valued continuous random sequences.Technology of ECG variant visual characteristic analysis is based on continuous signal sequence analysis technique of developing and original sequence cipher analysis model, this kind of technology can decompose effectively the one-dimensional multi-value ECG sequences into pseudo genes multiple discrete data sequence of four basic elements, and form the visual distribution diagrams and obtain ECG scatter diagrams. Data source of the paper is a batch of practical measured ECG data (22 ten thousands,500MB) of the first people’s Hospital of Yunnan Province, them are diagnosed by doctors. We chose relevant data to generate ECG scattered point maps, and make comparative analysis. Actual maps display:visual distributions of the normal ECG data sequences and the other types of abnormal ECG data sequences are very significant distinguishability. These preliminary results are using massive classified data to conduct further tests, so that the later stage of the project use this model and in order to lay the foundation for diagnosis of cardiovascular diseases.Firstly, the paper introduced research background, research status, research meaning and basic theory of the project; secondly, the paper described system architecture and processing of massive ECG data; then, the processing results of the serial ECG data were visualized and generated a series of ECG scatter plots, and the distribution of the maps were compared and analyzed; finally, summary and prospect.