Research And Analysis On The ST Segment In ECG

Electrocardiogram(ECG) has becamed one of the necessary instruments used for clinic heart disease diagnosing,the veracity of ECG instruments is correlative to human being health and life. It has very important applied value and significance to research high-powered ECG instruments for improve the exact rate of diagnosing.ST segment analysis is one of the most important parts of ECG analysis. The waveform changes of ST segment will appear in ECG whenever the cardiac cell depolarization were disturbed by ischemia,electrolyte turbulence,drug,etc. Therefore,it is significant to recognize and locate the ST segment, and to measure them quickly and accurately in clinical. However, the physicians are not satisfied with current technology of automatic analysis of ST segment. It requires engineers and physicians working together to seek more effective processing methods. Be directed against the difficulties and characteristic in ST segment analysis, this thesis consists of following several parts:â‘ On account of the special request of ST segment analysis, this thesis designs a filter based on the theory of Zero-phase Digital Filter. Contrast the ECG spectrum before and after filter and test triangle signal with simulative baseline wander, the Zero-phase Digital Filter designed in paper can effectively filter interference of low frequency and rippingly keep the shape of ST segment.â‘¡The detection of all waveform has been obtained. We deeply studied the P wave detection algorithm based on the detection of R wave. A novel P wave detection algorithm based on approximating functions in the optimum region is proposed. In the algorithm we use a conic to approach the P wave by this means of least square method in every region, which is shifted in local range. Then we find the optimum approach region of P wave by comparing the approximating error. Finally, we attain the characteristic points of P wave by using the approximating functions to approach P wave in the optimum region. The performance of the algorithm tested using the records of the MIT-BIH database is effective and accurate. The other characteristic points of ECG signal is also studied in thesis.â‘¢The thesis distinguish the shapes of ST segment using BP neural network based on learning the neural network theory. This has detailed elucidation about the choice of sample, confirming the parameters and learning algorithm of BP neural network. Finally, we tested the BP neural network designed in thesis using a plentiful sample with diversified shapes. It's satisfying in the veracity.â‘£The several algorithms of the domestic and overseas which ST segment electric level measured is researched. Compare the results of the several algorithms and expert's notes analyzing the database in Europe ST-T database and obtain the trend of ST segment offset.The affecting factors that ST segment detected and the merits and demerits of several algorithms is analyzed which provides an evidence for ST segment detected in the thesis.