A Digitizing Process System For Heart Sound Signal

Heart sound signal contains important information about activities of cardiovascular system,which can be as the basis for diagnosing cardiovascular diseases. Especially, animal experiments and invasive and non-invasive clinical trails show that there is a very close relationship between the amplitude of the first heart sound (S1) and the cardiac contractility, so, the change trend of the amplitude of the first heart sound can be used to evaluate cardiac contractility reserve. Like clinical medicine, sports medicine also applies many methods to evaluate cardiac function. For example, electrocardiogram, which is the optimal method for analyzing chronotropism and dromotropism, but it can not be used to measure myocardial inotropism. Cardiac catheterization is objective and quantitative, but it is an invasive procedure, requires a special catheterization chamber and aseptic manipulation, and can not be used routinely. Echocardiography can measure and evaluate cardiac function noninvasively, structurally, and functionally, but can not be conveniently used at bedside, in the field, and some indicators are not sensitive. Cardiac blood-pool developing with radionuclide and nuclear magnetic resonance detection have high sensitivity and specificity, but are expensive, not easy to popularize. Maximal oxygen uptake and anaerobic metabolism threshold determination are objective, practical, and quantitative, but are affected by respiratory function, require complex device and professional technician, can only be used in a special laboratory, are not appropriate to be used in the everyday settings. In view of the above-mentioned, the author presented the project, " A Digitizing Process System for Heart Sound Signal ". This author has accomplished the following research tasks. The key step for computerized automated extraction of cardiac contractility signal from heart sound signal is recognition of the first heart sound. In order to rapidly and accurately measure the values of the basic indicators (cardiac contractility and heart rate) of heart sound, the author developed a digitizing process system for heart sound signal, which included hardware and software systems. It can record and display the waveforms of various types of heart sound signals. By using a cursor system, the cardiac cycle, the duration between two points on a phonocardiogram, and the amplitude of the first and second heart sounds can be accurately measured, and so can the third and fourth sounds if they exist on a phonocardiogram. By using the developed digitizing process system for heart sound signal, the author first studied the relevant regularity of the values of the basic indicators of heart sound,measured a number of heart sound samples from various individuals under various conditions (at rest and following exercise),obtained change relationship curves between systole and cardiac cycle, systole and heart rate, diastole and systole. Based on this, a method was developed,which was used to automatically and rapidly recognize S1 and S2 at rest and following exercise, calculate the basic indicators related to cardiac contractility and heart rate. This method can obtain energy envelop using normalized Shannon average energy method,and then, obtain short time average amplitude envelop of heart sound signal from the energy envelop. Since the short time average value of S1 is greater than that of S2, the difficult in calculation of beginning location for all heart sound signals both at rest and following exercise was resolved. As there are some characteristics and value range for cardiac cycle, heart sound, systole, and diastole, for example, systolic duration < diastolic duration, the problem of location of S1 and S2 was resolved, and the values of the amplitudes of S1 and S2 as well as heart rate can be given. In order to evaluate the reliability and accuracy of this method, a comparison between reading value, manual calculation, and the result of automatic recognition for a series of heart sound signals at rest or following exercise was performed, and there was a good consistence, therefore indicating the reliability and accuracy of this method. After evaluating the reliability and accuracy of this method, recognition and calculation of 22 heart sound samples at rest or following exercise were performed, and the results were all accurate, which suggests that this method has a certain practical value.