The Development Of Intelligent Electronic Stethoscope Based On Embedded System

Heart and respiration are important medical indexes contain the physiological and pathological information of heart and lungs. They are important basis for heart disease and respiratory system disease diagnosis. Now, in the diagnosis of heart diseases and respiratory diseases, auscultation is still an important method for diagnosis of doctors. However, the traditional stethoscope has many defects in auscultation, such as: difficult to capture the faint heart sounds and breathing sound signal, and easy to produce noise and the diagnosis results are often susceptible to the doctor’s subjective judgment, these factors have restricted the development of the stethoscope in medicine. Therefore, develop an intelligent electronic stethoscope, which can separate the heart and lung sound signal from noise, and can make the judgment of auscultation standard, is very necessary and has a broad market demand.After comparing several digital signal de-noising algorithm,this thesis proposes using the blind extraction algorithm based on generalized self related to extraction of heart sounds and breathing signals, removing interacting noise. This method is applicable to the heart and respiration signals’ cycle is predictable. In this case, we can achieve very good results to extract the heart and breath sounds from the mixed signals. And the extracted signal is analyzed both in time domain and frequency domain, provides the quantified diagnosis parameters.Two paths of heart and breath sounds acquisition and pre-processing circuit were designed based on blind signal extraction assumption requirements, including amplifier circuit with low noise; filter circuit and power amplification circuit, in order to complete the acquisition and pre-processing of heart sounds signal and breathing sounds signal. And the STM32 development settings were completed, including ADC, DAC, LCD and Flash settings.