Analysis And Study Of Hypertrophic Cardiomyopathy Based On Time Domain And Frequency Domain Characteristic Parameters Of Heart Sound

With the popularity of medical knowledge,more and more diseases have been familiar with people.Hypertrophic cardiomyopathy(HCM)as a high hereditary,high morbidity and high risk disease,it can cause sudden death,heart failure and atrial fibrillation.More and more medical workers and researchers at home and abroad are involved in the diagnosis and treatment of HCM Over the years.Due to the level of medical constraints,HCM patients in China’s screening rate is not high,the vast majority of patients only in the onset of treatment to the hospital and then missed the early intervention of the golden period.It caused a heavy burden for the family and society.Besides,the patient will appear chest tightness,shortness of breath,syncope and other symptoms with the exacerbations,more serious cases may occur even sudden death.Therefore,the establishment and improvement of HCM early diagnosis system for the patient’s condition control is significance to practical treatment.Heart sounds contain heart valve,the basic information of the functional status of each organ,and have been widely used in the diagnosis and treatment of cardiovascular disease.The clinical auscultation shows that the second heart sound division may appear in the HCM patients of left ventricular outflow tract obstruction.What is worse,the systolic murmur will appear because of the left ventricular outflow tract turbulence and mitral regurgitation.The analysis of heart sounds is of great clinical value in the diagnosis and treatment of HCM,meanwhile,and has a very important significance to the HCM screening and even human health census.Based on the research work of Hypertrophic Cardiomyopathy Diagnosis and Genetics Counseling Center in the First Affiliated Hospital of the Fourth Military Medical University,the detection and analysis of heart sounds are applied in the diagnosis of HCM in this paper.Mainly completed the following work:(1)HCM heart sound signal acquisitionWe use the device produced by BIOPAC in the United States to collect and synchronize the heart sound signal in real time in the visual environment.The device is also simple to operate and the quality of the data acquisition is high.We used this system to collect the tricuspid valve,mitral valve and aortic valve second auscultation area three parts of the heart sound data between the age of 20 to 70 years old include 24 humas(15 males and 9 females)and 30 HCM patients(male 17,female 13),and then remove the low-quality heart sounds.Finally,57 patients with normal heart sounds(24 normal subjects)and 81 HCM heart sounds(28 HCM patients)were selected for this study.HCM heart sound signal preprocessingClinical HCM heart sound data acquisition process will inevitably be affected by the internal structure of the instrument and the external environment,resulting in additional noise.Therefore,this paper firstly analyzes the noise components and main sources of the heart sound signal,and then describes the noise reduction method and process of the heart sound signal.According to the nature of various types of noise,the use of 50 Hz frequency trap,digital bandpass filter and adaptive wavelet threshold noise reduction method to eliminate the heart sound in the frequency interference and additional noise to improve the heart sound signal signal to noise ratio,and finally get a better noise reduction effect.Feature Extraction of HCM Heart Sound SignalIn this paper,the HCM heart sound signal characteristics are extracted in the time domain and frequency domain.We discussed the effects of Hilbert transform(HT),normalized average Shannon energy(NASE),single degree of freedom model(SDOF)and frequency conversion homomorphic filtering(FMH)to the waveform.By comparing and analyzing,the FMH algorithm has better feature waveform extraction.After the feature waveform extraction,the HCM heart sound signal is segmented by local peak detection and adaptive double threshold threshold,and the time domain characteristic parameters and time domain energy characteristic parameters are defined.Finally,the wavelet packet is used to extract the frequency bands energy distribution of the normal heart sounds and HCM heart sounds in the VLF,LF,MF and HF,which lays the foundation for the study of the characteristics of follow-up clinical HCM heart sounds.(4)Study on Time Domain and Frequency Domain Characteristics of Clinical HCM Heart SoundThe characteristic parameters of HCM heart sound were extracted by statistical analysis in the time domain and frequency domain by using the noise reduction method and characteristic waveform extraction method.Comparing the parameters of heart sounds with normal and HCM,We discussed the HCM first heart sound,the second heart sound characteristics and HCM heart sound noise in the location,time,intensity and frequency range.By comparing the time domain energy of the obstruction and non-obstructive HCM heart sound in the resting state and the energy band parameters of the frequency domain wavelet packet,the characteristics of the noise can be obtained,and the obstruction of the left ventricular outflow tract is judged.It has significant importance to clinical value for the diagnosis of HCM.In summary,We have carried out a comprehensive and in-depth study on clinical HCM heart sound data acquisition,processing and analysis.Based on the analysis of clinical HCM heart sounds,the first heart sounds,second heart sounds and noise characteristics of HCM patients were described,and the clinical symptoms were analyzed.The clinical basis was established for HCM diagnosis based on phoneme analysis.Due to the relatively small amount of data can only be a preliminary analysis.Therefore,the future work will focus on the establishment of clinical HCM heart sound database,and put forward more effective HCM heart sound analysis method for the establishment of a sound clinical HCM diagnostic system.