| Congenital heart disease is a congenital genetic disease that can seriously affect the development of children and adolescents.Therefore,early detection and early treatment of such diseases have great significance to the healthy growth of young people.According to statistics,about 12 to 150,000 children with congenital heart disease are born every year in China.It has great significance for the healthy development of such children to be involved in the study of early diagnosis aid technology for congenital heart disease.Because of the heart sound is a non-stationary random signal,based on the short-time Fourier transform and neural network method,this paper designs an Android-based congenital heart sound heart sound acquisition and analysis equipment.The device can provide early clinical diagnosis for children in the western region who lack medical measures by determining whether the heart signal has pathological murmurs.The specific principle of the diagnosis system is that the heart sound signal of the detection sample is collected by a sensor,and after the digital-to-analog conversion circuit,the signal is digitized and then transmitted to the mobile device via Bluetooth.The software part completes the two functions of heart sound signal collection,signal playback.An algorithm was designed for this and the heart noise signal was detected by the cell phone recording.The heart sound preprocessing adopts fast wavelet transform and fragrant energy transform,and intuitively distinguishes S1 and S2 heart sound signals through the transformed envelope components.The short-time Fourier transform was used to extract the frequency components of heart sound signals of different diseases.The components were identified by BP neural network algorithm.By training and recognizing the 468 signals in the collected heart sound database,the recognition rate is 80%.The result indicates that the heart sound signal collected by the collection device is of good quality. |
PDF Full Text Request