Design And Implementation Of Heart Pre-diagnosis System Based On Smartphone Platform

In 2008 the World Health Organization released a report on that:Heart disease is becoming the number one killer of human health. The early prevention and accurate diagnosis of cardiovascular disease which is of great concern had becoming an important research. Heart sound signal contains plenty of heart information; phonocardiogram can reflect the early lesions information of cardiac. Therefore the heart sound detection has the irreplaceable advantages and clinical value, which cannot be replaced by ECG. Early prevention of heart disease is a long-term project. Due to time or financial reasons, people usually cannot adhere to this project. So the development of a portable and popular sound prevention device, which could reduce the incidence of heart disease, posses a great social meaningful and economic value.The heart pre-diagnosis system based on the Android intelligent mobile phone is the combination of pre-diagnosis and mobile terminal. It is a multifunctional heart sound pre-diagnosis system developed on the Android platform. The multi-function includes collection, display, classify and transmission. It is designed to achieve the popularization and portable of the sound pre-diagnosis. The system includes three parts:heart sound collection part, heart sound display part, heart sound pre-diagnosis part. The collection part comprises of hardware and software. The collection device records heart sound signal by using heart sound stethoscope, processes per-treatment of heart sound by using the filtering device, carries through analog-digital conversion by using ADC module and sends the heart sound signal to smart phone finally. The software part receives the heart sound signal and records it in WAV format via the microphone interface of the mobile phone. The display part mainly contains displaying, playing and transmission of heart sound. By analyzing the format of waveform audio file to read the audio file, defining the frame size of the waveform audio file and taking the maximum volume value of each frame as amplitude. Then, drawing the waveform audio file with time as the horizontal axis and volume size as vertical axis we can get PCG(phonocardiogram). Benefiting from the advantages of the Android platform, the transmission function can be realized by calling the existed transmission methods of the mobile phone. The pre-diagnosis part divides the heart sound into normal and abnormal two categories. The earlier work of the classification includes the extraction of heart sound envelop, partition and localization, feature extraction. Through the comparison of several classification methods, a graphical representation and discriminated classification strategy is proposed. A kind of graph and text representation which is user-friendly and simple is designed to describe the results of the classification.This paper contains the theoretical research and application development. The theoretical research studies the preprocessing methods of heart sound and classification methods of heart sound. The application development includes hardware side and software side. The hardware side designs a device for the heart sound collection which integrated with a sound filter function. The software development processes with Java in Eclipse integrated environment. The smart mobile phone classifier has two main innovation points:classification strategy and graphical representation of the results of pre-diagnosis. It has not only achieved the expected goal that makes the collection, display, transmission and pre-diagnosis function integrated together on the simulator, but also finished debugging on the real phone and can run very well on HTC Desire HD(G10).