Research On Key Technologies Of Terminal Oriented Portable ECG Signal Acquisition System

In recent years,with the increasing prevalence of cardiovascular diseases in China and the improvement of people's attention to medical care,the focus of medical treatment in the future will shift from doctors to patients and from hospitals to personal terminals,which requires the miniaturization and portability of biomedical devices,which stimulates the industrial development of portable and wearable medical electronic devices in disguise.On the other hand,wireless body area network(WBAN),human body communication technology(HBC),large-scale integrated circuit and digital signal processing technology provide system solutions for the miniaturization and portability of biomedical equipment.ECG is a very important bioelectric signal in human body.It can reflect the health state of human heart and help prevent cardiovascular diseases.However,ECG is a sparse signal with low frequency and small amplitude.It is easy to be disturbed by biological electrode resistance,polarization voltage,50/60 Hz power line power frequency interference,motion artifact,EMG signal and so on.Therefore,Portable ECG acquisition system demand for better performance.Starting with the key technologies and circuit implementation of analog front-end circuit,analog-to-digital converter(ADC)and feature extraction module,this dissertation focuses on the charge buffer technology,chopper modulation technology,charge buffer technology,low-power ADC design technology,preprocessing technology of feature parameter extraction module and adaptive threshold iteration technology.Those technlogies provide a technical scheme for the design of ECG signal acquisition system.Aiming at the interference and noise in ECG signal acquisition,this dissertation designs a low-power and high-performance analog front-end circuit,which integrates capacitive coupled chopper amplifier module,programmable gain amplifier module,programmable low-pass filter module and charge buffer module from the index of analog front-end circuit.The capacitive coupled chopper amplifier suppresses the influence of polarization voltage.The introduction of chopper technology eliminates the1/f noise and instrument amplifier offset.The programmable gain amplifier realizes the maximum signal-to-noise ratio of different ECG signal strengths.The programmable low-pass filter filters out the high-frequency clutter introduced by chopper to prevent signal aliasing in subsequent analog-to-digital converters.And the charge buffer eliminates the sampling error caused by voltage drop during sampling.The simulation results show that the equivalent input noise of the front-end circuit is only 0.6?V.The input impedance is 1.38 G?.The common mode rejection ratio and power rejection ratio is 116 d B and 87 d B,respectively.And the power consumption is 48?W.According to the sparse characteristics of ECG signal,this dissertation proposes a dynamic tracking algorithm suitable for ECG signal prediction,and designs a 12 bit low-power analog-to-digital converter based on the algorithm.While completing the digital conversion of ECG signal,the analog-to-digital converter can also output prediction information.Combined with the QRS wave detector designed in this dissertation,the system can identify the QRS complex.Therefore,the heart rate extraction is realised and further alleviates the hardware overhead of signal transmission and storage cost.the circuit and layout of the ECG acquisition system are conducted on the 130 nm complementary metal oxide semiconductor(CMOS)process.The test results of analog-to-digital converter in heart rate extraction module show that the effective number of bit of analog-to-digital converter is 10.85 and the minimum power consumption is 900 n W.also,the characteristic parameter extraction circuit can effectively detect QRS complex.In terms of the early warning of cardiovascular disease,this dissertation proposes a code-recombination quantization based intelligent pathological recognition module,which can output a characteristic signal K value characterizing the feature of the input signal,and further applied to make the learning set and test set of pathological recognition.Also,this dissertation proposes an adaptive threshold iterative technology based on Kalman filter theory,which can quickly respond to the change of input signal and improve the accuracy of R-wave detection in QRS complex.Based on the 130 nm CMOS process,the circuit and layout design,fabricating,packaging and testing of the code-recombination pathological recognition module are carried out.Test results show that the effective number of bit of the analog-to-digital converter in the pathological recognition module are 9.38.The power consumption is only 40 n W,and the area is only 0.11 mm~2.Meanwhile,the output K value is verified by the system model of off chip pathological recognition.The experimental results show that the accuracy of pathological recognition of four common cardiovascular diseases is above 96%.