Ultra low power switched-capacitor wavelet-based ADC for ECG analyzer

Abstract

A huge number (40%) of all human deaths are caused by cardiovascular diseases. In the past decades, a lot of research was done about the cause and the prevention of cardiovascular diseases. An implantable device, such as a pacemaker, is a well known method to prevent death by cardiac problems. This thesis explores the possibility of improving detection and reducing power consumption of ECG detectors. To find a suitable structure of our design, basic information about cardio signal and the heart is needed. Analog sense amplifier method is the method in this thesis to detect the QRS complex (obtains from depolarization of the ventricular in the heart) and the T-wave (creates by ventricle repolarisation) of IECG (intracardiac electrogram) signal. This method consists of analog signal processing block and an A/D converter. A wavelet filter (WF) is used to design the first part. In the introduction, wavelet transform analysis, mother wavelet, the desired approximation and suitable decomposition are discussed. These are parts of the wavelet filter. The wavelet filter is designed in discrete time domain due to its gain accuracy and low power consumption. Moreover, the switched-capacitors filters have an accurate frequency response as well as good linearity and dynamic range. The thesis discussed the design of discrete-time switched-capacitor (SC) fully differential folded cascode amplifier. The second part of the analog sense amplifier is the A/D converter, which consists of a sinh transconductor and a peak detector circuit. As the WF implemented in discrete-time, the detector block should do this as well. That is why the sample and hold (SH) block is used at its input of the detector block. The binary output is expected to be seen from the QRS complex and the T-wave of the heart signal. The whole system with an ECG signal input is built in five scales, which is needed to zoom in the different frequencies of the ECG signal. Every scale is implemented for different resolutions by using different clock frequency signals. Another way to implement the different scales is to change the ratio of the capacitors. This method is not used in this thesis, because the minimum values of the capacitors are already chosen. The system is able to detect the ECG signal with the ultra low power and small area design.