چكيده لاتين
Abstract
Bioimpedance is the electrical response of living tissues to the passage of an alternating current; and it is used in applications such as non-invasive health monitoring and wearable systems. In these applications, in addition to accuracy, ultra-low power consumption is essential to increase battery lifetime. With the development of low-power circuits, in many designs power reduction has been mainly focused on amplification, conversion, and processing blocks, while power dissipation in the output stage has received less attention.
In this research, a low-power sub-block for bioimpedance measurement and separation of its real and imaginary components is presented and simulated in Cadence using a 180-nm CMOS technology. The core of the proposed architecture is the use of a switching output stage in the excitation path so that, by storing the energy of the injected current in an inductive element and recovering part of it, the dissipated power is reduced. To extract the real and imaginary parts, small-signal AC analysis is performed on the proposed circuit; and then, at each frequency, the amplitude and phase are extracted and the complex impedance is formed. Subsequently, the Cole model parameters are fitted by solving a minimization problem, and the real and imaginary components of the impedance at the injection frequency are reconstructed. Moreover, to reduce power, the effect of changing the dimensions of the power transistors on the total power is investigated and the optimum point is selected.
Simulation results show that the total power consumption of the circuit is 37.2 µW and the dissipated power is 14.67 µW. Accuracy evaluation in four sets of experiments based on variations of the Cole model parameters indicates that the reconstruction error of |Z| remains below 2.1%, and an average measurement accuracy of 97.9% is achieved; specifically, the mean |Z| error is in the range of 0.17% to 1.2%, and the maximum |Z| error is reported to be less than 2.1%. Also, in the worst case, the errors of the real and imaginary components at the injection frequency are approximately 1.77% and 1.85%, respectively. Overall, the proposed architecture, while preserving the accuracy of component separation, enables reduction of power consumption and dissipated power, and is suitable for low-power wearable sensors.
Keywords: Low-power, Bio-impedance, Impedance measurement, Integrated circuit, Wearable sensor, Impedance spectroscopy (EIS).