Radio frequency integrated circuits for biomedical magnetic and magnetoelectric microsystems

Abstract

Design and analysis of the advanced biomedical circuit and systems in wide variety of applications has emerged a significant interest. Not only in different engineering disciplines, but also in a variety of applications such as neuroscience, COVID-19, etc. In this study, we are proposing an implantable device, handheld device for detecting different diseases and the RFIC design for the ME antenna and passive devices and sensor evaluations to diagnose different diseases. First, we show a miniaturized implantable device for deep brain implantation that provides wireless power transfer efficiency (PTE) of 1 to 2 orders of magnitude higher than the reported micro-coils for brain stimulation. The proposed device will simultaneously measure the ac magnetic field activity when neurons are firing and providing the energy harvesting for stimulation. A magnetoelectric (ME) antenna is fabricated to be integrated to the on-chip energy harvesting circuit for brain-computer interface applications. The proposed ME antenna resonates at the frequency of 2.57 GHz while providing a bandwidth of 3.37 MHz. The proposed rectangular ME antenna wireless power transfer efficiency is 0.304 %, which is considerably higher than that of micro-coils. Many of the next-generation implantable medical devices that are targeting sub-mm scale form factors are entirely powered wirelessly. The most commonly used form of wireless power transfer for ultra-small receivers is inductive coupling, and has been so for many decades. This might change with the advent of novel microfabricated magnetoelectric (ME) antennas which are showing great potential as high-frequency wireless powered receivers. We compare these two wireless power delivery methods using receivers that operate at 2.52 GHz with a surface area of 0.043 mm2. Measurements results show that the maximum achievable power transfer of a ME antenna outperforms that of an on-silicon coil by approximately 7 times for a Tx-Rx distance of 0.76 cm and 3.3 times for a Tx-Rx distance of 2.16 cm. The proposed ME antenna used in sound noise dosimetry system is also presented in this study to investigate and measure the sound noise level. The ME antenna is utilized to sense the sound noise level through the nonlinear antenna modulation (NAM) around the operating acoustic resonance frequency of 63.6 MHz. Applied noise level of 102 dBA at two different 300 kHz and 1 kHz frequencies results in 33.8 dB and 28.7 dB signal to noise ratio (SNR). The fabricated antenna provides a high frequency resonance of 2.49 GHz for energy harvesting and communications. In the second part we will go over the RFIC design for the bio-implant devices, evaluation of the ME antennas for communication purposes and the circuit interface to drive the ME and GMI sensors. First, we will discuss the RF energy harvesting circuit designs when interfaced with the discrete SMD inductor to improve the matching and energy harvesting performance is described. The proposed prototype provides 51% efficiency at 10kΩ output load while providing more than 10s of nC and µA injected charge and current which is enough for the neural stimulation. A low-noise amplifier (LNA) topology with tunable input matching and noise cancellation utilized in a Bluetooth receiver frontend is introduced and described in this study, which was designed and optimized to interface with a magnetoelectric (ME) antenna in a 0.35 µm MEMS-compatible CMOS process. Input matching at the LNA-antenna interface is controlled with a circuit that varies the effective impedance of the gate inductor using a control voltage. Tunability of 455 MHz around 2.4 GHz is achieved for the optimum S11 frequency with a control voltage range of 0.3 V to 1.2 V. The proposed LNA has a noise cancelling feedback loop that improves the noise figure by 4.1 dB. The post-layout simulation results of the LNA show a 1-dB compression point of -7.4 dBm with an S21 of 17.8 dB. The down conversion in receiver frontend is done from 2.41 GHz to 2.14 MHz with the conversion gain of 21.1 dB. Besides, a miniaturized CMOS oscillator using microelectromechanical system (MEMS) resonating at 159 MHz frequency is designed and simulated in 0.35µm XFAB technology to drive ME sensors. The fabricated ME sensor offers quality factor of 653. The proposed oscillator provides a phase noise as low as -131.3 dBc/Hz at 10 kHz and -137.9 dBc/Hz at 100 kHz offset frequencies while consuming 2.24 mW power. For final part, we will discuss the handheld device design for early diagnosis of different diseases such as, lung cancer, Alzheimer, Covid-19, etc through exhaled breath on the molecularly imprinted polymer (MIP) gas sensors. A novel gas sensor has been developed that might be applied to diagnose Covid-19 from the exhaled breath instantly. The handheld device is designed to read the sensor activities and send the data to the android phone to show if the patient is at risk or not. For this purpose, a lock-in amplifier is designed to read the resistance in ac domain and transmit the digitized data through Bluetooth communication link.--Author's abstract