This thesis presents a novel high performance miniature wireless neural recording system that is able to acquire neural activity sensed from microelectrodes implanted directly in the brain and transmit the data wirelessly to a remote receiver. The core of the system is a customized integrated circuit (ASIC) which is composed of 16 low-power and low-noise neural amplifiers, 16 low-power analog-to-digital converters (ADCs), a digital processor, a low-power short-range RF transmitter and a low-power phase-locked loop (PLL). All of these blocks are optimized for power consumption using innovative circuit design techniques. The integrated circuit amplifies, filters and digitizes neural signals from 16 channels and then transmits them out wirelessly, via the license-free ISM band (2.4-2.5GHz). The ASIC is fabricated in an AMS 0.18-μm CMOS process and assembled on a miniature PCB with a crystal reference clock, a small Zinc Air battery and its battery holder, a headstage connector, an external loop antenna and several decoupling capacitors. The thesis proves that, due to the innovative low power IC design techniques proposed, the 16-channel, 12-bit system’s power consumption can be reduced to 3.1mW, which makes possible continuous recording from a very small battery for 72 hours. The whole miniature system weighs 1.47g including a small battery. Experimental results demonstrate that the system is capable of acquiring all potential neural signals of interest, including EEG signals, local field potentials (LPF) and action potentials.
The major contribution of the proposed system is to significantly boost the battery lifetime of the relevant state-of-the-art systems from 1-15 hours to 72 hours while still maintaining the highest recording resolution, bandwidth, channel number and one of the smallest sizes and weights. As a result, it has provided the neuroscience community with a miniature, lightweight and high-performance wireless technology for recording neural activity without tethering or restraining the test subjects. Moreover, its small size and weight further improve the welfare of the test subjects and encourage them to move freely and behave normally and naturally without affecting their affective states, yielding the most accurate and faithful correlation between subjects’ neural activities and behavioral outcomes.