An adaptive ENG amplifier for tripolar cuff electrodes

Electroneurogram (ENG) recording from tripolar cuff electrodes is affected by interference signals, mostly generated by muscles nearby. Interference reduction may be achieved by suitably designed amplifiers such as the true-tripole and quasi-tripole systems. However, in practice their performance is severely degraded by cuff imbalance, resulting in very low output signal-to-interference ratios. Although some improvement may be offered by post filtering, this considerably increases complexity, size and power dissipation, rendering the approach unsuitable for the development of a high-performance ENG recording system which is fully implantable. This paper describes an integrated, fully implantable, adaptive ENG amplifier developed to automatically compensate for cuff imbalance, and thus significantly improve the quality of the recorded ENG. Measured results show that the adaptive ENG amplifier has a yield of 100%, a cuff imbalance correction range of more than ±40%, and an output signal-to-interference ratio of about 2/1 (6 dB) even for ±40% imbalance. The latter should be compared with an input signal-to-interference ratio of 1/500 (-54 dB). The circuit was fabricated in 0.8-μm BiCMOS technology, has a core area of 0.68 mm², and dissipates 7.2 mW from ±2.5 V power supplies. The adaptive ENG amplifier advances the state-of-the-art in implantable tripolar nerve cuff electrode recording techniques.