Litcius/Paper detail

An Implantable Neuromorphic Sensing System Featuring Near-Sensor Computation and Send-on-Delta Transmission for Wireless Neural Sensing of Peripheral Nerves

Yuming He, Federico Corradi, Chengyao Shi, Stan van der Ven, Martijn Timmermans, Jan Stuijt, Paul Detterer, Pieter Harpe, Lucas Lindeboom, Evelien Hermeling, Geert Langereis, Elisabetta Chicca, Yao‐Hong Liu

2022IEEE Journal of Solid-State Circuits75 citationsDOIOpen Access PDF

Abstract

This article presents a bioinspired, event-driven neuromorphic sensing system (NSS) capable of performing on-chip feature extraction and “send-on-delta” pulse-based transmission, targeting peripheral nerve neural recording applications. The proposed NSS employs event-based sampling which, by leveraging the sparse nature of electroneurogram (ENG) signals, achieves a data compression ratio of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$&gt; 125\times $ </tex-math></inline-formula> , while maintaining a low normalized rms error (NRMSE) of 4% after reconstruction. The proposed NSS consists of three sub-circuits. A clockless level-crossing (LC) analog-to-digital converter (ADC) with background offset calibration has been employed to reduce the data rate, while maintaining a high signal to quantization noise ratio (SQNR). A fully synthesized spiking neural network (SNN) extracts temporal features of compound action potential (CAP) signals and consumes only 13 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{W}$ </tex-math></inline-formula> . An event-driven, pulse-based body channel communication (Pulse-BCC) with serialized address-event representation (AER) encoding schemes minimizes transmission energy and form factor. The prototype is fabricated in 40-nm CMOS occupying a 0.32-mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> active area and consumes in total 28.2 and 50 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{W}$ </tex-math></inline-formula> power in feature extraction and full diagnosis mode, respectively. The presented NSS also extracts temporal features of CAP signals with 10- <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{s}$ </tex-math></inline-formula> precision.

Topics & Concepts

Neuromorphic engineeringComputer scienceCMOSChipElectronic engineeringComputer hardwareArtificial neural networkReal-time computingArtificial intelligenceEngineeringTelecommunicationsAdvanced Memory and Neural ComputingNeuroscience and Neural EngineeringFerroelectric and Negative Capacitance Devices