Search

This article presents the design and efficient hardware implementation of binarized neural networks (BNNs) for brain-implantable neural spike sorting. In contrast to the conventional artificial neural networks (ANNs), in which the weights and activation functions of neurons are represented using real values, the BNNs utilize binarized weights and activation functions to dramatically reduce the memory requirement and computational complexity of the ANNs. The designed BNN is trained using several realistic neural datasets to verify its accuracy for neural spike sorting. The application-specific integrated circuit (ASIC) implementation of the designed BNN in a standard 0.18- $\mu \text{m}$ CMOS process occupies 0.33 mm 2 of silicon area. Power consumption estimation of the ASIC layout shows that the BNN dissipates $2.02~\mu \text{W}$ of power from a 1.8 V supply while operating at 24 kHz. The designed BNN-based spike sorting system is also implemented on a field-programmable gate array and is shown to reduce the required on-chip memory by 89% compared to those of the alternative state-of-the-art spike sorting systems. To the best of our knowledge, this is the first work employing BNNs for real-time in vivo neural spike sorting. [ABSTRACT FROM AUTHOR]

Conventional in vivo neural signal processing involves extracting spiking activity within the recorded signals from an ensemble of neurons and transmitting only spike counts over an adequate interval. However, for brain-computer interface (BCI) applications utilizing continuous local field potentials (LFPs) for cognitive decoding, the volume of neural data to be transmitted to a computer imposes relatively high data rate requirements. This is particularly true for BCIs employing high-density intracortical recordings with hundreds or thousands of electrodes. This article introduces the first autoencoder-based compression digital circuit for the efficient transmission of LFP neural signals. Various algorithmic and architectural-level optimizations are implemented to significantly reduce the computational complexity and memory requirements of the designed in vivo compression circuit. This circuit employs an autoencoder-based neural network, providing a robust signal reconstruction. The application-specific integrated circuit (ASIC) of the in vivo compression logic occupies the smallest silicon area and consumes the lowest power among the reported state-of-the-art compression ASICs. Additionally, it offers a higher compression rate and a superior signal-to-noise and distortion ratio.

Intracortical brain computer interfaces (iBCIs) utilizing extracellular recordings mainly employ in vivo signal processing application-specific integrated circuits (ASICs) to detect action potentials (spikes). Conventionally, "brain-switches" based on spiking activity have been employed to realize asynchronous (self-paced) iBCIs, estimating when the user involves in the underlying BCI task. Several studies have demonstrated that local field potentials (LFPs) can effectively replace action potentials, drastically reducing the power consumption and processing requirements of in vivo ASICs. This article presents the first LFP-based brain-switch design and implementation using gated recurrent neural networks (RNNs). Compared to the previously reported brain-switches, our design requires no exhaustive learning phase for the estimation of optimal recording channels or frequency band selection, making it more applicable to practical asynchronous iBCIs. The synthesized ASIC of the designed in vivo LFP-based feature extraction unit, in a standard 180-nm CMOS process, occupies only 0.09 mm 2 of silicon area, and the post place-and-route synthesis results indicate that it consumes 91.87 nW of power while operating at 2 kHz. Compared to the previously published ASICs, the proposed LFP-based brain-switch consumes the least power for in vivo digital signal processing and achieves comparable state estimation performance to that of spike-based brain-switches.

Objective. Advances in brain-machine interfaces (BMIs) can potentially improve the quality of life of millions of users with spinal cord injury or other neurological disorders by allowing them to interact with the physical environment at their will. Approach. To reduce the power consumption of the brain-implanted interface, this article presents the first hardware realization of an in vivo intention-aware interface via brain-state estimation. Main Results. It is shown that incorporating brain-state estimation reduces the in vivo power consumption and reduces total energy dissipation by over 1.8× compared to those of the current systems, enabling longer better life for implanted circuits. The synthesized application-specific integrated circuit (ASIC) of the designed intention-aware multi-unit spike detection system in a standard 180 nm CMOS process occupies 0.03 mm 2 of silicon area and consumes 0.63  µ W of power per channel, which is the least power consumption among the current in vivo ASIC realizations. Significance. The proposed interface is the first practical approach towards realizing asynchronous BMIs while reducing the power consumption of the BMI interface and enhancing neural decoding performance compared to those of the conventional synchronous BMIs., (Creative Commons Attribution license.)

While brain-implantable neural spike sorting can be realized using efficient algorithms, the presence of noise may make it difficult to maintain high-peformance sorting using conventional techniques. In this article, we explore the use of partially binarized neural networks (PBNNs), to the best of our knowledge for the first time, for sorting of neural spike feature vectors. It is shown that compared to the waveform template-based methods, PBNNs offer robust spike sorting over various datasets and noise levels. The ASIC implementation of the PBNN-based spike sorting system in a standard 180-nm CMOS process is presented. The post place and route simulations results show that the synthesized PBNN consumes only 0.59 μ W of power from a 1.8 V supply while operating at 24 kHz and occupies 0.15 mm 2 of silicon area. It is shown that the designed PBNN-based spike sorting system not only offers comparable accuracy to the state-of-the-art spike sorting systems over various noise levels and datasets, it also occupies a smaller silicon area and consumes less power and energy. This makes PBNNs a viable alternative towards the implementation of brain-implantable spike sorting systems., Competing Interests: Conflict of interestThe authors have no competing interests to declare., (© Korean Society of Medical and Biological Engineering 2022, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.)