Neuromorphic Spike Data Classifier for Reconfigurable Brain-Machine Interface

Amir Zjajo; Sumeet Kumar; Rene van Leuken

doi:10.1109/NER.2017.8008314

Neuromorphic Spike Data Classifier for Reconfigurable Brain-Machine Interface

Amir Zjajo, Sumeet Kumar, Rene van Leuken

Signal Processing Systems

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

Abstract

In this paper, we propose a reconfigurable neural spike classifier based on neuromorphic event-based networks that can be directly interfaced to neural signal conditioning and quantization circuits. The classifier is set as a heterogeneity based, multi-layer computational network to offer wide flexibility in the implementation of plastic and metaplastic interactions, and to increase efficacy in neural signal processing. Built-in temporal control mechanisms allow the implementation of homeostatic regulation in the resulting network. The results obtained in a 90 nm CMOS technology show that an efficient neural spike data classification can be obtained with a low power (9.4 μW/core) and compact (0.54 mm2 per core) structure.

Original language	English
Title of host publication	8th International IEEE EMBS Conference on Neural Engineering (NER)
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	150-153
Number of pages	4
ISBN (Electronic)	978-1-5090-4603-4
DOIs	https://doi.org/10.1109/NER.2017.8008314
Publication status	Published - 2017
Event	8th International IEEE EMBS Conference on Neural Engineering - Shanghai, China Duration: 25 May 2017 → 28 May 2017 https://neuro.embs.org/2017/

Conference

Conference	8th International IEEE EMBS Conference on Neural Engineering
Abbreviated title	NER
Country/Territory	China
City	Shanghai
Period	25/05/17 → 28/05/17
Internet address	https://neuro.embs.org/2017/

Keywords

Neuromorphics
Kernel
Integrated circuit modeling
Axons
Computational modeling
Training

Access to Document

10.1109/NER.2017.8008314

http://ens.ewi.tudelft.nl/pubs/leuken17ner.pdf

Cite this

@inproceedings{48bd5b36b6a54d29894862b97057c33f,

title = "Neuromorphic Spike Data Classifier for Reconfigurable Brain-Machine Interface",

abstract = "In this paper, we propose a reconfigurable neural spike classifier based on neuromorphic event-based networks that can be directly interfaced to neural signal conditioning and quantization circuits. The classifier is set as a heterogeneity based, multi-layer computational network to offer wide flexibility in the implementation of plastic and metaplastic interactions, and to increase efficacy in neural signal processing. Built-in temporal control mechanisms allow the implementation of homeostatic regulation in the resulting network. The results obtained in a 90 nm CMOS technology show that an efficient neural spike data classification can be obtained with a low power (9.4 μW/core) and compact (0.54 mm2 per core) structure.",

keywords = "Neuromorphics, Kernel, Integrated circuit modeling, Axons, Computational modeling, Training",

author = "Amir Zjajo and Sumeet Kumar and {van Leuken}, Rene",

year = "2017",

doi = "10.1109/NER.2017.8008314",

language = "English",

pages = "150--153",

booktitle = "8th International IEEE EMBS Conference on Neural Engineering (NER)",

publisher = "IEEE",

address = "United States",

note = "8th International IEEE EMBS Conference on Neural Engineering, NER ; Conference date: 25-05-2017 Through 28-05-2017",

url = "https://neuro.embs.org/2017/",

}

TY - GEN

T1 - Neuromorphic Spike Data Classifier for Reconfigurable Brain-Machine Interface

AU - Zjajo, Amir

AU - Kumar, Sumeet

AU - van Leuken, Rene

PY - 2017

Y1 - 2017

N2 - In this paper, we propose a reconfigurable neural spike classifier based on neuromorphic event-based networks that can be directly interfaced to neural signal conditioning and quantization circuits. The classifier is set as a heterogeneity based, multi-layer computational network to offer wide flexibility in the implementation of plastic and metaplastic interactions, and to increase efficacy in neural signal processing. Built-in temporal control mechanisms allow the implementation of homeostatic regulation in the resulting network. The results obtained in a 90 nm CMOS technology show that an efficient neural spike data classification can be obtained with a low power (9.4 μW/core) and compact (0.54 mm2 per core) structure.

AB - In this paper, we propose a reconfigurable neural spike classifier based on neuromorphic event-based networks that can be directly interfaced to neural signal conditioning and quantization circuits. The classifier is set as a heterogeneity based, multi-layer computational network to offer wide flexibility in the implementation of plastic and metaplastic interactions, and to increase efficacy in neural signal processing. Built-in temporal control mechanisms allow the implementation of homeostatic regulation in the resulting network. The results obtained in a 90 nm CMOS technology show that an efficient neural spike data classification can be obtained with a low power (9.4 μW/core) and compact (0.54 mm2 per core) structure.

KW - Neuromorphics

KW - Kernel

KW - Integrated circuit modeling

KW - Axons

KW - Computational modeling

KW - Training

U2 - 10.1109/NER.2017.8008314

DO - 10.1109/NER.2017.8008314

M3 - Conference contribution

SP - 150

EP - 153

BT - 8th International IEEE EMBS Conference on Neural Engineering (NER)

PB - IEEE

CY - Piscataway, NJ

T2 - 8th International IEEE EMBS Conference on Neural Engineering

Y2 - 25 May 2017 through 28 May 2017

ER -

Neuromorphic Spike Data Classifier for Reconfigurable Brain-Machine Interface

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this