Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU

J. Arnin; D. Kahani; H. Lakany; B. A. Conway

doi:10.3217/978-3-85125-682-6-33

Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU

J. Arnin, D. Kahani, H. Lakany, B. A. Conway

Biomedical Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

40 Downloads (Pure)

Abstract

Online BCI has become a fascinating field of research nowadays. One of the main challenges in this field is to reduce the latency caused by the computational complexity of the signal processing algorithms. This issue leads to difficulty in processing real-time data. Usually, a trade-off needs to be considered between the number of input samples and precision of the processing algorithms. In this paper, heterogeneous computing concept is investigated to alleviate the computational complexity occurred in real-time processing. An OpenCL was utilized to implement signal processing algorithms in parallel. Feature extraction methods including band power and statistical moments were selected to examine the power of heterogeneous computing using parallel sum reduction. As a result, varying the number of work-group sizes which is an essential parameter of parallel processing provided dissimilar computing times. Also, running at a higher sampling rate yielded a higher benchmark ratio between sequential and parallel. However, system optimization is still necessary when processing BCI in real time.

Original language	English
Title of host publication	Proceedings of the 8th Graz Brain Computer Interface Conference 2019, Bridging Science and Application
Place of Publication	Graz, Austria
Number of pages	4
DOIs	https://doi.org/10.3217/978-3-85125-682-6-33
Publication status	Published - 20 Sept 2019
Event	40th International Conference of the IEEE Engineering in Medicine and Biology Society - Honolulu, Hawaii, United States Duration: 17 Jul 2018 → 21 Jul 2018 https://embc.embs.org/2018/

Conference

Conference	40th International Conference of the IEEE Engineering in Medicine and Biology Society
Abbreviated title	EMBC 2018
Country/Territory	United States
City	Honolulu, Hawaii
Period	17/07/18 → 21/07/18
Internet address	https://embc.embs.org/2018/

Keywords

signal processing
brain-computer interfaces
BCI systems
medicine
electroencephalogram
EEG
OpenCL

Access to Document

10.3217/978-3-85125-682-6-33

Arnin-etal-BCIC-2019-Heterogeneous-real-time-multi-channel-time-domain-feature-extraction-using-parallel-sum-reduction-on-GPUFinal published version, 571 KBLicence: CC BY-NC 4.0

IEEE Brain Data Bank Challenges and Competitions
Lakany, H. (Recipient), Kahani, D. (Recipient), ARNIN, J. (Recipient), Uttamchandani, D. (Recipient), Ren, J. (Recipient), Murray, P. (Recipient), Di Caterina, G. (Recipient) & Zabalza, J. (Recipient), 31 Oct 2017
Prize: Prize (including medals and awards)

Cite this

@inproceedings{945fb614c8cb43b68bd54fd554c30d18,

title = "Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU",

abstract = "Online BCI has become a fascinating field of research nowadays. One of the main challenges in this field is to reduce the latency caused by the computational complexity of the signal processing algorithms. This issue leads to difficulty in processing real-time data. Usually, a trade-off needs to be considered between the number of input samples and precision of the processing algorithms. In this paper, heterogeneous computing concept is investigated to alleviate the computational complexity occurred in real-time processing. An OpenCL was utilized to implement signal processing algorithms in parallel. Feature extraction methods including band power and statistical moments were selected to examine the power of heterogeneous computing using parallel sum reduction. As a result, varying the number of work-group sizes which is an essential parameter of parallel processing provided dissimilar computing times. Also, running at a higher sampling rate yielded a higher benchmark ratio between sequential and parallel. However, system optimization is still necessary when processing BCI in real time.",

keywords = "signal processing, brain-computer interfaces, BCI systems, medicine, electroencephalogram, EEG, OpenCL",

author = "J. Arnin and D. Kahani and H. Lakany and Conway, {B. A.}",

year = "2019",

month = sep,

day = "20",

doi = "10.3217/978-3-85125-682-6-33",

language = "English",

isbn = "9783851256826",

booktitle = "Proceedings of the 8th Graz Brain Computer Interface Conference 2019, Bridging Science and Application",

note = "40th International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 ; Conference date: 17-07-2018 Through 21-07-2018",

url = "https://embc.embs.org/2018/",

}

Arnin, J, Kahani, D, Lakany, H & Conway, BA 2019, Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU. in Proceedings of the 8th Graz Brain Computer Interface Conference 2019, Bridging Science and Application. Graz, Austria, 40th International Conference of the IEEE Engineering in Medicine and Biology Society, Honolulu, Hawaii, United States, 17/07/18. https://doi.org/10.3217/978-3-85125-682-6-33

TY - GEN

T1 - Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU

AU - Arnin, J.

AU - Kahani, D.

AU - Lakany, H.

AU - Conway, B. A.

PY - 2019/9/20

Y1 - 2019/9/20

N2 - Online BCI has become a fascinating field of research nowadays. One of the main challenges in this field is to reduce the latency caused by the computational complexity of the signal processing algorithms. This issue leads to difficulty in processing real-time data. Usually, a trade-off needs to be considered between the number of input samples and precision of the processing algorithms. In this paper, heterogeneous computing concept is investigated to alleviate the computational complexity occurred in real-time processing. An OpenCL was utilized to implement signal processing algorithms in parallel. Feature extraction methods including band power and statistical moments were selected to examine the power of heterogeneous computing using parallel sum reduction. As a result, varying the number of work-group sizes which is an essential parameter of parallel processing provided dissimilar computing times. Also, running at a higher sampling rate yielded a higher benchmark ratio between sequential and parallel. However, system optimization is still necessary when processing BCI in real time.

AB - Online BCI has become a fascinating field of research nowadays. One of the main challenges in this field is to reduce the latency caused by the computational complexity of the signal processing algorithms. This issue leads to difficulty in processing real-time data. Usually, a trade-off needs to be considered between the number of input samples and precision of the processing algorithms. In this paper, heterogeneous computing concept is investigated to alleviate the computational complexity occurred in real-time processing. An OpenCL was utilized to implement signal processing algorithms in parallel. Feature extraction methods including band power and statistical moments were selected to examine the power of heterogeneous computing using parallel sum reduction. As a result, varying the number of work-group sizes which is an essential parameter of parallel processing provided dissimilar computing times. Also, running at a higher sampling rate yielded a higher benchmark ratio between sequential and parallel. However, system optimization is still necessary when processing BCI in real time.

KW - signal processing

KW - brain-computer interfaces

KW - BCI systems

KW - medicine

KW - electroencephalogram

KW - EEG

KW - OpenCL

U2 - 10.3217/978-3-85125-682-6-33

DO - 10.3217/978-3-85125-682-6-33

M3 - Conference contribution book

SN - 9783851256826

BT - Proceedings of the 8th Graz Brain Computer Interface Conference 2019, Bridging Science and Application

CY - Graz, Austria

T2 - 40th International Conference of the IEEE Engineering in Medicine and Biology Society

Y2 - 17 July 2018 through 21 July 2018

ER -

Heterogeneous real-time multi-channel time-domain feature extraction using parallel sum reduction on GPU

Abstract

Conference

Keywords

Access to Document

Fingerprint

Prizes

IEEE Brain Data Bank Challenges and Competitions

Cite this