7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction

Sanjeev Tannirkulam Chandrasekaran; Imon Banerjee; Arindam Sanyal

doi:10.1109/ESSDERC53440.2021.9631777

7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction

Sanjeev Tannirkulam Chandrasekaran, Imon Banerjee, Arindam Sanyal

Diagnostic Radiology

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

Abstract

This work presents the first on-chip, mixed-signal echo state network (ESN) for early prediction of heart disease. The ESN comprises an input layer, a non-linear projection (NP) layer, and an output layer. Only the output layer of the ESN requires training. The input layer weights are time-invariant and drawn from a static binary random distribution. Thus, the proposed ESN has significantly lower trainable parameters compared to other non-linear neural networks used for similar prediction tasks. A 65nm prototype is validated with the Cleveland Heart Disease (CHD) dataset. The ESN achieves a mean accuracy of 84.6% over 5 test chips while consuming 7.5nJ energy/inference.

Original language	English (US)
Title of host publication	ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings
Publisher	Editions Frontieres
Pages	103-106
Number of pages	4
ISBN (Electronic)	9781665437479
DOIs	https://doi.org/10.1109/ESSDERC53440.2021.9631777
State	Published - 2021
Event	51st IEEE European Solid-State Device Research Conference, ESSDERC 2021 - Virtual, Online, France Duration: Sep 6 2021 → Sep 9 2021

Publication series

Name	European Solid-State Device Research Conference
Volume	2021-September
ISSN (Print)	1930-8876

Conference

Conference	51st IEEE European Solid-State Device Research Conference, ESSDERC 2021
Country/Territory	France
City	Virtual, Online
Period	9/6/21 → 9/9/21

Keywords

Machine learning
cardiac diseases prediction
data fusion and medical wearable
echo state network

ASJC Scopus subject areas

Electrical and Electronic Engineering
Safety, Risk, Reliability and Quality

Access to Document

10.1109/ESSDERC53440.2021.9631777

Cite this

Chandrasekaran, S. T., Banerjee, I., & Sanyal, A. (2021). 7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction. In ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings (pp. 103-106). (European Solid-State Device Research Conference; Vol. 2021-September). Editions Frontieres. https://doi.org/10.1109/ESSDERC53440.2021.9631777

7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction. / Chandrasekaran, Sanjeev Tannirkulam; Banerjee, Imon; Sanyal, Arindam.
ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings. Editions Frontieres, 2021. p. 103-106 (European Solid-State Device Research Conference; Vol. 2021-September).

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

Chandrasekaran, ST, Banerjee, I & Sanyal, A 2021, 7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction. in ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings. European Solid-State Device Research Conference, vol. 2021-September, Editions Frontieres, pp. 103-106, 51st IEEE European Solid-State Device Research Conference, ESSDERC 2021, Virtual, Online, France, 9/6/21. https://doi.org/10.1109/ESSDERC53440.2021.9631777

@inproceedings{6d5afa96ba1a4c7093da27e9bb530fd7,

title = "7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction",

abstract = "This work presents the first on-chip, mixed-signal echo state network (ESN) for early prediction of heart disease. The ESN comprises an input layer, a non-linear projection (NP) layer, and an output layer. Only the output layer of the ESN requires training. The input layer weights are time-invariant and drawn from a static binary random distribution. Thus, the proposed ESN has significantly lower trainable parameters compared to other non-linear neural networks used for similar prediction tasks. A 65nm prototype is validated with the Cleveland Heart Disease (CHD) dataset. The ESN achieves a mean accuracy of 84.6% over 5 test chips while consuming 7.5nJ energy/inference.",

keywords = "Machine learning, cardiac diseases prediction, data fusion and medical wearable, echo state network",

author = "Chandrasekaran, {Sanjeev Tannirkulam} and Imon Banerjee and Arindam Sanyal",

note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 51st IEEE European Solid-State Device Research Conference, ESSDERC 2021 ; Conference date: 06-09-2021 Through 09-09-2021",

year = "2021",

doi = "10.1109/ESSDERC53440.2021.9631777",

language = "English (US)",

series = "European Solid-State Device Research Conference",

publisher = "Editions Frontieres",

pages = "103--106",

booktitle = "ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings",

}

TY - GEN

T1 - 7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction

AU - Chandrasekaran, Sanjeev Tannirkulam

AU - Banerjee, Imon

AU - Sanyal, Arindam

PY - 2021

Y1 - 2021

N2 - This work presents the first on-chip, mixed-signal echo state network (ESN) for early prediction of heart disease. The ESN comprises an input layer, a non-linear projection (NP) layer, and an output layer. Only the output layer of the ESN requires training. The input layer weights are time-invariant and drawn from a static binary random distribution. Thus, the proposed ESN has significantly lower trainable parameters compared to other non-linear neural networks used for similar prediction tasks. A 65nm prototype is validated with the Cleveland Heart Disease (CHD) dataset. The ESN achieves a mean accuracy of 84.6% over 5 test chips while consuming 7.5nJ energy/inference.

AB - This work presents the first on-chip, mixed-signal echo state network (ESN) for early prediction of heart disease. The ESN comprises an input layer, a non-linear projection (NP) layer, and an output layer. Only the output layer of the ESN requires training. The input layer weights are time-invariant and drawn from a static binary random distribution. Thus, the proposed ESN has significantly lower trainable parameters compared to other non-linear neural networks used for similar prediction tasks. A 65nm prototype is validated with the Cleveland Heart Disease (CHD) dataset. The ESN achieves a mean accuracy of 84.6% over 5 test chips while consuming 7.5nJ energy/inference.

KW - Machine learning

KW - cardiac diseases prediction

KW - data fusion and medical wearable

KW - echo state network

UR - http://www.scopus.com/inward/record.url?scp=85123440862&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85123440862&partnerID=8YFLogxK

U2 - 10.1109/ESSDERC53440.2021.9631777

DO - 10.1109/ESSDERC53440.2021.9631777

M3 - Conference contribution

AN - SCOPUS:85123440862

T3 - European Solid-State Device Research Conference

SP - 103

EP - 106

BT - ESSDERC 2021 - IEEE 51st European Solid-State Device Research Conference, Proceedings

PB - Editions Frontieres

T2 - 51st IEEE European Solid-State Device Research Conference, ESSDERC 2021

Y2 - 6 September 2021 through 9 September 2021

ER -

7.5nJ/inference CMOS Echo State Network for Coronary Heart Disease prediction

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this