A MEMS network operating as a reservoir computer to classify peripheral artery disease signal from a healthy signal

Mohammad Okour; Hamed Nikfarjam; Mutaz Al Fayad; Farahnaz Fallah Tafti; Mohammad Ali Takallou; Iraklis I. Pipinos; Siavash Pourkamali; Sara A Myers; Fadi Alsaleem

doi:10.1016/j.sna.2025.116318

A MEMS network operating as a reservoir computer to classify peripheral artery disease signal from a healthy signal

Mohammad Okour, Hamed Nikfarjam, Mutaz Al Fayad, Farahnaz Fallah Tafti, Mohammad Ali Takallou, Iraklis I. Pipinos, Siavash Pourkamali, Sara A Myers, Fadi Alsaleem

Research output: Contribution to journal › Article › peer-review

Abstract

Microelectromechanical Systems (MEMS) reservoir computers (RC) utilize the vibration of micromechanical structures to perform computing rather than relying on digital computers, enabling low-power computing solutions. This paper introduces a novel application of MEMS RC for classifying the signals of patients with peripheral artery disease (PAD) versus healthy individuals. Using acceleration signals from PAD and non-PAD subjects, this study implements a MEMS RC model with three interconnected MEMS devices, achieving 89 % classification accuracy—substantially outperforming the 54 % accuracy obtained with a single MEMS device. This method reintroduces the benefits of parallel computing to RC, unlike the virtual-node RC approach, which uses one physical system in a serial setup, potentially introducing delays and requiring complex circuitry for high-speed sampling.

Original language	English (US)
Article number	116318
Journal	Sensors and Actuators A: Physical
Volume	385
DOIs	https://doi.org/10.1016/j.sna.2025.116318
State	Published - Apr 16 2025

Keywords

MEMS
PAD
Peripheral artery disease
Reservoir computers

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Instrumentation
Condensed Matter Physics
Surfaces, Coatings and Films
Metals and Alloys
Electrical and Electronic Engineering

Access to Document

10.1016/j.sna.2025.116318

Cite this

@article{a18d87329b024ae78f07dafb1689fd70,

title = "A MEMS network operating as a reservoir computer to classify peripheral artery disease signal from a healthy signal",

abstract = "Microelectromechanical Systems (MEMS) reservoir computers (RC) utilize the vibration of micromechanical structures to perform computing rather than relying on digital computers, enabling low-power computing solutions. This paper introduces a novel application of MEMS RC for classifying the signals of patients with peripheral artery disease (PAD) versus healthy individuals. Using acceleration signals from PAD and non-PAD subjects, this study implements a MEMS RC model with three interconnected MEMS devices, achieving 89 % classification accuracy—substantially outperforming the 54 % accuracy obtained with a single MEMS device. This method reintroduces the benefits of parallel computing to RC, unlike the virtual-node RC approach, which uses one physical system in a serial setup, potentially introducing delays and requiring complex circuitry for high-speed sampling.",

keywords = "MEMS, PAD, Peripheral artery disease, Reservoir computers",

author = "Mohammad Okour and Hamed Nikfarjam and Fayad, {Mutaz Al} and Tafti, {Farahnaz Fallah} and Takallou, {Mohammad Ali} and Pipinos, {Iraklis I.} and Siavash Pourkamali and Myers, {Sara A} and Fadi Alsaleem",

note = "Publisher Copyright: {\textcopyright} 2025 Elsevier B.V.",

year = "2025",

month = apr,

day = "16",

doi = "10.1016/j.sna.2025.116318",

language = "English (US)",

volume = "385",

journal = "Sensors and Actuators A: Physical",

issn = "0924-4247",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - A MEMS network operating as a reservoir computer to classify peripheral artery disease signal from a healthy signal

AU - Okour, Mohammad

AU - Nikfarjam, Hamed

AU - Fayad, Mutaz Al

AU - Tafti, Farahnaz Fallah

AU - Takallou, Mohammad Ali

AU - Pipinos, Iraklis I.

AU - Pourkamali, Siavash

AU - Myers, Sara A

AU - Alsaleem, Fadi

PY - 2025/4/16

Y1 - 2025/4/16

N2 - Microelectromechanical Systems (MEMS) reservoir computers (RC) utilize the vibration of micromechanical structures to perform computing rather than relying on digital computers, enabling low-power computing solutions. This paper introduces a novel application of MEMS RC for classifying the signals of patients with peripheral artery disease (PAD) versus healthy individuals. Using acceleration signals from PAD and non-PAD subjects, this study implements a MEMS RC model with three interconnected MEMS devices, achieving 89 % classification accuracy—substantially outperforming the 54 % accuracy obtained with a single MEMS device. This method reintroduces the benefits of parallel computing to RC, unlike the virtual-node RC approach, which uses one physical system in a serial setup, potentially introducing delays and requiring complex circuitry for high-speed sampling.

AB - Microelectromechanical Systems (MEMS) reservoir computers (RC) utilize the vibration of micromechanical structures to perform computing rather than relying on digital computers, enabling low-power computing solutions. This paper introduces a novel application of MEMS RC for classifying the signals of patients with peripheral artery disease (PAD) versus healthy individuals. Using acceleration signals from PAD and non-PAD subjects, this study implements a MEMS RC model with three interconnected MEMS devices, achieving 89 % classification accuracy—substantially outperforming the 54 % accuracy obtained with a single MEMS device. This method reintroduces the benefits of parallel computing to RC, unlike the virtual-node RC approach, which uses one physical system in a serial setup, potentially introducing delays and requiring complex circuitry for high-speed sampling.

KW - MEMS

KW - PAD

KW - Peripheral artery disease

KW - Reservoir computers

UR - http://www.scopus.com/inward/record.url?scp=85217819614&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85217819614&partnerID=8YFLogxK

U2 - 10.1016/j.sna.2025.116318

DO - 10.1016/j.sna.2025.116318

M3 - Article

AN - SCOPUS:85217819614

SN - 0924-4247

VL - 385

JO - Sensors and Actuators A: Physical

JF - Sensors and Actuators A: Physical

M1 - 116318

ER -

A MEMS network operating as a reservoir computer to classify peripheral artery disease signal from a healthy signal

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this