Improving ion-sensitive field-effect transistor selectivity with backpropagation neural network

Wan Fazlida Hanim Abdullah, Masuri Othman, Mohd Alaudin Mohd Ali, Md. Shabiul Islam

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

The ion-sensitive field-effect transistor (ISFET) that is designed to detect a specific ionic activity is susceptible to interfering ions in mixed-ion environments causing the sensor to produce deceptive signals. The objective of this work is to improve the interpretation of ISFET signals in mixed-ion environments. The focus of the research is relating sensor signal to the targeted ion concentration by applying supervised neural network as post-processing stage as a method to overcome low selectivity issues. In this paper, we acquire ISFET voltage response data in potassium and ammonium mixed-ion solutions for the training of a multilayer perceptron with backpropagation algorithm. A constant-voltage constant-current readout interface circuit is applied to maintain constant bias of the sensor throughout the data collection process. Primary data from measured observations was fed to a feed-forward multilayer perceptron trained to classify levels of ionic concentrations in various levels of mixed-ion solutions. Accuracy of sensor response interpretation of ionic activity estimation is compared between with and without neural network post-processing stage. Neural network performance was also compared for voltage values with and without pre-processing voltage signals by referencing sensor response in deionized water. Further improvement of the network was approached by using an ensemble of similar structures of networks trained with backpropagation constructed using the bagging algorithm. Results show that neural network fed with dc voltage response from 4-sensor array is able to improve concentration estimation by 15% improvement compared to direct estimation based on a look-up table. Pre-processing the sensor response significantly improves the sensor signal repeatability correlation factor by 15.5% and reduces mean-square error by 98.3%, with a typical 20% improvement in output-target regression factor network performance. Averaging from ensemble system is shown to give a further 5% improvement on the output-target regression factor with consistently stable ion concentration estimations.

Original languageEnglish
Pages (from-to)700-712
Number of pages13
JournalWSEAS Transactions on Circuits and Systems
Volume9
Issue number11
Publication statusPublished - Nov 2010

Fingerprint

Ion sensitive field effect transistors
Backpropagation
Neural networks
Sensors
Ions
Electric potential
Multilayer neural networks
Processing
Network performance
Backpropagation algorithms
Deionized water
Sensor arrays
Mean square error
Potassium
Networks (circuits)

Keywords

  • Electrochemical devices
  • Microsensors
  • MOSFET
  • Selectivity
  • Sensor array
  • Supervised learning

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Improving ion-sensitive field-effect transistor selectivity with backpropagation neural network. / Abdullah, Wan Fazlida Hanim; Othman, Masuri; Ali, Mohd Alaudin Mohd; Islam, Md. Shabiul.

In: WSEAS Transactions on Circuits and Systems, Vol. 9, No. 11, 11.2010, p. 700-712.

Research output: Contribution to journalArticle

Abdullah, Wan Fazlida Hanim ; Othman, Masuri ; Ali, Mohd Alaudin Mohd ; Islam, Md. Shabiul. / Improving ion-sensitive field-effect transistor selectivity with backpropagation neural network. In: WSEAS Transactions on Circuits and Systems. 2010 ; Vol. 9, No. 11. pp. 700-712.
@article{39fc622dac764e27b4ff6f5af799e38c,
title = "Improving ion-sensitive field-effect transistor selectivity with backpropagation neural network",
abstract = "The ion-sensitive field-effect transistor (ISFET) that is designed to detect a specific ionic activity is susceptible to interfering ions in mixed-ion environments causing the sensor to produce deceptive signals. The objective of this work is to improve the interpretation of ISFET signals in mixed-ion environments. The focus of the research is relating sensor signal to the targeted ion concentration by applying supervised neural network as post-processing stage as a method to overcome low selectivity issues. In this paper, we acquire ISFET voltage response data in potassium and ammonium mixed-ion solutions for the training of a multilayer perceptron with backpropagation algorithm. A constant-voltage constant-current readout interface circuit is applied to maintain constant bias of the sensor throughout the data collection process. Primary data from measured observations was fed to a feed-forward multilayer perceptron trained to classify levels of ionic concentrations in various levels of mixed-ion solutions. Accuracy of sensor response interpretation of ionic activity estimation is compared between with and without neural network post-processing stage. Neural network performance was also compared for voltage values with and without pre-processing voltage signals by referencing sensor response in deionized water. Further improvement of the network was approached by using an ensemble of similar structures of networks trained with backpropagation constructed using the bagging algorithm. Results show that neural network fed with dc voltage response from 4-sensor array is able to improve concentration estimation by 15{\%} improvement compared to direct estimation based on a look-up table. Pre-processing the sensor response significantly improves the sensor signal repeatability correlation factor by 15.5{\%} and reduces mean-square error by 98.3{\%}, with a typical 20{\%} improvement in output-target regression factor network performance. Averaging from ensemble system is shown to give a further 5{\%} improvement on the output-target regression factor with consistently stable ion concentration estimations.",
keywords = "Electrochemical devices, Microsensors, MOSFET, Selectivity, Sensor array, Supervised learning",
author = "Abdullah, {Wan Fazlida Hanim} and Masuri Othman and Ali, {Mohd Alaudin Mohd} and Islam, {Md. Shabiul}",
year = "2010",
month = "11",
language = "English",
volume = "9",
pages = "700--712",
journal = "WSEAS Transactions on Circuits and Systems",
issn = "1109-2734",
publisher = "World Scientific and Engineering Academy and Society",
number = "11",

}

TY - JOUR

T1 - Improving ion-sensitive field-effect transistor selectivity with backpropagation neural network

AU - Abdullah, Wan Fazlida Hanim

AU - Othman, Masuri

AU - Ali, Mohd Alaudin Mohd

AU - Islam, Md. Shabiul

PY - 2010/11

Y1 - 2010/11

N2 - The ion-sensitive field-effect transistor (ISFET) that is designed to detect a specific ionic activity is susceptible to interfering ions in mixed-ion environments causing the sensor to produce deceptive signals. The objective of this work is to improve the interpretation of ISFET signals in mixed-ion environments. The focus of the research is relating sensor signal to the targeted ion concentration by applying supervised neural network as post-processing stage as a method to overcome low selectivity issues. In this paper, we acquire ISFET voltage response data in potassium and ammonium mixed-ion solutions for the training of a multilayer perceptron with backpropagation algorithm. A constant-voltage constant-current readout interface circuit is applied to maintain constant bias of the sensor throughout the data collection process. Primary data from measured observations was fed to a feed-forward multilayer perceptron trained to classify levels of ionic concentrations in various levels of mixed-ion solutions. Accuracy of sensor response interpretation of ionic activity estimation is compared between with and without neural network post-processing stage. Neural network performance was also compared for voltage values with and without pre-processing voltage signals by referencing sensor response in deionized water. Further improvement of the network was approached by using an ensemble of similar structures of networks trained with backpropagation constructed using the bagging algorithm. Results show that neural network fed with dc voltage response from 4-sensor array is able to improve concentration estimation by 15% improvement compared to direct estimation based on a look-up table. Pre-processing the sensor response significantly improves the sensor signal repeatability correlation factor by 15.5% and reduces mean-square error by 98.3%, with a typical 20% improvement in output-target regression factor network performance. Averaging from ensemble system is shown to give a further 5% improvement on the output-target regression factor with consistently stable ion concentration estimations.

AB - The ion-sensitive field-effect transistor (ISFET) that is designed to detect a specific ionic activity is susceptible to interfering ions in mixed-ion environments causing the sensor to produce deceptive signals. The objective of this work is to improve the interpretation of ISFET signals in mixed-ion environments. The focus of the research is relating sensor signal to the targeted ion concentration by applying supervised neural network as post-processing stage as a method to overcome low selectivity issues. In this paper, we acquire ISFET voltage response data in potassium and ammonium mixed-ion solutions for the training of a multilayer perceptron with backpropagation algorithm. A constant-voltage constant-current readout interface circuit is applied to maintain constant bias of the sensor throughout the data collection process. Primary data from measured observations was fed to a feed-forward multilayer perceptron trained to classify levels of ionic concentrations in various levels of mixed-ion solutions. Accuracy of sensor response interpretation of ionic activity estimation is compared between with and without neural network post-processing stage. Neural network performance was also compared for voltage values with and without pre-processing voltage signals by referencing sensor response in deionized water. Further improvement of the network was approached by using an ensemble of similar structures of networks trained with backpropagation constructed using the bagging algorithm. Results show that neural network fed with dc voltage response from 4-sensor array is able to improve concentration estimation by 15% improvement compared to direct estimation based on a look-up table. Pre-processing the sensor response significantly improves the sensor signal repeatability correlation factor by 15.5% and reduces mean-square error by 98.3%, with a typical 20% improvement in output-target regression factor network performance. Averaging from ensemble system is shown to give a further 5% improvement on the output-target regression factor with consistently stable ion concentration estimations.

KW - Electrochemical devices

KW - Microsensors

KW - MOSFET

KW - Selectivity

KW - Sensor array

KW - Supervised learning

UR - http://www.scopus.com/inward/record.url?scp=79954528288&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79954528288&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:79954528288

VL - 9

SP - 700

EP - 712

JO - WSEAS Transactions on Circuits and Systems

JF - WSEAS Transactions on Circuits and Systems

SN - 1109-2734

IS - 11

ER -