Comparison of artificial neural network (ANN) and partial least squares (PLS) regression models for predicting respiratory ventilation: An exploratory study

Ming I.Brandon Lin, William A. Groves, Andris Freivalds, Eun Gyung Lee, Martin Harper

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

The objective of this study was to assess the potential for using artificial neural networks (ANN) to predict inspired minute ventilation (V I) during exercise activities. Six physiological/kinematic measurements obtained from a portable ambulatory monitoring system, along with individual's anthropometric and demographic characteristics, were employed as input variables to develop and optimize the ANN configuration with respect to reference values simultaneously measured using a pneumotachograph (PT). The generalization ability of the resulting two-hidden-layer ANN model was compared with a linear predictive model developed through partial least squares (PLS) regression, as well as other VIpredictive models proposed in the literature. Using an independent dataset recorded from nine 80-min step tests, the results showed that the ANN-estimated VIwas highly correlated (R2= 0.88) with VI measured by the PT, with a mean difference of approximately 0.9%. In contrast, the PLS and other regression-based models resulted in larger average errors ranging from 7 to 34%. In addition, the ANN model yielded estimates of cumulative total volume that were on average within 1% of reference PT measurements. Compared with established statistical methods, the proposed ANN model demonstrates the potential to provide improved prediction of respiratory ventilation in workplace applications for which the use of traditional laboratorybased instruments is not feasible. Further research should be conducted to investigate the performance of ANNs for different types of physical activity in larger and more varied worker populations.

Original languageEnglish (US)
Pages (from-to)1603-1611
Number of pages9
JournalEuropean Journal of Applied Physiology
Volume112
Issue number5
DOIs
StatePublished - May 1 2012

Fingerprint

Neural Networks (Computer)
Least-Squares Analysis
Ventilation
Ambulatory Monitoring
Exercise Test
Biomechanical Phenomena
Workplace
Linear Models
Reference Values
Demography
Research
Population

All Science Journal Classification (ASJC) codes

  • Orthopedics and Sports Medicine
  • Public Health, Environmental and Occupational Health
  • Physiology (medical)

Cite this

@article{65b3044aecc64c20b363303106ca81fd,
title = "Comparison of artificial neural network (ANN) and partial least squares (PLS) regression models for predicting respiratory ventilation: An exploratory study",
abstract = "The objective of this study was to assess the potential for using artificial neural networks (ANN) to predict inspired minute ventilation (V I) during exercise activities. Six physiological/kinematic measurements obtained from a portable ambulatory monitoring system, along with individual's anthropometric and demographic characteristics, were employed as input variables to develop and optimize the ANN configuration with respect to reference values simultaneously measured using a pneumotachograph (PT). The generalization ability of the resulting two-hidden-layer ANN model was compared with a linear predictive model developed through partial least squares (PLS) regression, as well as other VIpredictive models proposed in the literature. Using an independent dataset recorded from nine 80-min step tests, the results showed that the ANN-estimated VIwas highly correlated (R2= 0.88) with VI measured by the PT, with a mean difference of approximately 0.9{\%}. In contrast, the PLS and other regression-based models resulted in larger average errors ranging from 7 to 34{\%}. In addition, the ANN model yielded estimates of cumulative total volume that were on average within 1{\%} of reference PT measurements. Compared with established statistical methods, the proposed ANN model demonstrates the potential to provide improved prediction of respiratory ventilation in workplace applications for which the use of traditional laboratorybased instruments is not feasible. Further research should be conducted to investigate the performance of ANNs for different types of physical activity in larger and more varied worker populations.",
author = "Lin, {Ming I.Brandon} and Groves, {William A.} and Andris Freivalds and Lee, {Eun Gyung} and Martin Harper",
year = "2012",
month = "5",
day = "1",
doi = "10.1007/s00421-011-2118-6",
language = "English (US)",
volume = "112",
pages = "1603--1611",
journal = "European Journal of Applied Physiology",
issn = "1439-6319",
publisher = "Springer Verlag",
number = "5",

}

TY - JOUR

T1 - Comparison of artificial neural network (ANN) and partial least squares (PLS) regression models for predicting respiratory ventilation

T2 - An exploratory study

AU - Lin, Ming I.Brandon

AU - Groves, William A.

AU - Freivalds, Andris

AU - Lee, Eun Gyung

AU - Harper, Martin

PY - 2012/5/1

Y1 - 2012/5/1

N2 - The objective of this study was to assess the potential for using artificial neural networks (ANN) to predict inspired minute ventilation (V I) during exercise activities. Six physiological/kinematic measurements obtained from a portable ambulatory monitoring system, along with individual's anthropometric and demographic characteristics, were employed as input variables to develop and optimize the ANN configuration with respect to reference values simultaneously measured using a pneumotachograph (PT). The generalization ability of the resulting two-hidden-layer ANN model was compared with a linear predictive model developed through partial least squares (PLS) regression, as well as other VIpredictive models proposed in the literature. Using an independent dataset recorded from nine 80-min step tests, the results showed that the ANN-estimated VIwas highly correlated (R2= 0.88) with VI measured by the PT, with a mean difference of approximately 0.9%. In contrast, the PLS and other regression-based models resulted in larger average errors ranging from 7 to 34%. In addition, the ANN model yielded estimates of cumulative total volume that were on average within 1% of reference PT measurements. Compared with established statistical methods, the proposed ANN model demonstrates the potential to provide improved prediction of respiratory ventilation in workplace applications for which the use of traditional laboratorybased instruments is not feasible. Further research should be conducted to investigate the performance of ANNs for different types of physical activity in larger and more varied worker populations.

AB - The objective of this study was to assess the potential for using artificial neural networks (ANN) to predict inspired minute ventilation (V I) during exercise activities. Six physiological/kinematic measurements obtained from a portable ambulatory monitoring system, along with individual's anthropometric and demographic characteristics, were employed as input variables to develop and optimize the ANN configuration with respect to reference values simultaneously measured using a pneumotachograph (PT). The generalization ability of the resulting two-hidden-layer ANN model was compared with a linear predictive model developed through partial least squares (PLS) regression, as well as other VIpredictive models proposed in the literature. Using an independent dataset recorded from nine 80-min step tests, the results showed that the ANN-estimated VIwas highly correlated (R2= 0.88) with VI measured by the PT, with a mean difference of approximately 0.9%. In contrast, the PLS and other regression-based models resulted in larger average errors ranging from 7 to 34%. In addition, the ANN model yielded estimates of cumulative total volume that were on average within 1% of reference PT measurements. Compared with established statistical methods, the proposed ANN model demonstrates the potential to provide improved prediction of respiratory ventilation in workplace applications for which the use of traditional laboratorybased instruments is not feasible. Further research should be conducted to investigate the performance of ANNs for different types of physical activity in larger and more varied worker populations.

UR - http://www.scopus.com/inward/record.url?scp=84861527031&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84861527031&partnerID=8YFLogxK

U2 - 10.1007/s00421-011-2118-6

DO - 10.1007/s00421-011-2118-6

M3 - Article

C2 - 21861111

AN - SCOPUS:84861527031

VL - 112

SP - 1603

EP - 1611

JO - European Journal of Applied Physiology

JF - European Journal of Applied Physiology

SN - 1439-6319

IS - 5

ER -