Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer

Daniel Rodríguez-Martín, Albert Samà, Carlos Pérez-López, Andreu Català, Joan M.Moreno Arostegui, Joan Cabestany, Àngels Bayés, Sheila Alcaine, Berta Mestre, Anna Prats, M. Cruz Crespo, Timothy J. Counihan, Patrick Browne, Leo R. Quinlan, Gearóid Laighin, Dean Sweeney, Hadas Lewy, Joseph Azuri, Gabriel Vainstein, Roberta Annicchiarico & 2 others Alberto Costa, Alejandro Rodríguez-Molinero

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Among Parkinson's disease (PD) symptoms, freezing of gait (FoG) is one of the most debilitating. To assess FoG, current clinical practice mostly employs repeated evaluations over weeks and months based on questionnaires, which may not accurately map the severity of this symptom. The use of a non-invasive system to monitor the activities of daily living (ADL) and the PD symptoms experienced by patients throughout the day could provide a more accurate and objective evaluation of FoG in order to better understand the evolution of the disease and allow for a more informed decision-making process in making adjustments to the patient's treatment plan. This paper presents a new algorithm to detect FoG with a machine learning approach based on Support Vector Machines (SVM) and a single tri-axial accelerometer worn at the waist. The method is evaluated through the acceleration signals in an outpatient setting gathered from 21 PD patients at their home and evaluated under two different conditions: first, a generic model is tested by using a leave-one-out approach and, second, a personalised model that also uses part of the dataset from each patient. Results show a significant improvement in the accuracy of the personalised model compared to the generic model, showing enhancement in the specificity and sensitivity geometric mean (GM) of 7.2%. Furthermore, the SVM approach adopted has been compared to the most comprehensive FoG detection method currently in use (referred to as MBFA in this paper). Results of our novel generic method provide an enhancement of 11.2% in the GM compared to the MBFA generic model and, in the case of the personalised model, a 10% of improvement with respect to the MBFA personalised model. Thus, our results show that a machine learning approach can be used to monitor FoG during the daily life of PD patients and, furthermore, personalised models for FoG detection can be used to improve monitoring accuracy.

Original languageEnglish
Article numbere0171764
JournalPLoS One
Volume12
Issue number2
DOIs
Publication statusPublished - Feb 1 2017

Fingerprint

waist
gait
Gait
Accelerometers
Freezing
Support vector machines
freezing
Parkinson disease
Parkinson Disease
signs and symptoms (animals and humans)
artificial intelligence
Learning systems
monitoring
Social Adjustment
Patient treatment
Activities of Daily Living
support vector machines
Support Vector Machine
Decision Making
Outpatients

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Agricultural and Biological Sciences(all)

Cite this

Rodríguez-Martín, D., Samà, A., Pérez-López, C., Català, A., Arostegui, J. M. M., Cabestany, J., ... Rodríguez-Molinero, A. (2017). Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer. PLoS One, 12(2), [e0171764]. https://doi.org/10.1371/journal.pone.0171764

Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer. / Rodríguez-Martín, Daniel; Samà, Albert; Pérez-López, Carlos; Català, Andreu; Arostegui, Joan M.Moreno; Cabestany, Joan; Bayés, Àngels; Alcaine, Sheila; Mestre, Berta; Prats, Anna; Crespo, M. Cruz; Counihan, Timothy J.; Browne, Patrick; Quinlan, Leo R.; Laighin, Gearóid; Sweeney, Dean; Lewy, Hadas; Azuri, Joseph; Vainstein, Gabriel; Annicchiarico, Roberta; Costa, Alberto; Rodríguez-Molinero, Alejandro.

In: PLoS One, Vol. 12, No. 2, e0171764, 01.02.2017.

Research output: Contribution to journalArticle

Rodríguez-Martín, D, Samà, A, Pérez-López, C, Català, A, Arostegui, JMM, Cabestany, J, Bayés, À, Alcaine, S, Mestre, B, Prats, A, Crespo, MC, Counihan, TJ, Browne, P, Quinlan, LR, Laighin, G, Sweeney, D, Lewy, H, Azuri, J, Vainstein, G, Annicchiarico, R, Costa, A & Rodríguez-Molinero, A 2017, 'Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer', PLoS One, vol. 12, no. 2, e0171764. https://doi.org/10.1371/journal.pone.0171764
Rodríguez-Martín D, Samà A, Pérez-López C, Català A, Arostegui JMM, Cabestany J et al. Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer. PLoS One. 2017 Feb 1;12(2). e0171764. https://doi.org/10.1371/journal.pone.0171764
Rodríguez-Martín, Daniel ; Samà, Albert ; Pérez-López, Carlos ; Català, Andreu ; Arostegui, Joan M.Moreno ; Cabestany, Joan ; Bayés, Àngels ; Alcaine, Sheila ; Mestre, Berta ; Prats, Anna ; Crespo, M. Cruz ; Counihan, Timothy J. ; Browne, Patrick ; Quinlan, Leo R. ; Laighin, Gearóid ; Sweeney, Dean ; Lewy, Hadas ; Azuri, Joseph ; Vainstein, Gabriel ; Annicchiarico, Roberta ; Costa, Alberto ; Rodríguez-Molinero, Alejandro. / Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer. In: PLoS One. 2017 ; Vol. 12, No. 2.
@article{fa072afddadf488789af336d87245e96,
title = "Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer",
abstract = "Among Parkinson's disease (PD) symptoms, freezing of gait (FoG) is one of the most debilitating. To assess FoG, current clinical practice mostly employs repeated evaluations over weeks and months based on questionnaires, which may not accurately map the severity of this symptom. The use of a non-invasive system to monitor the activities of daily living (ADL) and the PD symptoms experienced by patients throughout the day could provide a more accurate and objective evaluation of FoG in order to better understand the evolution of the disease and allow for a more informed decision-making process in making adjustments to the patient's treatment plan. This paper presents a new algorithm to detect FoG with a machine learning approach based on Support Vector Machines (SVM) and a single tri-axial accelerometer worn at the waist. The method is evaluated through the acceleration signals in an outpatient setting gathered from 21 PD patients at their home and evaluated under two different conditions: first, a generic model is tested by using a leave-one-out approach and, second, a personalised model that also uses part of the dataset from each patient. Results show a significant improvement in the accuracy of the personalised model compared to the generic model, showing enhancement in the specificity and sensitivity geometric mean (GM) of 7.2{\%}. Furthermore, the SVM approach adopted has been compared to the most comprehensive FoG detection method currently in use (referred to as MBFA in this paper). Results of our novel generic method provide an enhancement of 11.2{\%} in the GM compared to the MBFA generic model and, in the case of the personalised model, a 10{\%} of improvement with respect to the MBFA personalised model. Thus, our results show that a machine learning approach can be used to monitor FoG during the daily life of PD patients and, furthermore, personalised models for FoG detection can be used to improve monitoring accuracy.",
author = "Daniel Rodr{\'i}guez-Mart{\'i}n and Albert Sam{\`a} and Carlos P{\'e}rez-L{\'o}pez and Andreu Catal{\`a} and Arostegui, {Joan M.Moreno} and Joan Cabestany and {\`A}ngels Bay{\'e}s and Sheila Alcaine and Berta Mestre and Anna Prats and Crespo, {M. Cruz} and Counihan, {Timothy J.} and Patrick Browne and Quinlan, {Leo R.} and Gear{\'o}id Laighin and Dean Sweeney and Hadas Lewy and Joseph Azuri and Gabriel Vainstein and Roberta Annicchiarico and Alberto Costa and Alejandro Rodr{\'i}guez-Molinero",
year = "2017",
month = "2",
day = "1",
doi = "10.1371/journal.pone.0171764",
language = "English",
volume = "12",
journal = "PLoS One",
issn = "1932-6203",
publisher = "Public Library of Science",
number = "2",

}

TY - JOUR

T1 - Home detection of freezing of gait using Support Vector Machines through a single waist-worn triaxial accelerometer

AU - Rodríguez-Martín, Daniel

AU - Samà, Albert

AU - Pérez-López, Carlos

AU - Català, Andreu

AU - Arostegui, Joan M.Moreno

AU - Cabestany, Joan

AU - Bayés, Àngels

AU - Alcaine, Sheila

AU - Mestre, Berta

AU - Prats, Anna

AU - Crespo, M. Cruz

AU - Counihan, Timothy J.

AU - Browne, Patrick

AU - Quinlan, Leo R.

AU - Laighin, Gearóid

AU - Sweeney, Dean

AU - Lewy, Hadas

AU - Azuri, Joseph

AU - Vainstein, Gabriel

AU - Annicchiarico, Roberta

AU - Costa, Alberto

AU - Rodríguez-Molinero, Alejandro

PY - 2017/2/1

Y1 - 2017/2/1

N2 - Among Parkinson's disease (PD) symptoms, freezing of gait (FoG) is one of the most debilitating. To assess FoG, current clinical practice mostly employs repeated evaluations over weeks and months based on questionnaires, which may not accurately map the severity of this symptom. The use of a non-invasive system to monitor the activities of daily living (ADL) and the PD symptoms experienced by patients throughout the day could provide a more accurate and objective evaluation of FoG in order to better understand the evolution of the disease and allow for a more informed decision-making process in making adjustments to the patient's treatment plan. This paper presents a new algorithm to detect FoG with a machine learning approach based on Support Vector Machines (SVM) and a single tri-axial accelerometer worn at the waist. The method is evaluated through the acceleration signals in an outpatient setting gathered from 21 PD patients at their home and evaluated under two different conditions: first, a generic model is tested by using a leave-one-out approach and, second, a personalised model that also uses part of the dataset from each patient. Results show a significant improvement in the accuracy of the personalised model compared to the generic model, showing enhancement in the specificity and sensitivity geometric mean (GM) of 7.2%. Furthermore, the SVM approach adopted has been compared to the most comprehensive FoG detection method currently in use (referred to as MBFA in this paper). Results of our novel generic method provide an enhancement of 11.2% in the GM compared to the MBFA generic model and, in the case of the personalised model, a 10% of improvement with respect to the MBFA personalised model. Thus, our results show that a machine learning approach can be used to monitor FoG during the daily life of PD patients and, furthermore, personalised models for FoG detection can be used to improve monitoring accuracy.

AB - Among Parkinson's disease (PD) symptoms, freezing of gait (FoG) is one of the most debilitating. To assess FoG, current clinical practice mostly employs repeated evaluations over weeks and months based on questionnaires, which may not accurately map the severity of this symptom. The use of a non-invasive system to monitor the activities of daily living (ADL) and the PD symptoms experienced by patients throughout the day could provide a more accurate and objective evaluation of FoG in order to better understand the evolution of the disease and allow for a more informed decision-making process in making adjustments to the patient's treatment plan. This paper presents a new algorithm to detect FoG with a machine learning approach based on Support Vector Machines (SVM) and a single tri-axial accelerometer worn at the waist. The method is evaluated through the acceleration signals in an outpatient setting gathered from 21 PD patients at their home and evaluated under two different conditions: first, a generic model is tested by using a leave-one-out approach and, second, a personalised model that also uses part of the dataset from each patient. Results show a significant improvement in the accuracy of the personalised model compared to the generic model, showing enhancement in the specificity and sensitivity geometric mean (GM) of 7.2%. Furthermore, the SVM approach adopted has been compared to the most comprehensive FoG detection method currently in use (referred to as MBFA in this paper). Results of our novel generic method provide an enhancement of 11.2% in the GM compared to the MBFA generic model and, in the case of the personalised model, a 10% of improvement with respect to the MBFA personalised model. Thus, our results show that a machine learning approach can be used to monitor FoG during the daily life of PD patients and, furthermore, personalised models for FoG detection can be used to improve monitoring accuracy.

UR - http://www.scopus.com/inward/record.url?scp=85013025234&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85013025234&partnerID=8YFLogxK

U2 - 10.1371/journal.pone.0171764

DO - 10.1371/journal.pone.0171764

M3 - Article

VL - 12

JO - PLoS One

JF - PLoS One

SN - 1932-6203

IS - 2

M1 - e0171764

ER -