Personalized advice on chronic low back pain interventions: a machine learning approach

Research output: Contribution to conferencePaperAcademicpeer-review

Abstract

Clinical decision support systems (CDSS) can support clinicians in selecting appropriate treatments for patients. The objective of this study was to examine if triaging patients with LBP to the most optimal treatment can be improved by using a data-driven approach with the help of machine learning technologies as base of such a CDSS.
A clinical database of the Groningen Spine Center containing patient-reported data from 1546 patients with LBP was used. From this dataset, a training dataset with 354 features was labeled on eight different treatments actually received by these patients. With this dataset, models were trained. A test dataset with 50 cases judged on treatments by 4 experts in LBP triage to treatments was used to test these models with data not used to train the models. Prediction accuracy and average area under curve (AUC) were used as performance measures for the models.
The AUC values indicated small to medium learning effects showing that machine learning on patient-reported data, to model decision-making processes on treatments for LBP, may be possible. One of the best performing models was the Bayesian Network (BN) model; e.g. predicted surgery with accuracy 0.78 (95% C.I. 0.68– 0.87) and AUC 0.70. Benefits to using BNs compared to other supervised machine learning techniques are that it is easy to exploit expert knowledge in BN models, meaning that advices generated by the model can be explained. The next step is to improve the BN so that it can be used in a CDSS.
Original languageEnglish
Pages25
Number of pages1
Publication statusPublished - 15 Nov 2018
EventSBPR 2018: Understanding the mechanisms of back pain: work, rest and play - University Medical Centre , Groningen, Netherlands
Duration: 15 Nov 201816 Nov 2018

Conference

ConferenceSBPR 2018
Abbreviated titleSBPR
CountryNetherlands
CityGroningen
Period15/11/1816/11/18

Fingerprint

Learning systems
Bayesian networks
Decision support systems
Surgery
Decision making

Cite this

@conference{7e41373cca0141aa9b81761c43bc5de9,
title = "Personalized advice on chronic low back pain interventions: a machine learning approach",
abstract = "Clinical decision support systems (CDSS) can support clinicians in selecting appropriate treatments for patients. The objective of this study was to examine if triaging patients with LBP to the most optimal treatment can be improved by using a data-driven approach with the help of machine learning technologies as base of such a CDSS.A clinical database of the Groningen Spine Center containing patient-reported data from 1546 patients with LBP was used. From this dataset, a training dataset with 354 features was labeled on eight different treatments actually received by these patients. With this dataset, models were trained. A test dataset with 50 cases judged on treatments by 4 experts in LBP triage to treatments was used to test these models with data not used to train the models. Prediction accuracy and average area under curve (AUC) were used as performance measures for the models.The AUC values indicated small to medium learning effects showing that machine learning on patient-reported data, to model decision-making processes on treatments for LBP, may be possible. One of the best performing models was the Bayesian Network (BN) model; e.g. predicted surgery with accuracy 0.78 (95{\%} C.I. 0.68– 0.87) and AUC 0.70. Benefits to using BNs compared to other supervised machine learning techniques are that it is easy to exploit expert knowledge in BN models, meaning that advices generated by the model can be explained. The next step is to improve the BN so that it can be used in a CDSS.",
author = "{Oude Nijeweme - d'Hollosy}, Wendy and Mannes Poel and {van Velsen}, {Lex Stefan} and C Oudshoorn and Hermens, {Hermanus J.} and P Stegeman and A.P. Wolff and M.F. Reneman and Remko Soer",
year = "2018",
month = "11",
day = "15",
language = "English",
pages = "25",
note = "SBPR 2018 : Understanding the mechanisms of back pain: work, rest and play, SBPR ; Conference date: 15-11-2018 Through 16-11-2018",

}

Oude Nijeweme - d'Hollosy, W, Poel, M, van Velsen, LS, Oudshoorn, C, Hermens, HJ, Stegeman, P, Wolff, AP, Reneman, MF & Soer, R 2018, 'Personalized advice on chronic low back pain interventions: a machine learning approach' Paper presented at SBPR 2018, Groningen, Netherlands, 15/11/18 - 16/11/18, pp. 25.

Personalized advice on chronic low back pain interventions: a machine learning approach. / Oude Nijeweme - d'Hollosy, Wendy; Poel, Mannes ; van Velsen, Lex Stefan; Oudshoorn, C; Hermens, Hermanus J.; Stegeman, P; Wolff, A.P.; Reneman, M.F.; Soer, Remko.

2018. 25 Paper presented at SBPR 2018, Groningen, Netherlands.

Research output: Contribution to conferencePaperAcademicpeer-review

TY - CONF

T1 - Personalized advice on chronic low back pain interventions: a machine learning approach

AU - Oude Nijeweme - d'Hollosy, Wendy

AU - Poel, Mannes

AU - van Velsen, Lex Stefan

AU - Oudshoorn, C

AU - Hermens, Hermanus J.

AU - Stegeman, P

AU - Wolff, A.P.

AU - Reneman, M.F.

AU - Soer, Remko

PY - 2018/11/15

Y1 - 2018/11/15

N2 - Clinical decision support systems (CDSS) can support clinicians in selecting appropriate treatments for patients. The objective of this study was to examine if triaging patients with LBP to the most optimal treatment can be improved by using a data-driven approach with the help of machine learning technologies as base of such a CDSS.A clinical database of the Groningen Spine Center containing patient-reported data from 1546 patients with LBP was used. From this dataset, a training dataset with 354 features was labeled on eight different treatments actually received by these patients. With this dataset, models were trained. A test dataset with 50 cases judged on treatments by 4 experts in LBP triage to treatments was used to test these models with data not used to train the models. Prediction accuracy and average area under curve (AUC) were used as performance measures for the models.The AUC values indicated small to medium learning effects showing that machine learning on patient-reported data, to model decision-making processes on treatments for LBP, may be possible. One of the best performing models was the Bayesian Network (BN) model; e.g. predicted surgery with accuracy 0.78 (95% C.I. 0.68– 0.87) and AUC 0.70. Benefits to using BNs compared to other supervised machine learning techniques are that it is easy to exploit expert knowledge in BN models, meaning that advices generated by the model can be explained. The next step is to improve the BN so that it can be used in a CDSS.

AB - Clinical decision support systems (CDSS) can support clinicians in selecting appropriate treatments for patients. The objective of this study was to examine if triaging patients with LBP to the most optimal treatment can be improved by using a data-driven approach with the help of machine learning technologies as base of such a CDSS.A clinical database of the Groningen Spine Center containing patient-reported data from 1546 patients with LBP was used. From this dataset, a training dataset with 354 features was labeled on eight different treatments actually received by these patients. With this dataset, models were trained. A test dataset with 50 cases judged on treatments by 4 experts in LBP triage to treatments was used to test these models with data not used to train the models. Prediction accuracy and average area under curve (AUC) were used as performance measures for the models.The AUC values indicated small to medium learning effects showing that machine learning on patient-reported data, to model decision-making processes on treatments for LBP, may be possible. One of the best performing models was the Bayesian Network (BN) model; e.g. predicted surgery with accuracy 0.78 (95% C.I. 0.68– 0.87) and AUC 0.70. Benefits to using BNs compared to other supervised machine learning techniques are that it is easy to exploit expert knowledge in BN models, meaning that advices generated by the model can be explained. The next step is to improve the BN so that it can be used in a CDSS.

UR - https://wencke4.housing.rug.nl/Cursuswinkel/public/Brochure/SBPR_program%20with%20abstracts.pdf

M3 - Paper

SP - 25

ER -