Detecting dynamical changes in vital signs using switching Kalman filter

Vania G. Almeida, Ian T. Nabney

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Vital signs contain valuable information about the health condition of patients during their stay in the ward, when deterioration process begins. The use of methods to predict and detect regime changes such as switching models can help to understand how vital sign dynamics are altered in health and disease. However, time series of vital signs are remarkably non-stationary in these scenarios. The objective of this study is to quantify the potential bias of the switching models in the presence of non-stationary time series, when the inputs are spectral, symbolic and entropy indices. To distinguish stationary periods from non-stationary, a stationarity test was used to verify the stability of the mean and variance over short periods. Then, we compared the results from a switching Kalman filter (SKF) model trained using only indices obtained over stationary periods, with a model trained using indices obtained solely over non-stationary periods. It was observed that the indices measured over stationary and non-stationary periods were significantly different. The results were highly dependent of what indices were used as input, being the multiscale entropy (MSE) the most efficient approach, achieving an average correlation coefficients of 38%.
Original languageEnglish
Title of host publication2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017
PublisherIEEE
Number of pages4
Publication statusPublished - 31 Aug 2017
Event39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society: in conjunction with International Biomedical Engineering Conference of KOSOMBE - Jeju Island, Korea, Republic of
Duration: 11 Jul 201715 Jul 2017

Conference

Conference39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society
Abbreviated titleEMBC'17
CountryKorea, Republic of
CityJeju Island
Period11/07/1715/07/17

Fingerprint

Kalman filter
entropy
time series
health and disease
index

Bibliographical note

© 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Funding: Marie Curie Actions under the project IPSIBiM (Grant number 656737 - H2020-MSCA-IF-2014

Cite this

Almeida, V. G., & Nabney, I. T. (2017). Detecting dynamical changes in vital signs using switching Kalman filter. In 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017 IEEE.
Almeida, Vania G. ; Nabney, Ian T. / Detecting dynamical changes in vital signs using switching Kalman filter. 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017. IEEE, 2017.
@inproceedings{d8d9df7e24d5478eaf1ca3a38549c8ed,
title = "Detecting dynamical changes in vital signs using switching Kalman filter",
abstract = "Vital signs contain valuable information about the health condition of patients during their stay in the ward, when deterioration process begins. The use of methods to predict and detect regime changes such as switching models can help to understand how vital sign dynamics are altered in health and disease. However, time series of vital signs are remarkably non-stationary in these scenarios. The objective of this study is to quantify the potential bias of the switching models in the presence of non-stationary time series, when the inputs are spectral, symbolic and entropy indices. To distinguish stationary periods from non-stationary, a stationarity test was used to verify the stability of the mean and variance over short periods. Then, we compared the results from a switching Kalman filter (SKF) model trained using only indices obtained over stationary periods, with a model trained using indices obtained solely over non-stationary periods. It was observed that the indices measured over stationary and non-stationary periods were significantly different. The results were highly dependent of what indices were used as input, being the multiscale entropy (MSE) the most efficient approach, achieving an average correlation coefficients of 38{\%}.",
author = "Almeida, {Vania G.} and Nabney, {Ian T.}",
note = "{\circledC} 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Funding: Marie Curie Actions under the project IPSIBiM (Grant number 656737 - H2020-MSCA-IF-2014",
year = "2017",
month = "8",
day = "31",
language = "English",
booktitle = "2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017",
publisher = "IEEE",
address = "United States",

}

Almeida, VG & Nabney, IT 2017, Detecting dynamical changes in vital signs using switching Kalman filter. in 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017. IEEE, 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Jeju Island, Korea, Republic of, 11/07/17.

Detecting dynamical changes in vital signs using switching Kalman filter. / Almeida, Vania G.; Nabney, Ian T.

2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017. IEEE, 2017.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Detecting dynamical changes in vital signs using switching Kalman filter

AU - Almeida, Vania G.

AU - Nabney, Ian T.

N1 - © 2017 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. Funding: Marie Curie Actions under the project IPSIBiM (Grant number 656737 - H2020-MSCA-IF-2014

PY - 2017/8/31

Y1 - 2017/8/31

N2 - Vital signs contain valuable information about the health condition of patients during their stay in the ward, when deterioration process begins. The use of methods to predict and detect regime changes such as switching models can help to understand how vital sign dynamics are altered in health and disease. However, time series of vital signs are remarkably non-stationary in these scenarios. The objective of this study is to quantify the potential bias of the switching models in the presence of non-stationary time series, when the inputs are spectral, symbolic and entropy indices. To distinguish stationary periods from non-stationary, a stationarity test was used to verify the stability of the mean and variance over short periods. Then, we compared the results from a switching Kalman filter (SKF) model trained using only indices obtained over stationary periods, with a model trained using indices obtained solely over non-stationary periods. It was observed that the indices measured over stationary and non-stationary periods were significantly different. The results were highly dependent of what indices were used as input, being the multiscale entropy (MSE) the most efficient approach, achieving an average correlation coefficients of 38%.

AB - Vital signs contain valuable information about the health condition of patients during their stay in the ward, when deterioration process begins. The use of methods to predict and detect regime changes such as switching models can help to understand how vital sign dynamics are altered in health and disease. However, time series of vital signs are remarkably non-stationary in these scenarios. The objective of this study is to quantify the potential bias of the switching models in the presence of non-stationary time series, when the inputs are spectral, symbolic and entropy indices. To distinguish stationary periods from non-stationary, a stationarity test was used to verify the stability of the mean and variance over short periods. Then, we compared the results from a switching Kalman filter (SKF) model trained using only indices obtained over stationary periods, with a model trained using indices obtained solely over non-stationary periods. It was observed that the indices measured over stationary and non-stationary periods were significantly different. The results were highly dependent of what indices were used as input, being the multiscale entropy (MSE) the most efficient approach, achieving an average correlation coefficients of 38%.

UR - http://doi.org/10.17036/researchdata.aston.ac.uk.00000244

M3 - Conference contribution

BT - 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017

PB - IEEE

ER -

Almeida VG, Nabney IT. Detecting dynamical changes in vital signs using switching Kalman filter. In 2017 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2017. IEEE. 2017