Abstract
Objective. Highly comparative time series analysis (HCTSA) is a novel approach involving massive feature extraction using publicly available code from many disciplines. The Prematurity-Related Ventilatory Control (Pre-Vent) observational multicenter prospective study collected bedside monitor data from > 700 extremely preterm infants to identify physiologic features that predict respiratory outcomes. Approach. We calculated a subset of 33 HCTSA features on > 7 M 10 min windows of oxygen saturation (SPO2) and heart rate (HR) from the Pre-Vent cohort to quantify predictive performance. This subset included representatives previously identified using unsupervised clustering on > 3500 HCTSA algorithms. We hypothesized that the best HCTSA algorithms would compare favorably to optimal PreVent physiologic predictor IH90_DPE (duration per event of intermittent hypoxemia events below 90%). Main Results. The top HCTSA features were from a cluster of algorithms associated with the autocorrelation of SPO2 time series and identified low frequency patterns of desaturation as high risk. These features had comparable performance to and were highly correlated with IH90_DPE but perhaps measure the physiologic status of an infant in a more robust way that warrants further investigation. The top HR HCTSA features were symbolic transformation measures that had previously been identified as strong predictors of neonatal mortality. HR metrics were only important predictors at early days of life which was likely due to the larger proportion of infants whose outcome was death by any cause. A simple HCTSA model using 3 top features outperformed IH90_DPE at day of life 7 (.778 versus .729) but was essentially equivalent at day of life 28 (.849 versus .850). Significance. These results validated the utility of a representative HCTSA approach but also provides additional evidence supporting IH90_DPE as an optimal predictor of respiratory outcomes.
Original language | English (US) |
---|---|
Article number | 055025 |
Journal | Physiological Measurement |
Volume | 45 |
Issue number | 5 |
DOIs | |
State | Published - May 1 2024 |
Keywords
- highly comparative time series analysis
- intermittent hypoxemia
- predictive models
- preterm infants
ASJC Scopus subject areas
- Biophysics
- Physiology
- Biomedical Engineering
- Physiology (medical)
Access to Document
Other files and links
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver
Qiu, J., Di Fiore, J. M., Krishnamurthi, N., Indic, P., Carroll, J. L., Claure, N., Kemp, J. S., Dennery, P. A., Ambalavanan, N., Weese-Mayer, D. E., Maria Hibbs, A., Martin, R. J., Bancalari, E., Hamvas, A., Randall Moorman, J., & Lake, D. E. (2024). Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants. Physiological Measurement, 45(5), Article 055025. https://doi.org/10.1088/1361-6579/ad4e91
Qiu, Jiaxing ; Di Fiore, Juliann M. ; Krishnamurthi, Narayanan et al. / Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants. In: Physiological Measurement. 2024 ; Vol. 45, No. 5.
@article{c2e4c3b4d5e44d4bbe3915671bf576ba,
title = "Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants",
abstract = "Objective. Highly comparative time series analysis (HCTSA) is a novel approach involving massive feature extraction using publicly available code from many disciplines. The Prematurity-Related Ventilatory Control (Pre-Vent) observational multicenter prospective study collected bedside monitor data from > 700 extremely preterm infants to identify physiologic features that predict respiratory outcomes. Approach. We calculated a subset of 33 HCTSA features on > 7 M 10 min windows of oxygen saturation (SPO2) and heart rate (HR) from the Pre-Vent cohort to quantify predictive performance. This subset included representatives previously identified using unsupervised clustering on > 3500 HCTSA algorithms. We hypothesized that the best HCTSA algorithms would compare favorably to optimal PreVent physiologic predictor IH90_DPE (duration per event of intermittent hypoxemia events below 90%). Main Results. The top HCTSA features were from a cluster of algorithms associated with the autocorrelation of SPO2 time series and identified low frequency patterns of desaturation as high risk. These features had comparable performance to and were highly correlated with IH90_DPE but perhaps measure the physiologic status of an infant in a more robust way that warrants further investigation. The top HR HCTSA features were symbolic transformation measures that had previously been identified as strong predictors of neonatal mortality. HR metrics were only important predictors at early days of life which was likely due to the larger proportion of infants whose outcome was death by any cause. A simple HCTSA model using 3 top features outperformed IH90_DPE at day of life 7 (.778 versus .729) but was essentially equivalent at day of life 28 (.849 versus .850). Significance. These results validated the utility of a representative HCTSA approach but also provides additional evidence supporting IH90_DPE as an optimal predictor of respiratory outcomes.",
keywords = "highly comparative time series analysis, intermittent hypoxemia, predictive models, preterm infants",
author = "Jiaxing Qiu and {Di Fiore}, {Juliann M.} and Narayanan Krishnamurthi and Premananda Indic and Carroll, {John L.} and Nelson Claure and Kemp, {James S.} and Dennery, {Phyllis A.} and Namasivayam Ambalavanan and Weese-Mayer, {Debra E.} and {Maria Hibbs}, Anna and Martin, {Richard J.} and Eduardo Bancalari and Aaron Hamvas and {Randall Moorman}, J. and Lake, {Douglas E.}",
note = "Publisher Copyright: {\textcopyright} 2024 The Author(s). Published on behalf of Institute of Physics and Engineering in Medicine by IOP Publishing Ltd.",
year = "2024",
month = may,
day = "1",
doi = "10.1088/1361-6579/ad4e91",
language = "English (US)",
volume = "45",
journal = "Physiological Measurement",
issn = "0967-3334",
publisher = "IOP Publishing Ltd.",
number = "5",
}
Qiu, J, Di Fiore, JM, Krishnamurthi, N, Indic, P, Carroll, JL, Claure, N, Kemp, JS, Dennery, PA, Ambalavanan, N, Weese-Mayer, DE, Maria Hibbs, A, Martin, RJ, Bancalari, E, Hamvas, A, Randall Moorman, J & Lake, DE 2024, 'Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants', Physiological Measurement, vol. 45, no. 5, 055025. https://doi.org/10.1088/1361-6579/ad4e91
Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants. / Qiu, Jiaxing; Di Fiore, Juliann M.; Krishnamurthi, Narayanan et al.
In: Physiological Measurement, Vol. 45, No. 5, 055025, 01.05.2024.
Research output: Contribution to journal › Article › peer-review
TY - JOUR
T1 - Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants
AU - Qiu, Jiaxing
AU - Di Fiore, Juliann M.
AU - Krishnamurthi, Narayanan
AU - Indic, Premananda
AU - Carroll, John L.
AU - Claure, Nelson
AU - Kemp, James S.
AU - Dennery, Phyllis A.
AU - Ambalavanan, Namasivayam
AU - Weese-Mayer, Debra E.
AU - Maria Hibbs, Anna
AU - Martin, Richard J.
AU - Bancalari, Eduardo
AU - Hamvas, Aaron
AU - Randall Moorman, J.
AU - Lake, Douglas E.
N1 - Publisher Copyright:© 2024 The Author(s). Published on behalf of Institute of Physics and Engineering in Medicine by IOP Publishing Ltd.
PY - 2024/5/1
Y1 - 2024/5/1
N2 - Objective. Highly comparative time series analysis (HCTSA) is a novel approach involving massive feature extraction using publicly available code from many disciplines. The Prematurity-Related Ventilatory Control (Pre-Vent) observational multicenter prospective study collected bedside monitor data from > 700 extremely preterm infants to identify physiologic features that predict respiratory outcomes. Approach. We calculated a subset of 33 HCTSA features on > 7 M 10 min windows of oxygen saturation (SPO2) and heart rate (HR) from the Pre-Vent cohort to quantify predictive performance. This subset included representatives previously identified using unsupervised clustering on > 3500 HCTSA algorithms. We hypothesized that the best HCTSA algorithms would compare favorably to optimal PreVent physiologic predictor IH90_DPE (duration per event of intermittent hypoxemia events below 90%). Main Results. The top HCTSA features were from a cluster of algorithms associated with the autocorrelation of SPO2 time series and identified low frequency patterns of desaturation as high risk. These features had comparable performance to and were highly correlated with IH90_DPE but perhaps measure the physiologic status of an infant in a more robust way that warrants further investigation. The top HR HCTSA features were symbolic transformation measures that had previously been identified as strong predictors of neonatal mortality. HR metrics were only important predictors at early days of life which was likely due to the larger proportion of infants whose outcome was death by any cause. A simple HCTSA model using 3 top features outperformed IH90_DPE at day of life 7 (.778 versus .729) but was essentially equivalent at day of life 28 (.849 versus .850). Significance. These results validated the utility of a representative HCTSA approach but also provides additional evidence supporting IH90_DPE as an optimal predictor of respiratory outcomes.
AB - Objective. Highly comparative time series analysis (HCTSA) is a novel approach involving massive feature extraction using publicly available code from many disciplines. The Prematurity-Related Ventilatory Control (Pre-Vent) observational multicenter prospective study collected bedside monitor data from > 700 extremely preterm infants to identify physiologic features that predict respiratory outcomes. Approach. We calculated a subset of 33 HCTSA features on > 7 M 10 min windows of oxygen saturation (SPO2) and heart rate (HR) from the Pre-Vent cohort to quantify predictive performance. This subset included representatives previously identified using unsupervised clustering on > 3500 HCTSA algorithms. We hypothesized that the best HCTSA algorithms would compare favorably to optimal PreVent physiologic predictor IH90_DPE (duration per event of intermittent hypoxemia events below 90%). Main Results. The top HCTSA features were from a cluster of algorithms associated with the autocorrelation of SPO2 time series and identified low frequency patterns of desaturation as high risk. These features had comparable performance to and were highly correlated with IH90_DPE but perhaps measure the physiologic status of an infant in a more robust way that warrants further investigation. The top HR HCTSA features were symbolic transformation measures that had previously been identified as strong predictors of neonatal mortality. HR metrics were only important predictors at early days of life which was likely due to the larger proportion of infants whose outcome was death by any cause. A simple HCTSA model using 3 top features outperformed IH90_DPE at day of life 7 (.778 versus .729) but was essentially equivalent at day of life 28 (.849 versus .850). Significance. These results validated the utility of a representative HCTSA approach but also provides additional evidence supporting IH90_DPE as an optimal predictor of respiratory outcomes.
KW - highly comparative time series analysis
KW - intermittent hypoxemia
KW - predictive models
KW - preterm infants
UR - http://www.scopus.com/inward/record.url?scp=85195327360&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85195327360&partnerID=8YFLogxK
U2 - 10.1088/1361-6579/ad4e91
DO - 10.1088/1361-6579/ad4e91
M3 - Article
C2 - 38772400
AN - SCOPUS:85195327360
SN - 0967-3334
VL - 45
JO - Physiological Measurement
JF - Physiological Measurement
IS - 5
M1 - 055025
ER -
Qiu J, Di Fiore JM, Krishnamurthi N, Indic P, Carroll JL, Claure N et al. Highly comparative time series analysis of oxygen saturation and heart rate to predict respiratory outcomes in extremely preterm infants. Physiological Measurement. 2024 May 1;45(5):055025. doi: 10.1088/1361-6579/ad4e91