From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters

Objective Current adjustments of continuous positive airway pressure (CPAP) may expose patients to risks of respiratory episodes or oxygen desaturation. Therefore, this study developed machine learning models using leading indicators, such as heart rate variability (HRV) and oximetry-related metrics...

Full description

Saved in:
Bibliographic Details
Main Authors: Chih-Fan Kuo, Yi-Chih Lin, Ze-Yu Chen, Jiunn-Horng Kang, Cheng-Chen Chang, Zhihe Chen, Arnab Majumdar, Yen-Ling Chen, Yi-Chun Kuan, Kang-Yun Lee, Po-Hao Feng, Kuan-Yuan Chen, Hsin-Chien Lee, Wun-Hao Cheng, Wen-Te Liu, Cheng-Yu Tsai
Format: Article
Language:English
Published: SAGE Publishing 2025-04-01
Series:Digital Health
Online Access:https://doi.org/10.1177/20552076251339273
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849722138656768000
author Chih-Fan Kuo
Yi-Chih Lin
Ze-Yu Chen
Jiunn-Horng Kang
Cheng-Chen Chang
Zhihe Chen
Arnab Majumdar
Yen-Ling Chen
Yi-Chun Kuan
Kang-Yun Lee
Po-Hao Feng
Kuan-Yuan Chen
Hsin-Chien Lee
Wun-Hao Cheng
Wen-Te Liu
Cheng-Yu Tsai
author_facet Chih-Fan Kuo
Yi-Chih Lin
Ze-Yu Chen
Jiunn-Horng Kang
Cheng-Chen Chang
Zhihe Chen
Arnab Majumdar
Yen-Ling Chen
Yi-Chun Kuan
Kang-Yun Lee
Po-Hao Feng
Kuan-Yuan Chen
Hsin-Chien Lee
Wun-Hao Cheng
Wen-Te Liu
Cheng-Yu Tsai
author_sort Chih-Fan Kuo
collection DOAJ
description Objective Current adjustments of continuous positive airway pressure (CPAP) may expose patients to risks of respiratory episodes or oxygen desaturation. Therefore, this study developed machine learning models using leading indicators, such as heart rate variability (HRV) and oximetry-related metrics, to proactively predict optimal adjustment timings. Methods CPAP titration data were first collected from a sleep center in northern Taiwan. Subsequently, Continuous HRV and oximetry-related metrics were retrieved (60-s window with 1-s stride) and then labeled based on the presence of pressure adjustments. The dataset, comprising seven HRV and two oximetry-related parameters, was independently divided into training/validation (80%) and test (20%) datasets based on patient information. Five cross-sectional and two time-series models were established. The model with the highest accuracy and area under the receiver operating characteristic curve (AUROC) in the training/validation dataset was applied to the test dataset to investigate feature importance through permutation analysis. Results A dataset comprising 14,629 time-series cases from 374 patients undergoing CPAP therapy was obtained. The InceptionTime model outperformed others during the training/validation phase with accuracies of 78.27% and AUROC of 78.07%, and achieved 80.92% accuracy and 76.52% AUROC in the test phase. Feature importance analysis identified peripheral arterial oxygen saturation, its standard deviation over a 60-s window, and normalized power in the very-low-frequency band of HRV as the most impactful predictors. Conclusions The findings demonstrated the feasibility of incorporating HRV and oximetry-related metrics, as leading indicators, to proactively predict CPAP adjustment timings. Further research should consider integrating these metrics to improve CPAP therapy.
format Article
id doaj-art-902c6bfff6f4441c906ef9a0e0cd080a
institution DOAJ
issn 2055-2076
language English
publishDate 2025-04-01
publisher SAGE Publishing
record_format Article
series Digital Health
spelling doaj-art-902c6bfff6f4441c906ef9a0e0cd080a2025-08-20T03:11:26ZengSAGE PublishingDigital Health2055-20762025-04-011110.1177/20552076251339273From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parametersChih-Fan Kuo0Yi-Chih Lin1Ze-Yu Chen2Jiunn-Horng Kang3Cheng-Chen Chang4Zhihe Chen5Arnab Majumdar6Yen-Ling Chen7Yi-Chun Kuan8Kang-Yun Lee9Po-Hao Feng10Kuan-Yuan Chen11Hsin-Chien Lee12Wun-Hao Cheng13Wen-Te Liu14Cheng-Yu Tsai15 Artificial Intelligence Center, , Taichung, Taiwan Department of Otolaryngology, , New Taipei City, Taiwan TMU Research Center of Artificial Intelligence in Medicine and Health, , Taipei, Taiwan College of Biomedical Engineering, Taipei Medical University, Taipei, Taiwan School of Medicine, , Taichung, Taiwan Department of Civil and Environmental Engineering, , London, UK Department of Civil and Environmental Engineering, , London, UK College of Biomedical Engineering, Taipei Medical University, Taipei, Taiwan Department of Neurology, School of Medicine, College of Medicine, , Taipei City, Taiwan Respiratory Therapy Room, Division of Pulmonary Medicine, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan Respiratory Therapy Room, Division of Pulmonary Medicine, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan Division of Pulmonary Medicine, Department of Internal Medicine, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan Department of Psychiatry, , Taipei, Taiwan School of Respiratory Therapy, College of Medicine, , Taipei, Taiwan Sleep Center, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan Sleep Center, Taipei Medical University-Shuang Ho Hospital, New Taipei City, TaiwanObjective Current adjustments of continuous positive airway pressure (CPAP) may expose patients to risks of respiratory episodes or oxygen desaturation. Therefore, this study developed machine learning models using leading indicators, such as heart rate variability (HRV) and oximetry-related metrics, to proactively predict optimal adjustment timings. Methods CPAP titration data were first collected from a sleep center in northern Taiwan. Subsequently, Continuous HRV and oximetry-related metrics were retrieved (60-s window with 1-s stride) and then labeled based on the presence of pressure adjustments. The dataset, comprising seven HRV and two oximetry-related parameters, was independently divided into training/validation (80%) and test (20%) datasets based on patient information. Five cross-sectional and two time-series models were established. The model with the highest accuracy and area under the receiver operating characteristic curve (AUROC) in the training/validation dataset was applied to the test dataset to investigate feature importance through permutation analysis. Results A dataset comprising 14,629 time-series cases from 374 patients undergoing CPAP therapy was obtained. The InceptionTime model outperformed others during the training/validation phase with accuracies of 78.27% and AUROC of 78.07%, and achieved 80.92% accuracy and 76.52% AUROC in the test phase. Feature importance analysis identified peripheral arterial oxygen saturation, its standard deviation over a 60-s window, and normalized power in the very-low-frequency band of HRV as the most impactful predictors. Conclusions The findings demonstrated the feasibility of incorporating HRV and oximetry-related metrics, as leading indicators, to proactively predict CPAP adjustment timings. Further research should consider integrating these metrics to improve CPAP therapy.https://doi.org/10.1177/20552076251339273
spellingShingle Chih-Fan Kuo
Yi-Chih Lin
Ze-Yu Chen
Jiunn-Horng Kang
Cheng-Chen Chang
Zhihe Chen
Arnab Majumdar
Yen-Ling Chen
Yi-Chun Kuan
Kang-Yun Lee
Po-Hao Feng
Kuan-Yuan Chen
Hsin-Chien Lee
Wun-Hao Cheng
Wen-Te Liu
Cheng-Yu Tsai
From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
Digital Health
title From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
title_full From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
title_fullStr From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
title_full_unstemmed From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
title_short From reactive to proactive: Machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry-related parameters
title_sort from reactive to proactive machine learning models for continuous positive airway pressure adjustments using heart rate variability and oximetry related parameters
url https://doi.org/10.1177/20552076251339273
work_keys_str_mv AT chihfankuo fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT yichihlin fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT zeyuchen fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT jiunnhorngkang fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT chengchenchang fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT zhihechen fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT arnabmajumdar fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT yenlingchen fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT yichunkuan fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT kangyunlee fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT pohaofeng fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT kuanyuanchen fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT hsinchienlee fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT wunhaocheng fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT wenteliu fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters
AT chengyutsai fromreactivetoproactivemachinelearningmodelsforcontinuouspositiveairwaypressureadjustmentsusingheartratevariabilityandoximetryrelatedparameters