Deep learning-based event counting for apnea-hypopnea index estimation using recursive spiking neural networks
- Author
- Lorin Werthen-Brabants (UGent) , Yolanda Castillo-Escario, Willemijn Groenendaal, Raimon Jane, Tom Dhaene (UGent) and Dirk Deschrijver (UGent)
- Organization
- Project
- Abstract
- Objective: To develop a novel method for improved screening of sleep apnea in home environments, focusing on reliable estimation of the Apnea-Hypopnea Index (AHI) without the need for highly precise event localization. Methods: RSN-Count is introduced, a technique leveraging Spiking Neural Networks to directly count apneic events in recorded signals. This approach aims to reduce dependence on the exact time-based pinpointing of events, a potential source of variability in conventional analysis. Results: RSN-Count demonstrates a superior ability to quantify apneic events (AHI MAE 6.17 +/- 2.21) compared to established methods (AHI MAE 8.52 +/- 3.20) on a dataset of whole-night audio and SpO2 recordings (N = 33). This is particularly valuable for accurate AHI estimation, even in the absence of highly precise event localization. Conclusion: RSN-Count offers a promising improvement in sleep apnea screening within home settings. Its focus on event quantification enhances AHI estimation accuracy. Significance: This method addresses limitations in current sleep apnea diagnostics, potentially increasing screening accuracy and accessibility while reducing dependence on costly and complex polysomnography.
- Keywords
- SLEEP-APNEA, PREDICTION, SOUND, Sleep apnea detection, AHI estimation, deep learning, deep learning, spiking neural networks, spiking neural networks, wearables, wearables, wearables
Downloads
-
8695 acc.pdf
- full text (Accepted manuscript)
- |
- open access
- |
- |
- 3.18 MB
-
(...).pdf
- full text (Published version)
- |
- UGent only
- |
- |
- 2.84 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01JQXFR3GCY7TX1F5BYQ9DHSP8
- MLA
- Werthen-Brabants, Lorin, et al. “Deep Learning-Based Event Counting for Apnea-Hypopnea Index Estimation Using Recursive Spiking Neural Networks.” IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, vol. 72, no. 4, 2025, pp. 1306–15, doi:10.1109/TBME.2024.3498097.
- APA
- Werthen-Brabants, L., Castillo-Escario, Y., Groenendaal, W., Jane, R., Dhaene, T., & Deschrijver, D. (2025). Deep learning-based event counting for apnea-hypopnea index estimation using recursive spiking neural networks. IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, 72(4), 1306–1315. https://doi.org/10.1109/TBME.2024.3498097
- Chicago author-date
- Werthen-Brabants, Lorin, Yolanda Castillo-Escario, Willemijn Groenendaal, Raimon Jane, Tom Dhaene, and Dirk Deschrijver. 2025. “Deep Learning-Based Event Counting for Apnea-Hypopnea Index Estimation Using Recursive Spiking Neural Networks.” IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING 72 (4): 1306–15. https://doi.org/10.1109/TBME.2024.3498097.
- Chicago author-date (all authors)
- Werthen-Brabants, Lorin, Yolanda Castillo-Escario, Willemijn Groenendaal, Raimon Jane, Tom Dhaene, and Dirk Deschrijver. 2025. “Deep Learning-Based Event Counting for Apnea-Hypopnea Index Estimation Using Recursive Spiking Neural Networks.” IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING 72 (4): 1306–1315. doi:10.1109/TBME.2024.3498097.
- Vancouver
- 1.Werthen-Brabants L, Castillo-Escario Y, Groenendaal W, Jane R, Dhaene T, Deschrijver D. Deep learning-based event counting for apnea-hypopnea index estimation using recursive spiking neural networks. IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING. 2025;72(4):1306–15.
- IEEE
- [1]L. Werthen-Brabants, Y. Castillo-Escario, W. Groenendaal, R. Jane, T. Dhaene, and D. Deschrijver, “Deep learning-based event counting for apnea-hypopnea index estimation using recursive spiking neural networks,” IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, vol. 72, no. 4, pp. 1306–1315, 2025.
@article{01JQXFR3GCY7TX1F5BYQ9DHSP8,
abstract = {{Objective: To develop a novel method for improved screening of sleep apnea in home environments, focusing on reliable estimation of the Apnea-Hypopnea Index (AHI) without the need for highly precise event localization. Methods: RSN-Count is introduced, a technique leveraging Spiking Neural Networks to directly count apneic events in recorded signals. This approach aims to reduce dependence on the exact time-based pinpointing of events, a potential source of variability in conventional analysis. Results: RSN-Count demonstrates a superior ability to quantify apneic events (AHI MAE 6.17 +/- 2.21) compared to established methods (AHI MAE 8.52 +/- 3.20) on a dataset of whole-night audio and SpO2 recordings (N = 33). This is particularly valuable for accurate AHI estimation, even in the absence of highly precise event localization. Conclusion: RSN-Count offers a promising improvement in sleep apnea screening within home settings. Its focus on event quantification enhances AHI estimation accuracy. Significance: This method addresses limitations in current sleep apnea diagnostics, potentially increasing screening accuracy and accessibility while reducing dependence on costly and complex polysomnography.}},
author = {{Werthen-Brabants, Lorin and Castillo-Escario, Yolanda and Groenendaal, Willemijn and Jane, Raimon and Dhaene, Tom and Deschrijver, Dirk}},
issn = {{0018-9294}},
journal = {{IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING}},
keywords = {{SLEEP-APNEA,PREDICTION,SOUND,Sleep apnea detection,AHI estimation,deep learning,deep learning,spiking neural networks,spiking neural networks,wearables,wearables,wearables}},
language = {{eng}},
number = {{4}},
pages = {{1306--1315}},
title = {{Deep learning-based event counting for apnea-hypopnea index estimation using recursive spiking neural networks}},
url = {{http://doi.org/10.1109/TBME.2024.3498097}},
volume = {{72}},
year = {{2025}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: