In-Ear Audio Wearable: Measurement of Heart and Breathing Rates for Health and Safety Monitoring
Abstract
OBJECTIVE: This paper examines the integration of a noninvasive vital sign monitoring feature into the workers' hearing protection devices (HPDs) by using a microphone positioned within the earcanal under the HPD. METHODS: 25 test-subjects were asked to breathe at various rhythms and intensities and these realistic sound events were recorded in the earcanal. Digital signal processing algorithms were then developed to assess heart and breathing rates. Finally, to test the robustness of theses algorithms in noisy work environments, industrial noise was added to the in-ear recorded signals and an adaptive denoising filter was used. RESULTS: The developed algorithms show an absolute mean error of 4.3 beats per minute (BPM) and 2.7 cycles per minute (CPM). The mean difference estimate is -0.44 BPM with a limit of agreement (LoA) interval of -14.3 to 13.4 BPM and 2.40 CPM with a LoA interval of -2.62 to 7.48 CPM. Excellent denoising is achieved with the adaptive filter, able to cope with ambient sound pressure levels of up to 110 dB SPL, resulting in a small error for heart rate detection, but a much larger error for breathing rate detection. CONCLUSION: Extraction of the heart and breathing rates from an acoustical measurement in the occluded earcanal under an HPD is possible and can even be conducted in the presence of a high level of ambient noise. SIGNIFICANCE: This proof of concept enables the development of a wide range of noninvasive health and safety monitoring audio wearables for industrial workplaces and life-critical applications where HPDs are used.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Heart rate and breathing rate are estimated from in-ear microphones, demonstrating potential of earables for health monitoring.
“Heart rate and breathing rate are also estimated from in-ear microphones, and demonstrated the potential of earables for health monitoring [44,84].”
From Human-centred artificial intelligence for mobile health sensing: challenges and opportunities (2023) · cites this paper as [84]
The citation lists haven’t been fetched yet
References 20 · Cited by 112
- Authors
- Martin, A., & Voix, J.
- Affiliations
- École de Technologie Supérieure (CA)
- Venue
- IEEE Transactions on Biomedical Engineering
- Published
- 2018-06
- Volume
- 65
- Issue
- 6
- Pages
- 1256-1263
- ISSN
- 0018-9294 (ISSN-L)
- 1558-2531
- OpenAlex
- W2724757864
- Cited by
- 112
- References
- 20
- Citations by year
- 2026: 6 · 2025: 19 · 2024: 18 · 2023: 14 · 2022: 20 · 2021: 12 · 2020: 8 · 2019: 13 · 2018: 2
- FWCI
- 2.83
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- IEEE Transactions on Biomedical Engineering (publishedVersion)
- PubMed (publishedVersion)
- Funders
- EERS Technologies, Inc.
- NSERC-EERS Industrial Research Chair in In-Ear Technologies
- Topics
- Physical Sciences › Engineering › Biomedical Engineering › Non-Invasive Vital Sign Monitoring
- Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › Heart Rate Variability and Autonomic Control
- Physical Sciences › Computer Science › Signal Processing › Speech and Audio Processing
- Keywords
- hearing protection devices, heart rate monitoring, breathing rate monitoring, ear canal acoustics, non-invasive vital sign monitoring, adaptive denoising, industrial noise, health and safety monitoring
- MeSH terms
- Adult, Algorithms, Ear, Equipment Design, Female, Heart Rate, Humans, Male, Middle Aged, Signal Processing, Computer-Assisted, Monitoring, Ambulatory, Young Adult, Respiratory Rate
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review