Toward Wearables for Bruxism Detection: Voluntary Oral Behaviors Sound Recorded Across the Head Depend on Transducer Placement

Mohammad Khair Nahhas, Jens C. Türp, Philippe C. Cattin, Nicolas Gerig, Elisabeth Wilhelm, Georg RauterClinical 2024 · DOI

TL;DR

This study investigates how the placement of acoustic transducers on the head affects the recording of various voluntary oral behaviors, including bruxism-related sounds. The findings suggest that the ear is a particularly advantageous location for transducer placement due to its ability to compensate for head movements and its social acceptability, offering a promising avenue for developing wearable bruxism detection devices.

What Problem Does This Paper Address?

Bruxism, a common parafunctional behavior involving teeth grinding or jaw clenching, can lead to significant health issues. While polysomnography is the gold standard for diagnosis, it's resource-intensive. Wearable devices offer a more accessible alternative, but accurately identifying bruxism-induced sounds amidst other oral noises (like eating, drinking, or speaking) is challenging. This paper tackles the problem of optimizing the detection of these subtle sounds by exploring how different transducer placements on the head influence the quality and distinctiveness of recorded acoustic emissions from various oral behaviors. The goal is to find optimal locations for wearable sensors that can reliably differentiate bruxism sounds from everyday activities.

Methods

The researchers employed a setup with eight bone-conduction transducers, six attached to the temporal, frontal, and zygomatic bones using medical tape, and two integrated into commercial earphones. Participants performed five distinct oral behaviors: jaw clenching, teeth grinding, reading, eating, and drinking, while also pressing a button to mark these events. The recorded acoustic data were analyzed using three key metrics: time-domain energy, spectral flux (measuring changes in spectral power between windows), and zero-crossing rate (ZCR, indicating signal noisiness). Statistical analysis, primarily a two-way ANOVA, was used to determine if transducer placement had a significant impact on these metrics for each behavior.

Datasets

Not discussed in the paper. The study used data collected from 15 participants during controlled experimental tasks.

Key Insights and Ideas

  • Oral behaviors leave distinct acoustic footprints: The study successfully recorded and analyzed sounds from various oral activities, demonstrating that they possess unique acoustic characteristics.
  • Transducer placement matters, especially for certain behaviors: The location of the acoustic sensor significantly impacts the quality of recordings for activities like reading and eating, suggesting that the path sound takes through different tissues and the proximity to the sound source are crucial.
  • The ear is a promising location: The ear emerged as a particularly advantageous placement for transducers. This is attributed to its ability to mitigate the effects of head movements and its social acceptability, making it suitable for long-term wearable applications.
  • Jaw clenching is tricky to isolate: While teeth grinding sounds were relatively consistent across placements, jaw clenching proved more challenging to characterize distinctly, with only a noticeable peak observed before the intended action, possibly related to tooth tapping.
  • Semiocclusion is a practical compromise: While full ear occlusion might enhance signal isolation, the study used semiocclusion for practical and hygienic reasons, acknowledging it as a trade-off.

Key Results

  • Teeth grinding: Transducer placement did not significantly impact the recording of teeth grinding sounds across energy, spectral flux, and ZCR.
  • Jaw clenching: Transducer placement had a significant impact on spectral flux (p < 0.01).
  • Reading and Eating: Transducer placement significantly impacted energy (p < 0.05 for reading, p < 0.01 for eating), spectral flux (p < 0.001 for reading, p < 0.01 for eating), and ZCR (p < 0.001 for both).
  • Drinking: Transducer placement had a significant impact only on ZCR (p < 0.01).
  • Pause Period: Transducer placement had a significant impact on energy and flux (p < 0.05 and p < 0.01, respectively).
  • Ear advantage: The ear transducers generally showed higher median and 75th percentile values for energy and flux, particularly for reading and eating.

Limitations and Future Work

The authors acknowledge several limitations:

  • Individual anatomical variations: Transducer attachment varied due to differences in head size, skin properties, and masseter muscle size.
  • Incomplete ear occlusion: The standard earpieces did not fully occlude the ear canal, potentially affecting signal isolation.
  • Voluntary vs. Involuntary behaviors: The bruxism-like movements were performed voluntarily, which may not perfectly replicate natural bruxism. Participants might have relied more on one side of the jaw, leading to imbalanced sound distribution.
  • Uncertainty in clenching peak: The observed peak before jaw clenching needs further investigation to confirm its origin (behavioral or protocol-induced).

Future work could focus on:

  • Investigating the impact of different ear occlusion types on signal quality.
  • Developing more robust methods for isolating jaw clenching sounds.
  • Exploring the long-term wearability and effectiveness of ear-based transducers for continuous bruxism monitoring in real-world settings.
  • Further research into the specific acoustic characteristics of various oral behaviors to improve classification algorithms.

Summary updated

Abstract

We were able to record the sounds of various oral behaviors from different locations on the head. However, the ears were an advantageous location to place the transducer, since they could compensate for various head movements and ear devices are socially tolerable.
BibTeX
@inproceedings{nahhas2024toward,
  author = {Mohammad Khair Nahhas and Jens C. Türp and Philippe C. Cattin and Nicolas Gerig and Elisabeth Wilhelm and Georg Rauter},
  title = {{Toward Wearables for Bruxism Detection: Voluntary Oral Behaviors Sound Recorded Across the Head Depend on Transducer Placement}},
  booktitle = {Clinical},
  year = {2024},
  doi = {10.1002/cre2.70001},
}

Loading related content...

Figure