Respiratory Rate Estimation during Walking and Running Using Breathing Sounds Recorded with a Microphone

Area of Science:

• Sports Science
• Biomedical Engineering
• Wearable Technology

Background:

• Respiratory frequency (f) is a key indicator of physical exertion.
• Monitoring f during exercise presents challenges like motion artifacts.
• Microphone sensors, less susceptible to motion artifacts, are underutilized for breathing monitoring.

Purpose of the Study:

• To evaluate a facemask-integrated microphone for estimating respiratory frequency (f) during various exercise intensities.
• To assess the accuracy of microphone-based f estimation against a reference flowmeter system.

Main Methods:

• A microphone embedded in a facemask captured breath sounds.
• Respiratory frequency (f) was calculated from exhalation events in 30-second intervals.
• Ten healthy subjects performed rest, walking, and running protocols.
• Data were compared against an orifice flowmeter using Mean Absolute Error (MAE) and Mean of Differences (MOD).

Main Results:

• The microphone system showed good agreement with the reference flowmeter.
• Accuracy decreased slightly with increased exercise intensity and ambient noise.
• Maximum MAE was 3.8 bpm and MOD was -2.0 bpm during running at 12 km/h.
• Overall MAE was 1.7 bpm, and MOD ± LOAs was -0.24 ± 5.07 bpm across all conditions.

Conclusions:

• Facemask-mounted microphone sensors are a suitable option for estimating respiratory frequency (f) during exercise.
• This technology can provide valuable insights into athletes' physical effort and training load.
• Further research may refine this non-invasive method for widespread athletic use.