Someone’s noise is someone else’s signal

It takes some effort

Whenever someone suffers from sleep-disordered breathing (SDB) or obstructive sleep apnea (OSA), it is useful to measure their respiratory effort to understand their breathing dynamics in detail. Even without clearly defined apneic events, patterns of increasing respiratory effort may be a marker of physiological dysfunctions [1].

In current clinical practice, measuring the respiratory effort involves invasive procedures, such as esophageal pressure sensors, that disrupt sleep or estimated with proxies, i.e., using respiratory belts. A more recent technique employs a pressure sensor placed on the suprasternal notch (SSN) to measure pressure swings in the trachea and allow for a more faithful detection of respiratory effort. [2]

Listening to a crowd

A downside of the suprasternal pressure sensor, common to other respiratory sensors, is its ability to detect signals extraneous to respiratory activity, such as snoring, and air swings caused by the beating of the heart, called cardiogenic oscillations. Snoring sits in the audible frequency range, so it’s easy to remove it. Instead cardiogenic oscillations are close to respiratory frequency and can be strong enough to hamper the quality of the respiratory signal and limit its clinical value.

Free two birds with one key

Although cardiogenic oscillations are considered noise and filtered out with various techniques [3,4], we see them as a potentially interesting source of information, as they may highlight certain cardiac mechanisms related to pulmonary arteries.

Therefore, we designed a method that estimates heart rate directly from the SSn signal, and tune a time-varying filter upon that. Then we can obtain two signals from a single sensor: respiratory activity cleaned of cardiogenic oscillations and cardiac activity uncoupled from respiration. 

We believe that similar techniques may lead to more compact, but information-rich, sensing methodologies to measure and analyse sleep, driving research further.

Check out Luca Cerina’s ePoster on “Extraction of cardiac-related signal from suprasternal pressure sensor in polysomnography” at Sleep Europe 2022, by logging in to the virtual platform. Alternatively, you can read the abstract here.

References

1. Chandra, Subani et al. “Respiratory effort-related arousals contribute to sympathetic modulation of heart rate variability.” Sleep and Breathing 17 (2013): 1193-1200.
2. Glos, Martin et al. “Characterization of Respiratory Events in Obstructive Sleep Apnea Using Suprasternal Pressure Monitoring.” Journal of clinical sleep medicine : JCSM : official publication of the American Academy of Sleep Medicine 14 3 (2018): 359-369.
3. Zara, Michael et al. “Suppression of Cardiogenic Oscillations in Esophageal Pressure Signals Using Ensemble Empirical Mode Decomposition.” (2018).
4. Pagano, Parwane P et al. “Separation of cardiogenic oscillations from airflow waveforms using singular spectrum analysis.” Computer methods and programs in biomedicine 220 (2022): 106803.