Philipp Dörfler

Bachelor's Thesis

Comparison of Algorithms for Respiratory Information Extraction from Wearable Sensors

Advisors
Daniel Krauß (M.Sc.), Robert Richer (M.Sc.), Prof. Dr. Björn Eskofier, Dr. Janina Beilner

Duration
06 / 2022 – 11 / 2022

Abstract
Unusual changes in biosignals such as heart rate, blood pressure, or respiration rate are indicators of a variety of diseases such as heart disorders, stroke, or obstructive pulmonary disease [1]. For that reason, accurate monitoring of vital signs is crucial for diagnosis and treatment. Thereby, wearable sensors represent a promising, unobtrusive way to capture various biosignals, providing physicians with more real-world data and contributing to better diagnosis and treatment [2]. While analyzing real-world cardiac information is already established in clinical treatment, the collection of respiratory information is often not taken into account due to obtrusive sensor systems like nasal flow devices [3], [4], or impedance pneumography [4], [5]. However, respiratory information should be examined more frequently, since it can indicate various disorders, thus increasing the quality of diagnosis and therapy [1]. Therefore, unobtrusive methods to collect real-world respiratory information are necessary.

Over the past decades, several methods and algorithms were developed to derive respiratory information from other, more unobtrusive sensor modalities. One of them is the measurement of chest movement via Inertial Measurement Unit (IMU) sensors to estimate respiration rate [6]. Another approach is to extract respiratory information from an electrocardiogram (ECG) [7], [8] since breathing modulates the ECG signal through respiratory sinus arrhythmia (RSA), which leads to shorter RR-Intervals during inspiration and longer ones during expiration. Furthermore, the rotation of the cardiac vector is influenced by respiration, leading to changes in the ECG waveform [9].

Although there are various publications that investigated different approaches to unobtrusively extract respiratory information from ECG or IMU signals [6]–[9], there is a lack of systematic comparison of both approaches, or whether combining both modalities would improve the signal quality.

The goal of this bachelor’s thesis is therefore to implement different algorithms to derive respiratory information from ECG and IMU and to systematically compare different approaches on a dataset containing ECG and IMU data, as well as synchronized respiratory ground truth data. Furthermore, the potential of combining IMU and ECG data will be explored to further improve the extraction of respiratory information.

Full Thesis

References

[1] J. Boyle, N. Bidargaddi, A. Sarela, and M. Karunanithi, ‘Automatic Detection of Respiration Rate From Ambulatory Single-Lead ECG’, IEEE Trans. Inf. Technol. Biomed., vol. 13, no. 6, pp. 890–896, Nov. 2009, doi: 10.1109/TITB.2009.2031239.
[2] D. Dias and J. Paulo Silva Cunha, ‘Wearable Health Devices—Vital Sign Monitoring, Systems and Technologies’, Sensors, vol. 18, no. 8, Art. no. 8, Aug. 2018, doi: 10.3390/s18082414.
[3] G. Ottaviano and W. J. Fokkens, ‘Measurements of nasal airflow and patency: a critical review with emphasis on the use of peak nasal inspiratory flow in daily practice’, Allergy, vol. 71, no. 2, pp. 162–174, 2016, doi: 10.1111/all.12778.
[4] M. Młyńczak and G. Cybulski, ‘Impedance pneumography: Is it possible?’, in Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2012, Oct. 2012, vol. 8454, pp. 542–555. doi: 10.1117/12.2000223.
[5] A. F. Pacela, ‘Impedance pneumography—A survey of instrumentation techniques’, Med. Biol. Eng., vol. 4, no. 1, pp. 1–15, Jan. 1966, doi: 10.1007/BF02474783.
[6] A. Cesareo, Y. Previtali, E. Biffi, and A. Aliverti, ‘Assessment of Breathing Parameters Using an Inertial Measurement Unit (IMU)-Based System’, Sensors, vol. 19, no. 1, p. 88, Dec. 2018, doi: 10.3390/s19010088.
[7] G. Moody, R. Mark, A. Zoccola, and S. Mantero, ‘Derivation of Respiratory Signals from Multilead ECGs’, Comput. Cardiol., vol. 12, Jan. 1985.
[8] B. Mazzanti, C. Lamberti, and J. de Bie, ‘Validation of an ECG-derived respiration monitoring method’, in Computers in Cardiology, 2003, Sep. 2003, pp. 613–616. doi: 10.1109/CIC.2003.1291230.
[9] R. Pallas-Areny, J. Colominas-Balague, and F. J. Rosell, ‘The effect of respirationinduced heart movements on the ECG’, IEEE Trans. Biomed. Eng., vol. 36, no. 6, pp. 585–590, Jun. 1989, doi: 10.1109/10.29452.