Multi-channel Impedance Pneumography (IP) system for non-invasive monitoring of spirometric parameters

Chronic Obstructive Pulmonary Disease (COPD) is one of the leading causes of morbidity and mortality worldwide, representing the third cause of death globally. Devices that monitor pulmonary function used to diagnose and track these conditions alter the patients’ regular breathing patterns and are typically uncomfortable. Indeed, there is currently no approved clinical method for monitoring of pulmonary flow and volume without highly obtrusive instrumentation like a mask. In this context, Impedance Pneumography (IP) is a promising method as a non-invasive alternative for continuous breathing monitoring. In literature, studies report its linear relation with respiratory volume and its capability to obtain typical respiratory parameters. This thesis aims to validate the IP method for continuous monitoring of pulmonary activity and function through the extraction of respiratory parameters useful in ambulatory and clinical scenarios. Two respiratory protocols were identified to investigate the behavior of Impedance Pneumography signal during passive and forced breathing conditions. The IP and spirometer volume signals were measured from 25 healthy subjects in static conditions from eight tetrapolar electrode configurations, distributed in both right and left hemithorax, using wet electrodes. The tests were carried out in collaboration with Dr. Nicola Launaro, head of the Ventilo-Therapy department of the Saluzzo Hospital (CN). The impedance measurement system was implemented at a prototype level using an evaluation board (EVAL- ADuCM355) equipped with an ARM-based microcontroller and the AD5940 Analog Front End (AFE). Multiple impedance signal acquisition was possible with analog MUXs mounted on PCBs realized with a semiconductor diode laser using an engraving method. In the first phase of the work, the IP signals under normal and quiet breathing conditions was investigated. The linear relationship of the IP signals with the reference volume was assessed by Pearson correlation (ρ). The results showed a mean value greater than 0.965 ± 0.02 for the best configurations in the central areas of the chest. The tidal volume and respiratory rate (RR), calculated cycle by cycle, were then estimated from IP signals. A root-mean-square error (RMSE) below 20% and less than 6% were found for the volume and the RR, respectively, for all configurations. Forced respiration conditions to extract spirometric parameters was then investigated. The Forced Exhalation Maneuver (FEM) must be performed as quickly as possible for the test to be considered acceptable, involving muscles such as the diaphragm, which play a minimal role under passive conditions. IP proved to be particularly sensitive to this effect. In addition, impedance variations attributed to subject movement compromised the good linearity observed during passive breathing, leading to errors in the estimation of spirometry parameters of over 40%. In conclusion, with this thesis work, it was shown that Impedance Pneumography measurement is a non-invasive and comfortable alternative to classical methods for continuous breathing monitoring in normal breathing. The IP can robustly estimate breath-by-breath volume and respiratory timings in static conditions. However, during forced breathing, the excessive movement of the subject and respiratory muscle contraction cause a drop of linearity with volume, making it complex to extract clinical parameters with high accuracy.