Skip to Main content Skip to Navigation
Theses

Cardiorespiratory Monitoring by Microphone via Tracheal Sounds in the Context of Implanted Phrenic Nerve Stimulation

Abstract : Introduction: Patients who have an artificial ventilation dependence are usually treated with mechanical ventilation. If their phrenic nerves and diaphragms are still functional, implanted diaphragm pacing (DP) can provide them a more natural respiration. But existing DP systems cannot monitor patient’s induced respiration and stimulate with a current of constant intensity and frequency. Adding adaptive abilities to existing systems would improve the efficiency of the delivered stimulation and could also deliver an alarm in case of an apnea detection. A respiratory monitoring method based on recordings from tracheal sounds by microphone is introduced. This method aims at being ambulatory and non-invasive for all-day long real-time use.Methods: Tracheal sounds were recorded by a microphone inserted into a 3D printed bell-shape support, which was stuck over patient’s neck. Recorded signals were filtered and pre-amplified then saved and processed in a computer. Four protocols were designed to record and analyze respiration in different contexts: (1) 15 healthy subjects; (2) 1 patient with high tetraplegia under IT-PNS; (3) 13 healthy subjects in sitting and lying positions with reference signals; (4) 30 patients with sleep apnoea and 10 patients with implanted IT-PNS will be included in short future. One real-time processing algorithm detected all inspiration/expiration phases by combining results from temporal envelope detection, frequency detection, and also PDR detection (PCG-derived respiration: variation of cardiac peak amplitude corresponds to respiration). Combining the detection results in these 3 domains allowed for a better specificity of system (less false positives).Results: The application of the algorithms to protocols 1 to 3 data lead to good detection results that met the minimum requirements for the system. These result also showed that noises like speech or environments and different body position do not influence much the detection results. The new proposed respiratory effort evaluation method -- PDR showed a good correlation with EDR, this demonstrates the feasibility to use the PDR to monitor respiration. At a good correlation between R-R intervals and S-S intervals also showed that it would also be possible to monitor heart activity from tracheal sounds.Conclusions: This thesis shows the feasibility of detecting the apnoea and of monitoring cardiac activity from tracheal sounds. The extracted signal PDR could be used to identify the type of apnoea (obstructive/central). Furthermore, the microphone-based recording system could capture stimulation signals which can indicate a dysfunction of the pacing system and recorded tracheal sounds could give a feedback of the electro-ventilation quality. The next step is to record patient's tracheal sounds signals and test if the same algorithm could still work well in this context (the protocol 4).Monitoring tracheal sounds could provide a non-invasive way to approximate inspiratory flow that would be useful in all patients requiring respiratory monitoring in acute situations (e.g. as a safety measure during the administration of morphine for acute pain) and in chronic situations (e.g. home mechanical ventilation). And in the construction of smart house, especially for nursing/retirement home, application of tracheal sounds monitoring could provide convenient, robust multi-vital signals monitoring.
Complete list of metadatas

Cited literature [71 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02968839
Contributor : Abes Star :  Contact
Submitted on : Friday, October 16, 2020 - 10:28:10 AM
Last modification on : Saturday, October 17, 2020 - 3:25:42 AM

File

87891_LU_2020_archivage.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02968839, version 1

Collections

Citation

Xinyue Lu. Cardiorespiratory Monitoring by Microphone via Tracheal Sounds in the Context of Implanted Phrenic Nerve Stimulation. Micro and nanotechnologies/Microelectronics. Université Montpellier, 2020. English. ⟨NNT : 2020MONTS011⟩. ⟨tel-02968839⟩

Share

Metrics

Record views

45

Files downloads

12