Skip to Main content Skip to Navigation
Theses

Développement d'outils pour le suivi non-invasif de la pression intracrânienne par des produits de distorsion acoustiques

Abstract : Non-invasive measurement of intracranial pressure (ICP) has been a research topic for decades, as invasive methods carry risks of hemorrhages and infections. Non-invasive monitoring by cochlear responses has been suggested as a reliable method. One of the most common solutions considered in these methods is the Distortion Products of Otoacoustic Emissions (DPOAE). The DPOAE measurement is quick and simple, as it only requires sending a sound stimulation and recording the cochlear acoustic response using a single probe (such as an earphone). The DPOAE phase shift is the parameter that allows us to monitor the ICP variations. As this is a relative measure, it is necessary to have an individual reference that should not be changed during or between measurements. In addition, the age of the patient influences the DPOAE levels which generally reduce with aging. They are also fragile and very sensitive to the acoustic environment, especially the noise from the patient. This work first presents the development of a signal processing method based on signal distribution analysis for automatic identification and rejection of noisy sections, in order to improve the robustness of the signal extraction from acoustic background noise. The developed method was compared to the rejection thresholds based on operator visual analysis (standard method). The results indicated a statistically lower noise level and more stable signals when using the automatic method. A second study presents a probe positioning and air leak detection technique (AFPS), based on the frequency response analysis of the pressure measured in the external ear canal after wideband stimulation. In this study, a tilt table was used to induce a slight variation in the ICP in four positions (60°, 0°, -20° and again 60°) into two series of measurements. We analyzed the reproducibility between the two tests and the effects on the test results of probe displacement and air leakage when they were identified. These analyses indicated that the AFPS is able to correctly classify the signals according to the occurrence of air leaks or probe movement. In the last experiment, both techniques (automatic rejection and AFPS) were adapted to be used in real time, to identify and correct any problems before or during the test. We compared the modified device and the commercial device in two test sessions, so that each device was used for both ears. In each session, three tests were performed (T1, T2 and T3), each with five values recorded for each of the three positions (45°, 0° "lying down" and -10°). These three tests were performed to allow the analysis of the reproducibility of the measurements, without and with a probe replacement effect, and its accuracy (expressed as the standard deviation of the differences). The AFPS increases robustness, providing more consistent values in all analyses, especially if the probe is repositioned. Automatic rejection reduces the variability between the five measurements taken for the same position, thus increasing the stability and accuracy of the responses.
Complete list of metadatas

Cited literature [89 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02498123
Contributor : Abes Star :  Contact
Submitted on : Wednesday, March 4, 2020 - 10:41:11 AM
Last modification on : Wednesday, March 11, 2020 - 1:51:23 AM

File

2019CLFAS010_DA_SILVA.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02498123, version 1

Collections

Citation

Jeferson Jhone da Silva. Développement d'outils pour le suivi non-invasif de la pression intracrânienne par des produits de distorsion acoustiques. Neurosciences [q-bio.NC]. Université Clermont Auvergne, 2019. Français. ⟨NNT : 2019CLFAS010⟩. ⟨tel-02498123⟩

Share

Metrics

Record views

56

Files downloads

21