Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Lung --- Asthma --- Total Lung Capacity --- Lung Volume Measurements --- Albuterol --- growth & development --- physiopathology --- diagnostic use

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

The goal of this thesis is to try to build an algorithm that would automatically de-noise the box pressure signal of the body plethysmography ( Is a lung test performed to measure the compliance of the lungs by determining how much air the lungs can hold). The box pressure signal plays a significant role to measure the thoracic gas volume. Hence, a noisy box pressure signal gives an inaccurate thoracic gas volume measurement. Accordingly, three methods were proposed for de-noising the box pressure signal with the smallest risk possible.

The first two methods are in combination with a signal decomposition method so-call empirical mode decomposition (EMD), the advantage of using this method is to decompose the signal into a series signals; from high-frequency to low-frequency oscillations in which highlight the random noises that are not visible in the original signal. The EMD will be in combination with two different kinds of filters. The first filter will be a FIR low pass filter; this so-called method EMD-LPF. The second filter is called Savitzky-Golay (smoothing filter), this so-called method EMD-SG. The third method based on wavelet analysis, the thresholding technique will be used with a discrete wavelet transform for de-noising the signal, this method so-called DWT-db6/ DWT-sym6; where the sym and db are the fundamental mother wavelets. 

The performance of these algorithms has been compared in term of SNR and PRD, on the one hand. Their performance has been validated by total lung capacity (TLC: is the maximum volume of air the lungs can accommodate) test and specific airway resistance (sRaw: Inverse slope of the plot of flow rate versus box pressure) test with noisy and normal data set, on the other hand.


Each of the methods shows good results. More specifically, the EMD-SG shows reasonable TLC values, consistent sRaw values, maximum SNR and minimum PRD, with noisy and normal data set.

Body plethysmography --- De-noising --- Total lung capacity --- Pressure sensors --- Airways resistance --- Ingénierie, informatique & technologie > Multidisciplinaire, généralités & autres