Jean-Pierre L'Huillier

S-transform applied to laser Doppler flowmetry reactive hyperemia signals

Laser Doppler flowmetry signals give information about many physiological activities of the cardiovascular system. The activities manifest themselves in rhythmic cycles. In order to explore these activities during the reactive hyperemia phenomenon, a novel time-frequency method, called the S-transform, based on a scalable Gaussian wavelet, is applied. The goal is to have a deeper understanding of reactive hyperemia. This paper focuses on the evaluation of the different activities between a rest signal and an hyperemia signal, both acquired simultaneously on the two forearms of healthy subjects. The results show that after the release of the occlusion, the myogenic, neurogenic, and endothelial related activities clearly increase on the forearm where the occlusion took place. Then, they return progressively to their basal level. However, on the rest forearm, no increase is noted for the three activities. The mechanisms that take place during reactive hyperemia are, therefore, local. The S-transform proves to be a suited time-frequency method, in order to analyze laser Doppler signal underlying mechanisms.

Time-frequency analysis of laser Doppler flowmetry signals recorded in response to a progressive pressure applied locally on anaesthetized healthy rats.

The laser Doppler flowmetry technique has recently been used to report a significant transient increase of the cutaneous blood flow signal, in response to a local non-noxious pressure applied progressively on the skin of both healthy humans and rats. This phenomenon is not entirely understood yet. In the present work, a time-frequency analysis is applied to signals recorded on anaesthetized healthy rats, at rest and during a cutaneous pressure-induced vasodilation (PIV). The comparison, at rest and during PIV, of the scalogram relative energies and scalogram relative amplitudes in five bands, corresponding to five characteristic frequencies, shows an increased contribution for the endothelial related metabolic activity in PIV signals, till 400 s after the beginning of the progressive pressure application. The other subsystems (heart, respiration, myogenic and neurogenic activities) contribute relatively less during PIV than at rest. The differences are statistically significant for all the relative activities in the interval 0-200 s following the beginning of the pressure. These results and others obtained on patients, such as diabetics, could increase the understanding of some cutaneous pathologies involved in various neurological diseases and in the pathophysiology of decubitus ulcers.

Spectral components of laser Doppler flowmetry signals recorded in healthy and type 1 diabetic subjects at rest and during a local and progressive cutaneous pressure application: scalogram analyses

A significant transient increase in laser Doppler flowmetry (LDF) signals is observed in response to a local and progressive cutaneous pressure application in healthy subjects. This reflex may be impaired in diabetic patients. The work presents a signal processing providing the clarification of this phenomenon. Scalogram analyses of LDF signals recorded at rest and during a local and progressive cutaneous pressure application are performed on healthy and type 1 diabetic subjects. Three frequency bands, corresponding to myogenic, neurogenic and endothelial related metabolic activities, are studied. The results show that, at rest, the scalogram energy of each frequency band is significantly lower for diabetic patients than for healthy subjects, but the scalogram relative energies do not show any statistical difference between the two groups. Moreover, the neurogenic and endothelial related metabolic activities are significantly higher during the progressive pressure than at rest, in healthy and diabetic subjects. However, the relative contribution of the endothelial related metabolic activity is significantly higher during the progressive pressure than at rest, in the interval 200-400 s following the beginning of the pressure application, but only for healthy subjects. These results may improve knowledge on cutaneous microvascular responses to injuries or local pressures initiating diabetic complications.

Simplified model of laser Doppler signals during reactive hyperaemia.

Laser Doppler flowmetry (LDF) is a non-invasive method to measure tissue blood flow. During reactive hyperaemia, the LDF signal increases to a peak and then returns to a resting value. A simplified model is developed to explain these variations. The emphasis is on simulating the effects occurring rather than on trying to mimic the anatomical structure of the microcirculation. A single blood vessel is therefore analysed. The increasing value of blood velocity is studied, and vasodilatation as well as vasoconstriction are taken into account. The model parameters are calculated using wavelets. For a 2-min occlusion on a healthy subject, the radius of the vessel is initially 15 μm, increasing to 24.6 μm at the peak, reached 14 s after the release of the occlusion. The model shows that the high value of the LDF signal during the initial phase of reactive hyperaemia is produced by an increasing number of erythrocytes in a cross-section, due to vasodilatation rather than an increase in moving blood cell velocities. Moreover, the rapidity of the vasodilatation and vasoconstriction effects determine the rapidity of the signal variations. The paper aims to give a basic solution to develop a numerical model.

Signal processing methodology to study the cutaneous vasodilator response to a local external pressure application detected by laser Doppler flowmetry

The existence of a cutaneous pressure-induced vasodilation (PIV) has recently been reported. This paper proposes a signal processing methodology to improve PIV knowledge. Temporal variations of laser Doppler signals rhythmic activities are first analyzed on anesthetized rats. The results lead to a method that provides a better PIV understanding.

Physiological effects of indomethacin and celecobix: an S-transform laser Doppler flowmetry signal analysis

Conventional signal processing typically involves frequency selective techniques which are highly inadequate for nonstationary signals. In this paper, we present an approach to perform time-frequency selective processing of laser Doppler flowmetry (LDF) signals using the S-transform. The approach is motivated by the excellent localization, in both time and frequency, afforded by the wavelet basis functions. Suitably chosen Gaussian wavelet functions are used to characterize the subspace of signals that have a given localized time-frequency support, thus enabling a time-frequency partitioning of signals. In this paper, the goal is to study the influence of various pharmacological substances taken by the oral way (celecobix (Celebrex), indomethacin (Indocid) and placebo) on the physiological activity behaviour. The results show that no statistical differences are observed in the energy computed from the time-frequency representation of LDF signals, for the myogenic, neurogenic and endothelial related metabolic activities between Celebrex and placebo, and Indocid and placebo. The work therefore proves that these drugs do not affect these physiological activities. For future physiological studies, there will therefore be no need to exclude patients having taken cyclo-oxygenase 1 inhibitions.

Wavelet de-noising of laser Doppler reactive hyperemia signals to diagnose peripheral arterial occlusive diseases

In order to improve peripheral arterial occlusive diseases (PAOD) diagnoses, five de-noising algorithms based on a multiresolution analysis computed with wavelets are applied on reactive hyperemia signals obtained with the laser Doppler flowmetry technique. Results are presented on recordings acquired on patients suffering from PAOD and on healthy subjects.

Empirical Mode Decomposition Applied to Laser Doppler Flowmetry Signals : Diagnosis Approach

Empirical mode decomposition (EMD) is a recently introduced tool for decomposing signals into so-called intrinsic mode functions (IMF). These IMF represent the data by means of oscillating waves with local zero mean. In some sense the decomposition can be compared with a time-varying filter bank, i.e., signals are decomposed using band limited filters with band widths that vary in time. The main attribute of EMD compared to other time-frequency tools is that it does not use any predetermined filters or transforms. It is therefore a self-contained method that preserves the physical properties in the separate IMF, explaining why it has been successfully applied in many engineering fields. This method is applied here on laser Doppler flowmetry signals and particularly on the hyperemia signals. Two interested hyperemia parameters are the maximum perfusion value and the corresponding time instant of appearance. Accurate values parameters are determined from the fifth IMF component. Computing these parameters allows us to improve diagnosis of some pathologies as peripheral arterial occlusive diseases

S-Transform Time-Frequency Feature Extraction Of Laser Doppler Flowmetry Signal Using Svd Decomposition

Extracting the most important features of input data is extremely important. In this paper we introduced a new approach of feature extraction based on singular value decomposition of the S-transform time-frequency matrix of the laser Doppler flowmetry signal. We used this new approach to study reactive hyperemia blood flow recordings. This kind of bio-data provides an interesting biomedical investigation for estimating microcirculation perfusion. The technique uses the singular vector associated with the time-frequency distribution to characterize the physiological activities behaviors during the hyperemia phenomena

Dynamic characteristics of the cutaneous vasodilator response to a local external pressure application detected by the laser Doppler flowmetry technique on anesthetized rats

The laser Doppler flowmetry technique has recently been used to report a significant transient increase of the cutaneous blood flow signal when a local non-noxious pressure is applied progressively on the skin (11.1 Pa/s). The present work analyses the dynamic characteristics of this vasodilatory reflex response on anaesthetised rats. A de-noising algorithm using wavelets is proposed to obtain accurate values of these dynamic characteristics. The blood flow peak and the time to reach this peak are computed on the de-noised recordings. The results show that the mean time to reach the peak of perfusion is 85.3 s (time t = 0 at the beginning of the pressure application). The mean peak value is 188.3 arbitrary units (a.u.), whereas the mean value of the perfusion before the pressure application is 113.4 a.u. The mean minimum value obtained at the end of the experiment is 60.7 a.u. This latter value is, on the average, reached 841.3 s after the beginning of the pressure application. The comparison of the dynamic characteristics, computed with the de-noising algorithm on signals obtained in other situations, will give a better understanding on some cutaneous lesions such as those present on diabetic people.