MEDICAL ANALYSIS METHOD USING IMPEDANCE SIGNAL PROCESSING

Description

[0001] Since the 1950s, the counting and volumetry of different blood cells in hematology machines have been carried out by impedance measurement, according to a method known as the Coulter Principle. This method consists of passing the cells suspended in a conductive liquid through a polarized micro-orifice and detecting the variations in electrical resistance (or impedance variations) induced by the passage of particles in the orifice. The detection of the different pulses thus generated allows the counting of the elements.

[0002] Various solutions have been developed to address the problems associated with flow through an orifice (rotations due to edge or hydrodynamic effects, masking of doublets, etc.). As none of these solutions were satisfactory, hydrodynamic focusing technology, or hydrofocusing, was developed. This solution consists of hydrodynamically sheathing the flow of cells to be analyzed, which allows it to be centered in the orifice and limits the effects related to passages at the edge. This technique is nevertheless very complex to implement and particularly expensive.

[0003] The Applicant has recently developed a solution that provides very satisfactory results while avoiding hydrofocusing. It has protected this solution in patent application FR 1904410. During this development, the Applicant realized that its work could also be used to characterize cells from their impedance signals, in order to return information on normality or abnormality or to characterize the morphology of these cells.

[0004] The state of the art on existing techniques for measuring the deformability of red blood cells reveals two main families.

[0005] The first family concerns precise but time-consuming and complex measurement techniques to implement. For some of them, analytical models make it possible to trace rheological parameters such as elastic and viscous membrane moduli, among others. Among these methods, we can cite micropipette aspiration, the principle of which is to aspirate a part of a red blood cell into a pipette by imposing a known depression. By measuring certain quantities related to the shape of the red blood cell after aspiration, the shear modulus or the shear viscosity can be deduced (see for example the article by EA Evans "New membrane concept applied to the analysis of fluid shear and micropipette deformed red blood cells" Biophysical Journal, 1973). Other methods, such as the method « optical tweezer » (see for example the article by Brandao et al. “Optical tweezers for measuring red blood cell elasticity: application to the study of drug response in sickle cell disease” European Journal of Haematology, 70:207-211, 2003), or the method “ optical stretcher » (see for example the article by Guck et al. "The optical stretcher: a novel laser tool to micromanipulate cells" Biophysical Journal, 81:767-784, 2001) consist of stretching the globule using optical lasers and observing its shape.

[0006] The second family includes techniques that allow for faster and more autonomous processing of a large number of cells and thus provide a statistical idea of ​​the deformability of the red blood cells in a sample. This type of technique studies the deformation of red blood cells using a shape parameter, called the Deformation Index (DI), which is in fact a measure of the elongation of the red blood cell subjected to a well-known mechanical stress. The deformation index combines the mechanical and morphological parameters of the red blood cell: it is therefore simpler to study but does not provide precise rheological information. Among these techniques, we can mention the one presented in the article by Cha et al. "Cell stretching measurement utilizing viscoelastic particle focusing" Analytical chemistry 2012, in which the red blood cells are stretched in an extensional flow and observed using a camera.On the same principle, but by subjecting the red blood cells to shear flow, the method described in the article by Dobbe et al. "Analyzing red blood cell-deformability distributions", Blood Cells, Molecules, and Diseases, 28:373-384, 2002, has shown the impact of certain pathologies on the distributions of the deformation index of red blood cells. The article by Mohandas et al. "Analysis of factors regulating erythrocyte deformability", Journal of Clinical Investigations, 66:563-573, 1980, describes an ektacytometer in which red blood cells are subjected to shear flow, and for which the light diffraction spectra are observed to measure the deformation index. By studying the curve of the deformation index as a function of the osmolarity of the suspending medium (see for example the article by Clark et al.“Osmotic gradient ektacytometry: comprehensive characterization of red cell volume and surface maintenance”, Blood, 61:899-910, 1983), it has been shown that certain rheological and / or morphological parameters can be measured.

[0007] Although richer in information, the methods of the first family do not allow for high-speed hematological analysis. Indeed, they are too long and complex to implement and require the intervention of a specialized manipulator. On the other hand, the methods of the second family, although easier to implement, remain complicated to industrialize in the context of a medical analysis laboratory, although the study of a more global response with regard to the deformation index that they offer makes it possible to isolate subpopulations of pathological red blood cells.

[0008] A method of analyzing a blood sample is known from US3919050A.

[0009] All of the above methods require an optical or video microscopy acquisition system. This is significantly more complicated and expensive to implement than an impedance measurement system. However, meters operating under the Coulter Principle are not capable of determining the strain index or a simplified version of it.

[0010] There is therefore a need to offer a simple measuring device allowing the discrimination of cell populations according to their morphological characteristics.

[0011] The invention improves the situation. To this end, it relates to a method for classifying a blood sample according to independent claim 1 comprising the following operations: a) receiving pulse data sets, each pulse data set comprising impedance value data associated each time with a time marker, these data together representing a curve of cellular impedance values ​​measured during the passage of a cell in a polarized orifice, the cells being derived from the blood sample, b) for each pulse data set b1. determining a maximum impedance value of the pulse data set, b2.calculating a high impedance value by multiplying the maximum impedance value by a high coefficient chosen from the range [0.7; 0.95], and determining in the pulse data set the time markers whose associated impedance value in the pulse data set is equal to the high impedance value, and calculating a high duration corresponding to the maximum duration between these time markers, and a low impedance value by multiplying the maximum impedance value by a low coefficient chosen from the range [0.1; 0.6], and determining in the pulse data set the time markers whose associated impedance value in the pulse data set is equal to the low impedance value, and calculating a low duration corresponding to the maximum duration between these time markers, b3.calculating a peak position value equal to the division of the difference between the time associated with the maximum impedance value and the first time corresponding to the low impedance value, and the low duration, and optionally a rotation value equal to the division of the high duration by the low duration, c) determining the statistical distribution of the pulse data sets according to the low duration / peak position value pairs or optionally the rotation value / peak position value pairs, the statistics being established in relation to a set of ranges of values ​​of pairs, d) classifying the blood sample by comparing the distribution of operation c) to a reference distribution of a healthy blood sample.

[0012] Other characteristics and advantages of the invention will appear more clearly on reading the following description, taken from examples given for illustrative and non-limiting purposes, taken from the drawings in which: There figure 1 represents a principle view of the measuring orifice within the framework of the invention, as well as the trajectories that a cell can take in it, The figure 2 represents pulse measurements for the trajectories of the figure 1 , There figure 3 and the figure 4 represent characterization diagrams of red blood cell impedance pulses measured with the arrangement of the figure 1 , There figure 5 represents a characterization diagram of areas of interest, The figures 6 à 9 represent proportions of pulses in some of the areas of interest of the figure 5 , There figure 10 represents a diagram of an example of a device which is not part of the invention, The figure 11 represents a diagram of a neural network implemented in the example of the figure 10 , There figure 12 represents a diagram of a neural network implemented in another example, The figure 13 represents a characterization diagram of areas of interest, The figures 14 à 16 represent representations of population statistics of pulses in some of the areas of interest of the figure 13 , and The figure 17 represents the analysis of two healthy blood samples over a period of 11 days, and The figure 18 represents the statistical evolution over time of the measurements of the figure 17 corresponding to the Box3' of the figure 5 .

[0013] The drawings and the description below contain, for the most part, elements of a certain character. They may therefore not only serve to better understand the present invention, but also contribute to its definition, if necessary.

[0014] There figure 1 represents a principle view of the measuring orifice within the framework of the invention, as well as the trajectories that a cell can take in it. The orifice has walls represented in dotted lines, the abscissa and the ordinate being expressed in µm.

[0015] There figure 2 represents the measured impedance pulses (taken from the signal corresponding to the variation in impedance of the system due to the passage of a cell in the micro-orifice) for each of the trajectories of the figure 1 The abscissa is expressed in µs, while the ordinate is expressed in ohms. As can be seen, the closer the cell has an incident trajectory to one of the walls of the orifice, the more chaotic the measurement is, and the cause of errors explained in the introduction.

[0016] The Applicant's work set out in application FR 1904410 led it to develop a new quantity to characterize impedance pulses. This quantity is called WR, and it is a ratio between two pulse widths. These widths make it possible to indicate the presence of a peak in the pulse, or on the contrary a bell-shaped pulse.

[0017] To do this, we first determine the maximum pulse height in the pulse dataset. The maximum height is used to calculate a high impedance value and a low impedance value.

[0018] The high impedance value is obtained by multiplying the maximum impedance value (which corresponds to the maximum height) by a high coefficient. This high coefficient is used to determine two instants which, in general, allow a good approximation of the width of the impedance peak of a pulse. For this, the high coefficient is chosen in the range [0.7; 0.95], and preferably [0.8; 0.9], which ensures that there are at least two instants, and that these instants correspond to the peak of the pulses (in order to limit the cases where several peaks are present).

[0019] Thus, the high impedance value is lower than the maximum impedance value and higher than 70% of it. The Applicant's work has shown that this range allows the peaks of the pulses produced to be captured well. The Applicant has identified that the value of 0.875 is particularly advantageous and gives the best results: it allows the peaks of the pulses to be estimated most accurately. Indeed, the peaks around the maximum height are generally quite narrow.

[0020] The low impedance value is obtained by multiplying the maximum impedance value by a low coefficient. This low coefficient is used to determine two instants which, in general, allow the width of the impedance pulse to be approximated well. For this, the low coefficient is chosen in the range [0.1; 0.6], and preferably [0.3; 0.6] which ensures that there are two instants, and that these instants correspond to the general width of the pulse.

[0021] Thus, the low impedance value is between 30% and 60% of the maximum impedance value. The Applicant's work has shown that this range allows the width of the pulses produced to be captured well while eliminating noise. The Applicant has identified that the value of 0.5 is particularly advantageous and gives the best results: the slopes of the pulses below 50% of the maximum height are very steep, and this value avoids any risk of noisy measurements.

[0022] Once the high impedance value and the low impedance value have been determined, the duration between the two instants in the pulse data set that are the furthest apart in time, and that have the high impedance value or the low impedance value respectively, is determined. The duration associated with the high impedance value is called the high duration, and the duration associated with the low impedance value is called the low duration. Instinctively, it appears that the high duration corresponds approximately to the width of the impedance peak of a pulse data set, and that the low duration corresponds approximately to the pulse width. Finally, the WR quantity is determined by calculating the ratio between the high duration and the low duration.

[0023] The second quantity is called PP and is associated with the pulse peak. For this, this quantity is calculated by taking the ratio between the difference between the instant when the pulse is maximal and the first instant that corresponds to the low impedance, and the low duration. The result is a percentage that indicates the position of the pulse maximum in it.

[0024] The Applicant worked on pulse representations, in particular by establishing graphs of type (PP; WR). Indeed, it discovered that this type of graph allowed it to identify interesting behaviors, in particular for pulses characteristic of cells that had been rotated.

[0025] In order to validate its hypotheses, the Applicant altered the morphology of the red blood cells by adding specific molecules to the electrolytic solution.

[0026] Different concentrations of glutaraldehyde and N-dodecyl-N,N-dimethyl-3-ammonio-1-propanesulfonate (also called sulfobetaine 3-12, hereinafter SB3-12) were added to the dilution reagent, and then the impedances of healthy red blood cells immersed in these solutions were measured. Specifically, preparations comprising glutaraldehyde at concentrations between 0% and 0.5% were prepared on the one hand, and solutions comprising SB3-12 at concentrations between 0 mg / L and 90 mg / L were prepared on the other hand. The preparations were made separately, i.e., they each contain only an addition of glutaraldehyde or only an addition of SB3-12.

[0027] It is known that the use of glutaraldehyde has a fixative effect and allows the red blood cells to be rigidified while maintaining their discocyte shape. Similarly, it is known that the use of SB3-12 tends to spherize the cells.

[0028] A blood sample from a healthy patient (verified to be free of abnormalities, hereinafter referred to as "healthy blood") was analyzed with the different concentrations of SB3-12, and another was analyzed with all the different concentrations of glutaraldehyde. Each acquisition was performed twice, for a preliminary assessment of the repeatability of the proposed developments.

[0029] Finally, the Applicant calculated for each preparation the graph (PP; WR) of the pulses which were generated. figure 3 represents the graphs obtained for the preparations incorporating SB3-12, while the figure 4 represents the graphs obtained for preparations incorporating glutaraldehyde.

[0030] These graphs validated the Applicant's intuitions, namely that they contain information on the morphological characteristics of red blood cells. Thus, the Applicant established pulse zones (Box1' to Box6') to be studied from the graph (PP; WR) of healthy blood, as shown in the figure 5 . The pulse zones of the figure 5 were chosen to highlight the differences between acquisitions with glutaraldehyde and acquisitions with SB3-12 (see figure 3 And figure 4 ).

[0031] On the figure 5 , the pulse zones were defined as follows, each pair expressing a PP (min; max) and WR (min; max) range: Box1': (25; 60) - (58; 76) Box2': (60; 83) - (65; 76) Box3': (25, 85) - (76, 86) Box4': (5, 20) - (5, 70) Box5': (5, 25) - (70, 85) Box6': (20, 85) - (10, 58)

[0032] In each area of ​​the figure 5 , the Applicant calculated the proportion of pulses, and the averages of the PP and WR quantities. For a given sample, there are therefore 18 parameters in total (3 parameters for each of the 6 zones).

[0033] Then, the evolutions of the proportion of pulses in the Box3' zone and the Box5' zone were represented, as a function of the concentration in SB3-12 (on the figures 6 et 7 respectively), and depending on the glutaraldehyde concentration (on the figures 8 et 9 , respectively).

[0034] In each of the figures 6 à 9 , normality is represented by horizontal lines. The margin of error is represented by the dotted lines, and is defined as twice the standard deviation. The solid line between the dotted lines represents the average evaluated on 22 healthy blood samples defining normality.

[0035] The analysis of the figures 6 à 9 reveals that when the concentration of SB3-12 or glutaraldehyde is increased, the calculated parameters deviate from normality. These figures clearly show the influence of SB3-12 and glutaraldehyde on red blood cells, and how they change the morphological characteristics of the latter visibly in the pulses.

[0036] Assuming that glutaraldehyde concentration and SB3-12 concentration are correlated with RBC rigidity and sphericity, respectively, it seems possible to measure these parameters. In fact, by combining the pulse proportion and PP magnitude averages for the 3' zone on a graph, it becomes possible to quantitatively distinguish between RBCs mixed with a glutaraldehyde preparation and those mixed with a SB3-12 preparation.

[0037] All these elements have made it possible to empirically validate the fact that impedance pulses contain information concerning the morphological characteristics of red blood cells, but there does not seem to be a simple function allowing the normality of red blood cells to be measured, nor to precisely characterize their morphological abnormality if necessary.

[0038] The Applicant therefore had the idea of ​​developing a device using a first neural network trained and configured to indicate the normality or abnormality of a cell on the basis of its impedance pulse, and a device using a second neural network trained and configured to classify a cell by indicating whether it has normal morphological characteristics, morphological characteristics of a rigidified cell, or morphological characteristics of a spherized cell.

[0039] There figure 10 represents a generic diagram of this device. The device comprises a memory 4, and a classifier (or classifier) ​​6.

[0040] Memory 4 can be any type of data storage suitable for receiving digital data: hard disk, solid-state drive (SSD), flash memory in any form, RAM, magnetic disk, locally or cloud-distributed storage, etc. The data calculated by the device can be stored on any type of memory similar to memory 4, or on it. This data can be erased after the device has performed its tasks or retained.

[0041] In the example described here, memory 4 receives pulse data sets. A pulse data set represents all the data that can be used to characterize an impedance pulse represented on the figure 2 . It is therefore a set of couples (measured impedance value; time marker), which together define a curve like those of the figure 2 . In practice, the pulse data set will generally be a sampling of the orifice detection output. The pulse data set can also be a continuous curve, in which case the calculator 6 will be adapted accordingly.

[0042] The classifier 6 is an element directly or indirectly accessing the memory 4. It can be implemented in the form of appropriate computer code executed on one or more processors. By processors, it is meant any processor suitable for the calculations described below. Such a processor can be implemented in any known manner, in the form of a microprocessor for a personal computer, a dedicated chip of the FPGA or SoC type (“system on chip” in English), a computing resource on a grid or in the cloud, a microcontroller, or any other form suitable for providing the computing power necessary for the implementation described below. One or more of these elements can also be implemented in the form of specialized electronic circuits such as an ASIC. A combination of processor and electronic circuits can also be envisaged.

[0043] It should be noted that the device can advantageously be integrated into a hematological analysis device, or be remote. It can therefore be fully integrated into the hematological analysis device or, for example, be a web service to which the hematological analysis device connects when necessary or desired.

[0044] As suggested above, classifier 6 is a neural network. Indeed, pulses can be likened to images, and, with appropriate training, the Applicant considered that a neural network could be particularly effective in classifying pulses into sets of pulse data sets with or without rotation.

[0045] More particularly, the Applicant identified that a convolutional neural network was best suited. Thus, the architecture of the first neural network is shown in the figure 11 , while that of the second neural network is represented on the figure 12 .

[0046] In both cases, the neural network is a convolutional neural network that includes two convolution layers. Thus, a pulse 100 dataset (composed of 50 variables) is processed by a first convolution layer 110 that extracts 6 features, then a second convolution layer 120 extracts 3 features from layer 110.

[0047] The filters (or convolution kernels) of the first convolution layer 110 have a size of 8, those of the second convolution layer 120 have a size of 3.

[0048] The convolution layer 120 is connected to a fully connected layer 130 of the neural network which comprises a chain of 4 layers of neurons comprising respectively 80, 40, 20 and finally 10 neurons.

[0049] In the case of the first neural network, the fully connected layer 130 returns a value 140 in the output layer. In the example described here, the value 140 is 1 if the cell is normal and 0 if it is abnormal.

[0050] For all the neurons making up the different layers of the model, the activation function retained is the sigmoid function.

[0051] For this first neural network, training was performed using data from acquisitions of healthy blood defining normality, acquisitions with SB3-12 concentrations between 50 mg / L and 90 mg / L, and acquisitions with glutaraldehyde concentrations between 0.3% and 0.5%. The neural network was therefore trained to detect highly impacted cells. Each time, the training pulses were labeled with the value 1 if the associated cell was normal and with the value 0 if the associated cell was abnormal.

[0052] The training adequacy check was performed by discarding some data from the training and introducing them into the trained neural network. The results were excellent, and, with a threshold of 0.5 on the output layer (i.e., the value 1 is returned if the output layer returns a value greater than 0.5 and 0 otherwise), the false positive rate was 4.3% and the false negative rate was 3.1% on the validation pulse set.

[0053] In the case of the second neural network, the fully connected layer 130 returns a triplet 150 in the output layer. In the example described here, the three components of the triplet take the value 0 or 1.

[0054] For this second neural network, training is performed in a similar manner to the first neural network, except that the training pulses are labeled with triplets indicating whether a given pulse is normal ([1; 0; 0]), with spherization morphological characteristics ([0; 1; 0]), or with rigidification morphological characteristics ([0; 0; 1]).

[0055] The training adequacy check was performed by discarding some data from the training and introducing them into the trained neural network. The results were excellent, each element on the output layer being reduced to its maximum component, i.e. [0.92; 0.02; 0.06] returns the triplet [1; 0; 0], [0.01; 0.99; 0] returns the triplet [0; 1; 0], and [0.05; 0.25; 0.7] returns the triplet [0; 0; 1]. Under these conditions, on the validation pulse set, the false positive rate was 4% on cells classified as normal, 7.5% on cells classified with morphological characteristics of spherization, and 8.2% on cells classified with morphological characteristics of rigidification.

[0056] These results are excellent and demonstrate the relevance of the device according to the invention, which makes it possible to obtain extremely precise results with simple impedance measurements without hydrofocusing.

[0057] The Applicant's work enabled it to establish that a single convolution layer could suffice, as well as a fully connected layer which would contain only 2 layers or less.

[0058] Alternatively, the Applicant considers it possible to use a multi-layer perceptron (MLP) instead of the convolutional neural network described above. Indeed, even if this type of neural network provides less precise results with an equivalent number of parameters (typically, 5 to 8% additional false positives), it nevertheless constitutes a plausible alternative.

[0059] These results paved the way for the detection of pathologies that result in a modification of the morphological characteristics of red blood cells or other cells, such as malaria or sickle cell disease. Each time, it is sufficient to test healthy blood and diseased blood to label the corresponding pulses, and to train the neural network of the figures 11 ou 12 with this data.

[0060] For example, by analyzing a culture sample in which all the red blood cells are infected with malaria, a series of signatures labeled for this pathology is obtained and makes it possible to generate a classifier specific to this infection.

[0061] The Applicant has also identified that coupling the low duration with the PP quantity or the WR quantity with the PP quantity makes it possible to make a simple distinction between normal cells and abnormal cells on a statistical basis. This makes it possible, for example, to implement the invention without using a neural network described below.

[0062] Thus, on the basis of a set of measurements on healthy blood, the Applicant established the figure 13 In this figure, the Applicant grouped together all the quantities taken from the pulses, which enabled it to define the pulse zones Box1 to Box8.

[0063] On the figure 13 , the pulse zones were defined as follows, each pair expressing a PP (min; max) and WR (min; max) range: Box1: (70; 80) - (15.5; 18) Box2: (69; 79) - (18.5; 21) Box3: (50, 58) - (19, 24) Box4: (36, 42) - (20, 27) Box5: (26, 32) - (22, 32) Box6: (8, 22) - (26, 32) Box7: (8; 22) - (32; 38) Box8: (8; 22) - (38; 44)

[0064] Then, the Applicant repeated the same operation with the pulse data sets represented on the figure 3 and the figure 4 and established two criteria of normality by comparing the population statistics for each pulse zone, for healthy blood on the one hand and for the blood of figures 3 And 4 on the other hand. Indeed, the comparison of population statistics makes it possible to establish significant differences for each pulse zone.

[0065] So, on the figure 14 , the Applicant represented on the one hand the statistical distributions of the populations for each pulse zone for healthy blood (in solid lines, defining a zone in solid color), and on the other hand the value of healthy blood in dashed lines.

[0066] On the figure 15 (respectively the figure 16 ), the Applicant reproduced on the one hand the statistical distributions of the populations for each pulse zone for healthy blood (in solid lines, defining a zone in solid color, identical to those of the figure 14 ), and on the other hand in dashed lines these same distributions but for pulses obtained on blood containing glutaraldehyde (respectively SB3-12).

[0067] THE figures 14 à 16 clearly show that it is sufficient to perform population statistics for the pulses taken from a blood sample and compare them with the normality envelope defined by the figure 14 (and reproduced on the figures 15 et 16 ) to determine whether the sample is healthy or has a tendency to spherization or rigidification.

[0068] Thus, by analyzing a sufficient portion of a sample (for example approximately 10,000 cells, the volume of blood analyzed being dependent on the counting conditions), it is possible to quickly return information of the type "healthy blood sample" or "abnormal blood sample", without using a neural network.

[0069] Alternatively, instead of the (low duration; PP) graph, one could use the (WR; PP) graph to establish the pulse zones allowing the definition of normality criteria.

[0070] There figure 17 represents the analysis of two healthy blood samples over a period of 11 days (duplicate run for each analysis). This figure shows that the distribution of pulse data sets on a WR / PP graph changes as a function of sample age.

[0071] For each analysis, the statistical distribution of the pulse data sets according to the WPP and WR / PP pairs is calculated, the statistics being established relative to a set of range of metrics presented on the Fig. 5 And 13 .

[0072] On the figure 18 , only the results relating to BOX3' are presented for the sake of brevity. This figure represents the statistical evolution over time of the pulses associated with BOX3' of the Fig.5 , depending on the age of the sample, the time origin being the day the sample was taken from the patient.

[0073] This figure shows that blood samples remain "normal" (do not change) for 7 days, then become "abnormal," in the sense that the statistical marker begins to drift sharply.

[0074] The invention can therefore be used to monitor the evolution over time of a blood sample. This figure is consistent with scientific publications in the field: when the sample ages, the red blood cells present in the sample see their biomechanical characteristics deteriorate. In particular, it is known that red blood cells exhibit a decrease in elasticity over time. The evolution over time of a sample can be affected by the storage duration, storage conditions, etc. In all cases, the figures 17 And 18 show that the arrangement according to the invention makes it possible to obtain information relating to the evolution over time of the sample, indicative for example of the validity of this sample.

[0075] Although the above refers primarily to the study of red blood cells, the invention would be applicable to any other type of cell whose morphological characteristics are likely to change, such as platelets for example.

Claims

1. A method for classifying a blood sample comprising the following operations: a) receiving pulse data sets, each pulse data set comprising impedance value data each time associated with a time marker, these data together representing a curve of cell impedance values measured when a cell passes through a polarised opening, the cells originating from the blood sample, b) for each pulse data set b1. determining a maximum impedance value of the pulse data set, the method being characterised in that it comprises the following additional operations: b2. calculating a high impedance value by multiplying the maximum impedance value by a high coefficient chosen in the range [0.7; 0.95], and by determining in the pulse data set the time markers whose associated impedance value in the pulse data set is equal to the high impedance value, and by calculating a high duration corresponding to the maximum duration between these time markers, and a low impedance value by multiplying the maximum impedance value by a low coefficient chosen in the range [0.1; 0.6], and by determining in the pulse data set the time markers whose associated impedance value in the pulse data set is equal to the low impedance value, and by calculating a low duration corresponding to the maximum duration between these time markers, b3. calculating a peak position value (PP) equal to the division of the difference between the instant associated with the maximum impedance value and the first instant that corresponds to the low impedance value, and the low duration, and optionally a rotation value (WR) equal to the division of the high duration by the low duration, c) determining the statistical distribution of the pulse data sets according to the low duration / peak position value (PP) pairs or optionally the rotation value / peak position value (WR) pairs, the statistics being established with respect to a set of pair value ranges (Box 1, Box 2, Box 3, Box 4, Box 5, Box 6, Box 7, Box 8, Box 1', Box 2', Box 3', Box 4', Box 5', Box 6'), d) classifying the blood sample by comparing the distribution of operation c) to a reference distribution of a sample of healthy blood.