Techniques for processing data and estimating blood pressure

Abstract

Disclosed is a method for processing data including at least one signal forming a plurality of patterns, for example a photoplethysmographic signal the patterns of which are representative of heartbeats of an individual, comprising a comparison of a plurality of said patterns to one another and a selection of patterns meeting at least one predetermined similarity criterion. Also disclosed is a method for training an artificial-intelligence model using training data formed from data processed by such a processing method. Also disclosed is a method for estimating a physiological parameter such as a blood-pressure value of one or more individuals, using a smart model trained with such a training method. Also disclosed is a device and computer program relating thereto.

Description

Data processing and blood pressure estimation techniques

[0001] The invention relates to the field of signal and data processing, particularly physiological data.

[0002] The invention is of particular, non-limiting interest for estimating blood pressure, in particular from a signal obtained by photoplethysmography or another type of cardiac signal. State of the prior art

[0003] Several techniques have been proposed in recent years to estimate the blood pressure of individuals from photoplethysmographic (PPG) signals, notably using artificial intelligence models.

[0004] The inventors found that known techniques produce frequent estimation errors and that the accuracy of the estimates depends heavily on individuals.

[0005] The invention aims to overcome all or part of the drawbacks of the techniques of the prior art.

[0006] According to a first aspect of the invention, the subject of the invention is a method for processing data, in particular physiological data, also called “data to be processed”.

[0007] Generally speaking, the processing method of the invention aims to constitute exploitable data from such data to be processed.

[0008] The data to be processed includes at least one signal that forms a plurality of patterns.

[0009] By way of non-limiting example, the at least one signal is a photoplethysmographic signal whose patterns are representative of an individual's heartbeats.

[0010] According to the invention, the processing method comprises: a comparison of several of said patterns with respect to each other, and a selection of the patterns which respect at least one predetermined similarity criterion.

[0011] In other words, the pattern(s) not meeting at least one similarity criterion are not selected and are therefore not used to form said exploitable data.

[0012] Such a selection makes it possible to exploit only patterns likely to contain useful information. Indeed, in a signal comprising patterns of which at least some are representative of the same phenomenon, in this example a heartbeat, a pattern which is not similar to patterns actually carrying useful information – the latter being in this example relating to a heartbeat – is probably a pattern not containing useful or exploitable information.

[0013] In one embodiment, the at least one similarity criterion is chosen from a list including: a correlation coefficient greater than a predetermined value, a deviation between a value of at least one pattern parameter and a predetermined value less than a predetermined deviation, the at least one parameter being able to be chosen from a list including a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time and a dicrotic notch time.

[0014] This list is of course not exhaustive and may include other similarity criteria, for example using a machine learning technique such as data partitioning.

[0015] In one embodiment, the processing method comprises: a construction of average patterns each from at least two respective of said selected patterns, a comparison of the average patterns with respect to each other, a selection of the average patterns if these respect at least one predetermined similarity criterion.

[0016] In a non-limiting manner, the at least one similarity criterion used to select average patterns can be chosen from the list indicated above, that is to say from a list including: a correlation coefficient greater than a predetermined value, a difference between a value of at least one pattern parameter and a predetermined value less than a predetermined difference, the at least one parameter being able to be chosen from a list including a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time and a dicrotic notch time.

[0017] In other words, the average pattern(s) not meeting at least one similarity criterion are not selected and are therefore not used to form said exploitable data.

[0018] In one embodiment, said plurality of patterns of the at least one signal are patterns selected from initial patterns meeting at least one predetermined quality criterion.

[0019] The initial pattern(s) not meeting the at least one quality criterion are not selected and therefore do not constitute said plurality of patterns subject to the aforementioned steps, in particular the selection based on the at least one similarity criterion.

[0020] In one embodiment, the at least one quality criterion is selected from a list including: a signal-to-noise ratio less than a predetermined value, a deviation between a value of at least one pattern parameter and a predetermined value less than a predetermined deviation, the at least one parameter being selectable from a list including a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time, a dicrotic notch time, an instantaneous heart rate, an instantaneous heart rate variation, a pattern amplitude and an amplitude difference between a pattern start time and a pattern end time.

[0021] This list is of course not exhaustive and may include other quality criteria, for example a criterion relating to similarity to a predetermined reference pattern.

[0022] In one embodiment, said at least one signal comprises several signals and said physiological data comprises one or more groupings each comprising several data sets, each of these data sets comprising a respective one of said signals as well as at least one blood pressure value.

[0023] Without limitation, the at least one blood pressure value may include both a systolic blood pressure value and a diastolic blood pressure value, or only one of these values.

[0024] In one embodiment, the processing method comprises selecting the one or more clusters in each of which the blood pressure values ​​of the datasets in that cluster are within a corresponding predetermined range of values ​​and / or have a variability less than a predetermined value.

[0025] In other words, the other group(s) are not selected and the signals they comprise therefore do not form said plurality of patterns which are the subject of the aforementioned steps, in particular the selection based on at least one similarity criterion.

[0026] In one embodiment, the processing method comprises an operation of collecting said data.

[0027] The collection operation preferably comprises, for each of one or more individuals, an acquisition of several groupings of data sets, each of the data sets comprising:a signal which forms a plurality of patterns, preferably a photoplethysmographic signal whose patterns are representative of heartbeats of this individual,at least one blood pressure value of this individual, for example a systolic blood pressure value and / or a diastolic blood pressure value.

[0028] It is preferred that the signals be acquired consecutively within each of the groupings.

[0029] The term "consecutive" means that the time interval between two successive data sets from the same grouping is relatively short compared to the time interval between two successive data sets each belonging to a different grouping. As an indication, the time interval between two signals acquired successively within the same grouping may be of the order of a few seconds or minutes, while the time interval between two signals acquired successively within two different groupings may be of the order of several hours.

[0030] According to a second aspect, the invention relates to a method for training an artificial intelligence model, also called “model” in this document.

[0031] The training of the model is preferably carried out using training data which is formed from data processed by a processing method as defined above.

[0032] Alternatively, the training data can be trained using any other method.

[0033] In one embodiment, the training data is formed from data collected using a collection operation similar to that described above.

[0034] The collected data may thus comprise, for each of one or more individuals, several groupings of data sets, each of the data sets comprising: a signal which forms a plurality of patterns, preferably a photoplethysmographic signal whose patterns are representative of heartbeats of this individual, at least one blood pressure value of this individual, for example a systolic blood pressure value and / or a diastolic blood pressure value.

[0035] It is preferred that the signals have been acquired consecutively within each of the groupings.

[0036] In one embodiment, the data is used such that:a first input of the model receives data obtained based on the signal and the at least one blood pressure value of one of the data sets of one of the individuals, this set being called the personalization set,a second input of the model receives data obtained based on the signal of another of said data sets of this individual, this other set being called the test set.

[0037] In one embodiment, during training, an output of the model is formed by a difference between the at least one blood pressure value of the test set and the at least one blood pressure value of the personalization set.

[0038] In one embodiment, a third input of the model receives data obtained from data in the test set and / or the personalization set.

[0039] In an alternative embodiment, the data provided to one or more of said model inputs comprises two-dimensional data which may for example be obtained by time-frequency transformation of one or more of said signals and / or patterns.

[0040] Such a training process allows the model to perform learning that inherently involves individual personalization.

[0041] A single model thus trained can then be used to correctly estimate a blood pressure value with input data from different individuals.

[0042] According to a third aspect, the invention relates to a method for estimating a physiological parameter such as a blood pressure value of one or more individuals, using an artificial intelligence model trained with a training method as defined above.

[0043] Of course, the various methods described above are preferably implemented by computer.

[0044] Furthermore, the various methods described above are independent of each other and can each be implemented independently, optionally in combination with other techniques not described herein.

[0045] The various methods of the invention can of course be combined with each other, the invention also having as its subject a method implementing a processing method as described above and / or a training method as described above and / or an estimation method as described above.

[0046] According to a fourth aspect, the invention relates to a device comprising a processing unit configured to implement a method according to any one of the aspects defined above, preferably to implement all of the methods described above.

[0047] In one embodiment, the device comprises at least one sensor for measuring at least one physiological signal and / or at least one physiological parameter.

[0048] For example, the at least one sensor may include a sensor for measuring a heartbeat signal, for example a piezoelectric sensor or a pressure sensor or a sensor comprising a camera capable of collecting a stream of images of an individual's face in order to generate a photoplethysmographic signal.

[0049] The at least one sensor may also include a blood pressure monitor type sensor, for measuring the blood pressure of an individual.

[0050] According to a fifth aspect, the invention relates to a computer program comprising executable computer instructions which, when executed by computer, implement a method according to any one of the aspects defined above, preferably several and more preferably all of the methods described above and steps which they comprise.

[0051] The computer program can be in any computer language, for example, machine language, C, C++, JAVA, Python, etc.

[0052] Other advantages and characteristics of the invention will appear on reading the detailed, non-limiting description which follows.

[0053] The following detailed description refers to the appended drawings in which:is a schematic view of a method according to the invention, comprising a phase of constructing training data, a phase of training an artificial intelligence model using the training data, a phase of constructing input data and a prediction phase in which the input data are provided to the trained model;is a schematic view of a data processing method according to the invention, this method being able to be implemented in the phase of constructing the method of the;is a schematic view of a device according to the invention, this device being able to be used to implement all or part of the method of the. Detailed description of embodiments

[0054] In the following non-limiting description, several innovative techniques are combined for the non-limiting purpose of estimating a blood pressure value of one or more individuals. These techniques can be implemented independently of each other, in applications and / or methods and / or devices different from those described herein.

[0055] Diagrammatically illustrates a method in accordance with the invention.

[0056] Without limitation, this method implements a conventional artificial intelligence model of the regressive neural network type, in this example a deep learning convolutional neural network.

[0057] Generally, the method comprises a phase 1 of building training data, a phase 2 of training the model using the training data, a phase 3 of building input data and a phase 4 of prediction in which the input data are provided to the trained model. These different phases and / or the steps and / or operations they contain may each constitute or be the subject of a respective method that can be implemented separately.

[0058] In the example of the, construction phase 1 comprises an operation 6 of data collection, an operation 7 of preprocessing the collected data, an operation 8 of processing the preprocessed data, and an operation 9 of evaluating the data thus processed.

[0059] Collection operation 6 comprises an acquisition of N series of data sets, called “collected data sets”, each of the series comprising data from a respective individual.

[0060] The number N of series, and therefore of individuals, is generally greater than one, preferably greater than two and more preferably greater than ten or one hundred. In this example, N is equal to 1000.

[0061] For each individual, the number M of data sets forming the series associated with this individual is generally greater than one, preferably greater than two and more preferably of the order of ten or one hundred.

[0062] In this example, each of the data sets in each of the series includes:a cardiac signal comprising patterns, also called pulses or beats, which each correspond to a respective heartbeat of the corresponding individual,a systolic blood pressure value of that individual,a diastolic blood pressure value of that individual.

[0063] For each of the data sets in each of the series, the cardiac signal may correspond to a signal having a predetermined duration T, for example a duration of 30 seconds, acquired for example using a photoplethysmography (PPG) technique, thus forming a photoplethysmographic signal.

[0064] In this example, for each of the data sets in each of the series, the systolic blood pressure and the diastolic blood pressure are measured using a conventional blood pressure monitor during the acquisition of the corresponding cardiac signal.

[0065] In the example of the, the collection 6 is carried out so as to acquire, for each of the individuals, a series of data sets forming several groupings which each comprise a number L of data sets acquired consecutively, that is to say by defining a time interval between two successive data sets of the same grouping which is relatively short. For information purposes, this intra-grouping time interval may be of the order of ten seconds between two signals. By comparison, the data of two successive groupings of the same series are acquired by defining a relatively long inter-grouping time interval, for example of the order of four hours between two signals.

[0066] In this example, for each of the groupings of each of the series of data sets, the number L of data sets of this grouping is generally greater than one, preferably greater than two, for example equal to three.

[0067] So, in this example, each of the dataset series includes a number of groupings equal to M / L.

[0068] As a guide, all data for each series can be acquired over an overall period of several weeks or months, for example five weeks.

[0069] The preprocessing operation 7 generally comprises one or more steps of preprocessing the cardiac signal of each of the collected data sets.

[0070] In this example, the preprocessing step(s) are chosen from a list including:an interpolation and resampling step,a filtering step, typically including windowing, for example using a Hamming window,a derivation step, for example by applying a second derivative.

[0071] Advantageously, the preprocessing operation 7 implements, for the cardiac signal of each of the collected data sets, the three steps listed above, preferably carried out in the order in which they are listed.

[0072] Such preprocessing makes it possible to remove noise from cardiac signals and more generally to clean them in order to improve the detection of useful information they contain.

[0073] The data sets containing the signals thus preprocessed during operation 7 are called “preprocessed data sets”.

[0074] The processing operation 8 comprises in this example a succession of steps of selection and / or cleaning of the preprocessed data sets, in a non-limiting manner steps 11, 12, 13, 14 and 15 described below and shown diagrammatically in.

[0075] Step 11 is performed based on the blood pressure values ​​of the datasets of each of the series.

[0076] Step 11 comprises, for each of the groupings of data sets in each of the series, a selection of the groupings in each of which the blood pressure values ​​of the corresponding data sets meet one or more blood pressure quality criteria.

[0077] In this example, a first blood pressure quality criterion relates to the absolute value of the blood pressure values ​​and a second blood pressure quality criterion relates to the variability of these values.

[0078] More specifically, in the context of this non-limiting example, step 11 comprises, for each of the groupings of data sets of each of the series, a selection of all the data sets of this grouping if:the systolic blood pressure value of at least one of the data sets of this grouping is in a predetermined range of values, for example in the range 80-180 mmHg, or / andthe diastolic blood pressure value of at least one of the data sets of this grouping is in a predetermined range of values, for example in the range 40-120 mmHg, or / andthe variability of the systolic blood pressure values ​​of the data sets of this grouping is less than a predetermined value, for example less than 5 mmHg, or / andthe variability of the diastolic blood pressure values ​​of the data sets of this grouping is less than a predetermined value, for example less than 5 mmHg.

[0079] Data sets from clusters not selected in step 11 are not used for further processing.

[0080] Step 12 is performed on the basis of the preprocessed signals, in particular the signals from the datasets selected in step 11.

[0081] In a non-limiting manner, step 12 comprises, for each of these signals, a selection of the data set containing this signal if the patterns of this signal are representative of a heart rate having a value in a predetermined range of values, for example in the range 40-180 beats per minute.

[0082] To make this selection, one or more other criteria may be used alternatively or in addition, for example a criterion based on a variation in instantaneous heart rate.

[0083] Datasets not selected in step 12 are not used in further processing.

[0084] Step 13 is performed based on the preprocessed signals belonging to the datasets that were selected in step 12.

[0085] Step 13 includes, for each of the patterns of each of these signals, a selection of this pattern if it meets one or more signal quality criteria.

[0086] In this example, a first signal quality criterion relates to a fluctuation and / or a signal-to-noise ratio of the portion of the signal comprising this pattern and a second signal quality criterion relates to one or more pattern parameters.

[0087] The pattern parameters may be selected from a list including, but not limited to, a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time, a dicrotic notch time, an instantaneous heart rate, an instantaneous heart rate variation, a pattern amplitude, and an amplitude difference between a pattern start time and a pattern end time.

[0088] More specifically, in the context of this non-limiting example, step 13 comprises, for each of the patterns of each of these signals, a selection of this pattern if: the signal-to-noise ratio of the part of the signal comprising this pattern is less than a predetermined value, for example less than 0.9 for a ratio that can be included in a range having a minimum value equal to 0 and a maximum value equal to 1, or / and a value of one or more of said parameters of this pattern has a deviation less than a predetermined deviation from a corresponding predetermined value, for example if the absolute difference in amplitude between the pattern start time and the pattern end time is greater than 0.4 for a maximum amplitude normalized to 1, or if the systolic peak time is greater than 0.4 times the difference between the pattern start time and the pattern end time.

[0089] Step 14 is performed on the basis of the preprocessed signals in which several patterns were selected in step 13. Data sets whose signal does not contain any pattern selected in step 13, or too few patterns are selected, for example less than two, are not used in further processing.

[0090] Step 14 includes, for each of these signals, a selection of patterns meeting one or more similarity criteria with respect to one or more other patterns of this signal.

[0091] Similarity criteria can be based on correlation and / or absolute value difference.

[0092] In this particular example, for each of the patterns of each of these signals, step 14 comprises a selection of this pattern if: the correlation coefficient of this pattern with one or more other patterns of this signal is greater than a predetermined coefficient, for example greater than 0.95, or / and the absolute difference between the value of at least one parameter of this pattern and the value of the same parameter of one or more other patterns of this signal is less than a corresponding predetermined value, for example a difference on the systolic peak time less than 7 ms.

[0093] Step 15 is performed on the basis of the preprocessed signals in which several patterns were selected in step 14. Data sets whose signal does not contain any pattern selected in step 14, or too few patterns are selected, for example less than four, are not used in further processing.

[0094] Generally, step 15 comprises, for each of these signals, processing this signal to form average patterns and comparing these average patterns so as to select this signal if the average patterns are sufficiently similar.

[0095] In this particular example, step 15 comprises, for each of these signals, a calculation of the number of patterns in several parts of this signal, these signal parts each being able to correspond to a fraction 1 / P of this signal, P preferably being an integer greater than or equal to two, for example equal to two. In this example, this signal is selected if the number of patterns in at least one of these parts is greater than a predetermined value, for example greater than one.

[0096] For each of the selected signals, step 15 implements signal processing consisting of normalizing the patterns of this signal and correcting their shape.

[0097] Without limitation, shape correction may consist of modifying the signal so that each pattern has an identical amplitude between the pattern start time and the pattern end time.

[0098] For each part of the signal thus processed, an average pattern is then calculated.

[0099] An average pattern may, for example, correspond to a pattern each of whose values ​​corresponds to the average of the corresponding values ​​of the patterns in that part. As a non-limiting variant, the values ​​of the patterns used to calculate the average may be weighted by a quality factor associated with the corresponding pattern, for example based on a similarity rate of that pattern compared to the other patterns in the signal.

[0100] Without limitation, this signal is selected here if the average patterns of this signal meet one or more similarity criteria. The similarity criteria can be based on a correlation and / or an absolute difference in values.

[0101] In this particular example, for each of the signals, step 15 comprises a selection of this signal if: the correlation coefficient of the average patterns of this signal is greater than a predetermined coefficient, for example greater than 0.95, or / and the absolute difference between the average patterns of this signal, according to at least one pattern parameter, is less than a predetermined value, for example a difference in systolic peak time less than 7 ms.

[0102] The pattern parameters for the average patterns may be the same as those used for the initial patterns and may therefore be chosen from a list including, but not limited to, a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time, a dicrotic notch time, and an amplitude difference between a pattern start time and a pattern end time.

[0103] The data sets containing a signal selected at the end of operation 15 are called “processed data sets”.

[0104] With reference to the, the evaluation operation 9 comprises a step of evaluating the quality of the processed data sets for each of the series, in order to determine whether the processed data of each of the individuals can be used for the training phase 2.

[0105] In this example, this step includes, for each of the processed data sets of each of the series, a calculation of a correlation between one or more parameters of the average patterns of the signal of this set – the average patterns being in this example calculated during step 15 of processing 8 – and the systolic and / or diastolic blood pressure value.

[0106] If the correlation thus calculated is greater than a predetermined coefficient, for example 0.9, for at least one of the parameters and for all the processed data sets of a given series, the data sets of this series are selected.

[0107] In this example, operation 9 also includes a step of evaluating the homogeneity of all the series of data sets thus selected.

[0108] This step typically allows one or more of said series to be discarded and / or one or more new series to be acquired, so as to obtain a set of data set series which is homogeneous according to one or more homogeneity criteria.

[0109] For information purposes, said homogeneity criteria may refer to a homogeneous inter-series distribution in terms of blood pressure values, and / or in terms of age and / or gender of the corresponding individuals.

[0110] Phase 1, which has just been described, thus makes it possible to construct data sets forming several series, each of which corresponds to a respective individual.

[0111] In the non-limiting example of the, the datasets thus constructed are used to train the model, in the manner described below. Alternatively, the datasets resulting from the implementation of phase 1 can be used for other purposes, for example for an analysis not implementing an artificial intelligence model.

[0112] With reference to the, phase 2 of the method is a phase of training the model using series of training data sets formed by the series of data sets constructed in phase 1 described above. Of course, the training data can be constructed in a different way, for example by performing a processing of collected data that includes only part of the selection steps described above and / or that includes one or more other processing steps.

[0113] In this example, training phase 2 comprises a step in which:one of the training data sets of one of the series is used to form so-called personalization data constituting a first input of the model, the personalization data comprising in this example on the one hand the average patterns of this set obtained during step 15 (see above) and, on the other hand, at least one of the blood pressure values ​​of this set,one of the other training data sets of this series is used to form so-called test data constituting a second input of the model, the test data comprising in this example the average patterns of this set obtained during step 15 (see above),an output of the model is formed by the difference between at least one of the blood pressure values ​​of the set forming the test data and at least one of the blood pressure values ​​of the personalization data.

[0114] This step is repeated iteratively using as test data each of the training datasets of the same series not forming the personalization data, this by iteratively using as personalization data each of the training datasets of the same series. These multiple combinations make it possible to improve learning.

[0115] This step and these iterations are then repeated in a similar manner for each of the other series.

[0116] Optionally, the learning may be enhanced by using a third input formed by at least one parameter of the average patterns of the set forming the test data and / or the set forming the personalization data.

[0117] The signals and patterns forming the model inputs can be one-dimensional, i.e. temporal in nature, or two-dimensional, for example by performing a prior time-frequency transformation of the corresponding signals or patterns.

[0118] In a manner known per se, training can be performed using a conventional training algorithm, for example a gradient backpropagation algorithm.

[0119] Regarding phase 3 of the method, this allows the input data of the model to be constituted during the prediction phase 4.

[0120] Generally speaking, phase 3 comprises, for at least one individual who is typically different from those whose data were used to train the model, here called “target individual”: an operation 21 of collecting one or more data sets intended to form so-called personalization data sets, and an operation 22 of collecting one or more signals intended to form so-called prediction signals.

[0121] In this example, for each target individual, operation 21 comprises a consecutive acquisition of three data sets analogous to the data sets of a grouping obtained during collection operation 6 of phase 1, the corresponding description applying by analogy.

[0122] Thus, in this example, for each target individual, each of the personalization datasets will be constructed based on a set comprising:a cardiac signal comprising patterns that each correspond to a respective heartbeat of the corresponding individual,a systolic blood pressure value of this individual,a diastolic blood pressure value of this individual.

[0123] The signals acquired during collection operation 22 are cardiac signals similar to the signals acquired during collection operation 6 of phase 1, the corresponding description applying by analogy.

[0124] Thus, in this example, for each target individual, the prediction signals are constructed based on photoplethysmographic signals each comprising patterns that each correspond to a respective heartbeat of that individual.

[0125] In the example of the, phase 3 of the method comprises a preprocessing operation 23 and a processing operation 24 of these different data which are analogous to operations 7 and 8 of phase 1, the corresponding description of which applies by analogy, a step such as step 11 of phase 1 being of course implemented only for the personalization data. These operations 23 and 24 thus make it possible to construct said personalization data sets and said prediction signals. In this example, phase 3 of the method does not of course implement steps analogous to those of operation 9 described above.

[0126] Of course, in the event that the data collected from a given target individual is not usable, phase 3 can be repeated in whole or in part in order to obtain usable input data.

[0127] During prediction phase 4, the prediction data of a given target individual are used as input data for the model: the personalization data sets constructed during phase 3 constitute a first input to the model, the personalization data comprising in this example on the one hand average patterns of this set obtained during step 24 and, on the other hand, at least one of the blood pressure values ​​of this set, the prediction signals each iteratively constitute a second input to the model, the prediction signals comprising in this example average patterns obtained during step 24.

[0128] In prediction phase 4, the model provides as output, for each target individual, a blood pressure value for each of the prediction signals.

[0129] In this example, for each target individual, a processing of the blood pressure values ​​provided by the model can be carried out by removing, for example, the outliers and calculating the average of the other values.

[0130] The schematically represents a device 30 making it possible to implement the method of the.

[0131] The device 30 comprises a processing unit 31 and two sensors 32 and 33.

[0132] In a non-limiting manner, the processing unit 31 is configured to implement each of phases 1 to 4 of the method.

[0133] The processing unit 31 can generally form a hardware module such as a processor, an electronic chip, a calculator, a computer or even a server, and a software module such as an application, a computer program or even a virtual machine.

[0134] The processing unit 31 can be integrated into a device such as a smartphone, a tablet or even a computer.

[0135] In this example, sensor 32 is a contact photoplethysmography sensor configured to measure a heartbeat signal by being placed in contact with the skin of an individual and sensor 33 is a blood pressure monitor for measuring the blood pressure of an individual.

[0136] Many variations may be made to the preceding description without departing from the scope of the invention. In particular, each of the phases of the method may be implemented without resorting to one or more of the other phases of this method, or by replacing one or more of these phases with techniques different from those described above.

[0137] Among other variants of the invention, the signals used may be signals acquired using a non-contact photoplethysmography (rPPG) technique or a different technique, for example a technique for acquiring a pulse signal using a pressure sensor. More generally, the data may be data of a periodic or quasi-periodic nature.

[0138] In an alternative embodiment, not shown, the construction of the data which is the subject of phase 1 of the method of la is carried out within the framework of a method without a collection operation, the data to be processed being able in this case to be available in a database. Of course, such a method can also be without one or more other steps and / or operations which are the subject of phase 1 of the method of la. By way of example, the invention thus covers a method comprising one or more of the steps described above with reference to la and not comprising operations 6, 7 and 9 of phase 1 of the method of la.

Claims

A method of processing (1) data including at least one signal which forms a plurality of patterns, for example a photoplethysmographic signal whose patterns are representative of an individual's heartbeats, the method comprising: a comparison of several of said patterns with respect to each other, a selection of the patterns which respect at least one predetermined similarity criterion, a construction of average patterns each from at least two respective of said selected patterns, a comparison of the average patterns with respect to each other, a selection of the average patterns if these respect at least one predetermined similarity criterion. Processing method (1) according to claim 1, wherein the at least one similarity criterion is chosen from a list including: a correlation coefficient greater than a predetermined value, a difference between a value of at least one pattern parameter and a predetermined value less than a predetermined difference, the at least one parameter being able to be chosen from a list including a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time and a dicrotic notch time. Processing method (1) according to claim 1 or 2, wherein said plurality of patterns of the at least one signal are patterns selected from initial patterns meeting at least one predetermined quality criterion, the at least one quality criterion being selectable from a list including: a signal-to-noise ratio less than a predetermined value, a deviation between a value of at least one pattern parameter and a predetermined value less than a predetermined deviation, the at least one parameter being selectable from a list including a peak time, a systolic peak time, a diastolic peak time, an interval between the systolic peak time and the diastolic peak time, a dicrotic notch time, an instantaneous heart rate, an instantaneous heart rate variation, a pattern amplitude and an amplitude difference between a pattern start time and a pattern end time. Processing method (1) according to any one of claims 1 to 3, wherein said at least one signal comprises several signals and wherein said physiological data comprises one or more groupings each comprising several data sets, each of these data sets comprising a respective one of said signals as well as at least one blood pressure value, the method comprising a selection of the grouping(s) in each of which the blood pressure values ​​of the data sets of this grouping are in a corresponding predetermined range of values ​​and / or have a variability less than a predetermined value. Processing method (1) according to any one of claims 1 to 4, comprising an operation of collecting (6) said data, the collecting operation (6) preferably comprising, for each of one or more individuals, an acquisition of several groupings of data sets, each of the data sets comprising: a signal which forms a plurality of patterns, preferably a photoplethysmographic signal whose patterns are representative of heartbeats of this individual, at least one blood pressure value of this individual, the signals being acquired consecutively within each of the groupings. A method (2) for training an artificial intelligence model using training data formed from data processed by a processing method (1) according to any one of claims 1 to 5, the training data preferably being formed from data collected according to a method comprising the characteristics of claim 5 and used such that:a first input of the model receives data obtained on the basis of the signal and the at least one blood pressure value of one of the data sets of one of the individuals, this set being called the personalization set,a second input of the model receives data obtained on the basis of the signal of another of said data sets of this individual, this other set being called the test set,an output of the model is formed by a difference between the at least one blood pressure value of the test set and the at least one blood pressure value of the personalization set. Method for estimating (4) a physiological parameter such as a blood pressure value of one or more individuals, using an artificial intelligence model trained with a training method (2) according to claim 6. Device (30) comprising a processing unit (31) configured to implement a method according to any one of claims 1 to 7 and, preferably, at least one sensor (32, 33) for measuring at least one physiological signal and / or at least one physiological parameter. A computer program comprising executable computer instructions which, when executed by a computer, implement a method according to any one of claims 1 to 7.

Images