Method and system for monitoring the health status of an individual

Abstract

The invention relates to a method for monitoring the health status of an individual (1), implemented by a monitoring system (2) comprising a measuring device (10) worn by the individual (1) and comprising at least one physiological sensor (11a-11b), at least one wave sensor (20), worn by the individual (1), configured to detect a pulse wave, and an analysis device (30) comprising a database associating reference evolution vectors with predetermined health states, the method comprising the steps of detecting a pulse wave at a detection time (t0), determining at least one evolution vector by comparing a first data vector prior to the detection time (t0) with a second data vector subsequent to the detection time (t0), and determining a health status of the individual (1) by consulting the database. Abstract figure: Figure 1

Description

Title of the invention: Method and system for monitoring the health status of an individual Technical field

[0001] The present invention relates to the field of monitoring the health status of an individual, in particular, of a soldier during an operation.

[0002] Today, there are already several systems for monitoring the health status of an individual in a combat zone. For example, one of these systems comprises, on the one hand, instrumented portable equipment (vest, armband, etc.) comprising a plurality of sensors configured to measure physiological data of the soldier and, on the other hand, an analysis terminal connected to said portable equipment in order to process the data from said sensors. For example, the portable equipment may comprise sensors for measuring the body temperature, heart rate or even the respiratory rate of the individual.

[0003] The analysis terminal is capable of receiving the various physiological measurements taken by all the sensors of the portable equipment. The analysis terminal makes it possible to carry out a diagnosis of the soldier's state of health, based on the measurements taken by the sensors of the vest. For example, the analysis terminal detects when a measured physiological parameter varies beyond predefined thresholds and makes it possible, for example, to detect an increase in the individual's temperature, heart rate or respiratory rate.

[0004] In a known manner, the analysis terminal includes a communication device to enable the diagnosis of the soldier's state of health to be communicated to a remote command center. This allows the command center to know the state of health of all the soldiers in a combat zone in order to make strategic decisions. For example, if the analysis terminal indicates that a soldier is injured, or that he is suffering from a malaise, the control center will decide to send a team of rescuers to help and treat said soldier.

[0005] However, the diagnosis made by the analysis terminal is not always reliable. Indeed, the variation of the parameters measured by the sensors can sometimes be associated with different health states of the individual. For example, the same profile of variations of measured parameters can be associated with either discomfort of the individual or a recovery of breath after a sprint. However, it is sometimes not possible to differentiate between the two possible states of the individual.

[0006] An objective of the present invention is to carry out a relevant diagnosis of the state of health of a soldier, in particular, in a combat zone.

[0007] The invention thus aims to eliminate at least some of these drawbacks. PRESENTATION OF THE INVENTION

[0008] The invention relates to a method for monitoring the health status of an individual, implemented by a monitoring system comprising: • a measuring device worn by the individual and comprising at least one physiological sensor, each physiological sensor being capable of measuring physiological data characterizing the individual over time, the measuring device being configured to measure at different times a data vector grouping together the physiological data measurements measured at a given time, • at least one wave sensor, worn by the individual, configured to detect a pulse wave, • a device for analyzing the individual's state of health, comprising a memory area in which a database is recorded, the database associating reference evolution vectors with predetermined states of health, • the method comprising the steps of: • Detect a pulse wave at a detection instant, • Determine at least one evolution vector by comparing a first data vector measured at a first instant, prior to the detection instant, to a second data vector measured at a second instant, subsequent to the detection instant, • Determine the individual's health status by consulting the database based on the determined evolution vector.

[0009] Physiological data includes, for example, heart rate, respiratory rate, and temperature. A pulse wave is, in particular, a wave generated following an explosion or a gunshot.

[0010] The method thus makes it possible to determine the state of health of a soldier reliably, by comparison between the physiological data characterizing the individual before the detection of the pulse wave and the physiological data characterizing the individual after the detection of the pulse wave. It is thus possible to determine the impact of a pulse wave on the physiological data. A state of health is thus only determined under certain conditions, which makes it possible to limit the risk of false positives.

[0011] When this monitoring system is used by soldiers during an operation, it allows the operations command center to obtain a relevant health status of each soldier following an explosion and thus decide on the most effective strategy. adapted to the situation and the health of each soldier. In particular, if a soldier is injured, the command center can trigger a rescue or first aid operation for said soldier. This monitoring system also allows the command center to quantify the explosion to which the soldiers have been subjected in order to adapt the tactical system.

[0012] Preferably, the first instant belongs to a first time window [to-dla; to] in which the duration dla is greater than or equal to 10 seconds. Such a time window makes it possible to determine in a relevant manner the physiological data before the explosion.

[0013] More preferably, the first instant belongs to a first time window [t0-dla; t0-dlb] in which the duration dlb is greater than or equal to 0 seconds.

[0014] More preferably, the second instant belongs to a second time window [t0; t0+d2b] in which the duration d2b is less than or equal to 10 seconds. This makes it possible to quickly carry out a diagnosis following the detection of a wave.

[0015] Preferably still, the second instant belongs to a second time window [t0+d2a; t0+d2b] in which the duration d2a is greater than or equal to 5 seconds. Such a time window makes it possible to exclude all physiological data in the vicinity of the detection time which could lead to false positives. This makes it possible to ensure that the consequences of the explosion are visible on the physiological data of the soldier. Thus, in the event of a hemorrhagic injury at the detection time, the physiological data will only deteriorate after the duration t0+d2a and not immediately after the detection time.

[0016] Preferably, the step of determining at least one evolution vector consists of determining one or more abnormal variations by comparing the first data vector to the second data vector. Preferably, at least one abnormal variation is determined by a computer learning method, in particular, by a classification method. By carrying out learning, an abnormal variation can be detected in a practical and rapid manner. Advantageously, the larger the dimension of the vectors, the greater the determination of abnormal variations.

[0017] According to a preferred aspect, the monitoring system comprising at least one physical sensor, each physical sensor being capable of measuring physical data over time, the measuring device being configured to measure at different times a data vector grouping the measurements of at least one physiological data item and at least one physical data item at a given time. The measurement of physical data has the first advantage of increasing the dimension of the measurement vector, which improves the determination of abnormal variations. In addition, the use of data Physical measurements allow physiological data to be corrected / transformed in such a way as to facilitate the detection of abnormal variation. An external temperature is relevant, for example, to interpret an individual's body temperature.

[0018] Preferably, the physical sensor is chosen from an acceleration sensor, an individual position sensor, a sound sensor, an outside temperature sensor, a visible or infrared imaging sensor, a weather data sensor, or a radio wave sensor.

[0019] The invention also relates to a system for monitoring the health status of an individual, comprising: • a measuring device worn by the individual and comprising at least one physiological sensor, each physiological sensor being capable of measuring physiological data characterizing the individual over time, the measuring device being configured to measure at different times a data vector grouping together the physiological data measurements measured at a given time, • at least one wave sensor, worn by the individual, configured to detect a pulse wave, • a device for analyzing the individual's state of health, comprising a memory area in which a database is recorded, the database associating reference evolution vectors with predetermined states of health, the analysis device being configured to: • Determine at least one evolution vector by comparing a first data vector measured at a first instant, prior to the detection instant, to a second data vector measured at a second instant, subsequent to the detection instant, • Determine the individual's health status by consulting the database based on the determined evolution vector.

[0020] The invention also relates to a computer program type product, comprising at least one sequence of instructions stored and readable by a processor and which, once read by this processor, causes the steps of the method as presented previously to be carried out.

[0021] The invention further relates to a computer-readable medium comprising the computer program product as presented above. PRESENTATION OF FIGURES

[0022] The invention will be better understood on reading the description which follows, given by way of example, and referring to the following figures, given by way of non-limiting examples, in which identical references are given to objects similar.

[0023] [Fig.l] is a schematic representation of a system for monitoring the health status of an individual according to one embodiment of the invention,

[0024] [Fig.2] is a schematic representation of the variation of a representation of physiological data measured by a sensor of the measuring device of the system according to [Fig.l],

[0025] [Fig.3] is a simplified schematic representation of the variation of a representation of two physiological data and one physical data measured by sensors of the measuring device of the system according to [Fig.l].

[0026] It should be noted that the figures set out the invention in detail for implementing the invention, said figures can of course be used to better define the invention where appropriate. DETAILED DESCRIPTION OF THE INVENTION

[0027] The invention relates to a system 2 for monitoring the state of health of an individual 1, and in particular of a soldier during an operation, in a combat zone. It can also be used by emergency personnel.

[0028] With reference to [Fig.l], the monitoring system 2 comprises a measuring device 10, a wave sensor 20 and an analysis device 30. The monitoring system 2 is worn by the individual in order to measure certain of his physical and physiological data.

[0029] The measuring device 10 is worn by the individual and in this example comprises three sensors 1 la-1 le but it goes without saying that the number of sensors could be different.

[0030] In this example, the sensors 11a-11b are physiological sensors capable of measuring physiological data Da-Db characterizing the individual over time. For example, the sensor 11a measures the heart rate, noted "fc", the sensor 11b measures the respiratory rate, noted "fr". It goes without saying that the physiological sensors could be more numerous, in particular, to measure the body temperature of the individual.

[0031] In this example, the sensor 11c is a physical sensor capable of measuring physical data De. For example, the sensor 11c measures the acceleration of the individual Acc. It goes without saying that the physical sensors could be more numerous and of various types, for example, a position sensor of the individual, a sound sensor, an outside temperature sensor, a visible or infrared imaging sensor, or even a meteorological data sensor.

[0032] Such physiological and physical sensors are known per se to those skilled in the art and will not be presented in further detail.

[0033] With reference to [Fig.l], all of the sensors are carried by a single measuring device 10 but it goes without saying that they could be distributed over several measuring devices 10 carried by one or more individuals. The measuring device 10 can for example be mounted on a bra and / or on a vest and / or on an armband, put on by the individual.

[0034] More precisely, the measuring device 10 is configured to measure at different times a data vector VE grouping the physical / physiological data measurements Da, Db, De measured at a given time. According to the example presented here, each data vector VE comprises the physiological data Da, Db and the physical data De, in other words the data fc, fr, Acc measured at the same time t.

[0035] The data thus measured are identified in order to define the measuring device 10 which determined them. The data vectors VE are also sent to the analysis device 30.

[0036] The wave sensor 20 is also worn by the individual and is capable of detecting a pulse wave. More specifically, since the wave sensor 20 is worn by the user, the wave sensor 20 makes it possible to detect a pulse wave experienced by the user. The pulse wave may be caused by an explosion or a gunshot, for example.

[0037] More precisely, the wave sensor 20 can be mounted on the measuring device 10 and / or be an independent element and physically separated from the measuring device 10, it can for example be positioned on another soldier of the same team. Such a wave sensor 20 is known per se. Preferably, the monitoring system 2 comprises at least three wave sensors 20.

[0038] With reference to [Fig.l], the analysis device 30 firstly comprises a memory area in which a database is recorded, associating reference evolution vectors of physical / physiological data with predetermined health states, each reference evolution vector comprising a set of variations of physical / physiological data. The physical / physiological data characterized in the reference evolution vector correspond to those measured by the sensor(s) 11a, 11b, 11c of the measuring device 10. The database is produced by learning or obtained by feedback.

[0039] For example, by "health states" are designated in particular the following health states: "malaise", or "catching one's breath" or "death" or "injured" or "seriously injured". Each of these health states is therefore associated with a set of variations in physical and physiological data. For example, the state "malaise" is associated with a reference evolution vector comprising an abnormal variation in heart rate, an abnormal variation in respiratory rate and an abnormal variation in body temperature. For example, a variation in a set of physical / physiological data vectors is considered abnormal compared to nominal data which correspond to activities not related to a deterioration in health status.

[0040] Abnormal variations are determined by learning using a computer, in particular, by means of a classifier. Preferably, abnormal variations are determined by an analysis of the density of a set, for example in a k-nearest neighbor (k-NN) analysis, local outlier factor or isolation forest analysis. Abnormal variations may also correspond to an exclusion of nominal classes or associated with an abnormal class with support vector machine classes, autoencoders or recurrent neural networks. A nominal variation may also be determined by data partitioning or by ensemble learning.

[0041] The analysis device 30 is in the form of mobile equipment, such as a tablet or a smartphone for example, placed near at least one individual wearing a measuring device 10. Preferably, the analysis device 30 is worn by the soldier. The analysis device comprises a computing unit, in particular a processor, in order to practically determine a state of health. Preferably, the analysis device 30 comprises a communication module, in particular encrypted, for communicating the state of health to a remote command center.

[0042] The analysis device 30 is connected to each measuring device 10 and to each wave sensor 20 by a wired or wireless link. Thus, the analysis device 30 is configured to recover the physical / physiological data vectors measured by each measuring device 10 and the data measured by each wave sensor 20.

[0043] Thus, the analysis device 30 is configured to receive, on a regular basis, a vector of physical / physiological data, each vector being associated with a measuring device 10 and associated with a given instant. The analysis device 30 is also capable of receiving an alert of detection of a pulse wave sent by the wave sensor 20 at a given detection instant as will be presented later.

[0044] When the analysis device 30 receives an alert of detection of a pulse wave concerning an individual, the analysis device 30 is configured to define at least one first vector and at least one second vector.

[0045] With reference to [Fig. 2], an example of the variation of a physiological datum Da is shown, in this case a representation of the deviation of the heart rate from a nominal heart rate as a function of time t. The instant t0 represents the instant at which a pulse wave was detected by the wave sensor 20.

[0046] A first vector corresponds to a vector measured at a first instant tl, prior to the detection instant t0. More precisely, the first instant tl belongs to a first time window fl defined between [t0-dla; t0] in which the duration dla is equal to 10 seconds. Even more precisely, the first instant tl belongs to a first time window [to-dla; to-dlb] in which the duration dlb is equal to 0 seconds.

[0047] Similarly, a second vector corresponds to a vector measured at a second time t2, subsequent to the detection time t0. More precisely, the second time t2 belongs to a second time window f2 defined between [t0; t0+d2b] in which the duration d2b is equal to 10 seconds. Even more precisely, the second time t2 belongs to a second time window [t0+d2a; t0+d2b] in which the duration d2a is equal to 5 seconds. Such a second time window f2 makes it possible to characterize the physical / physiological data of the soldier after the explosion without being disturbed by the explosion itself. Preferably, the duration d2a makes it possible to ensure that the possible consequences of the explosion are visible on the physical / physiological data. Thus, in the event of injury, the drop in temperature occurs after a certain delay and is not measurable directly after the explosion.The d2b duration is low to allow a health status to be determined reactively following the detection of a wave.

[0048] The analysis device 30 is configured to compare at least one first data vector measured at a first instant t1, prior to the detection instant t0, with at least one second vector measured at a second instant t2, subsequent to the detection instant t0, in order to obtain an evolution vector.

[0049] For example, still with reference to [Fig.2], in the case where the analysis device 30 compares a single first vector with a single second vector, the analysis device 30 is configured to determine, for each physical / physiological data item characterized by the first and second vectors, an abnormal variation Aa between the first instant t1 and the second instant t2.

[0050] According to a variant, when the analysis device 30 compares several first vectors to several second vectors ([Fig.3]), the analysis device 30 determines abnormal variations Aa, Ab, Ac, according to several dimensions, which improves the determination of a diagnosis.

[0051] In other words, the evolution vector comprises all the determined difference values, for example a value representing a rise or a loss of temperature and a value representing an increase or decrease in heart rate and a value of decrease or increase in acceleration. It goes without saying that the addition of other physiological / physical data makes it possible to increase the dimension of the evolution vector. This makes it possible to refine the determination of variations abnormal by learning.

[0052] The analysis device 30 is also configured to consult the database and determine the state of health of the individual, by selecting, in the database, the state of health whose reference evolution vector corresponds to the determined evolution vector. In other words, the state of health which is closest to the detected abnormal variations is determined.

[0053] The use of physiological data and physical data makes it possible to reliably characterize abnormal variations following a pulse wave. For example, a meteorological data sensor makes it possible to characterize a drop in body temperature as a function, in particular, of the outside temperature. An image and acceleration sensor makes it possible to confirm the variation in the physical movement of the individual.

[0054] With reference to [Fig. 3], an embodiment of the method for monitoring the health status of an individual 1 will be presented, implemented by a system 2 for monitoring the health status of an individual 1 as presented previously. A method will be presented in the case where a single individual 1 wears a measuring device 10 and a wave sensor 20. The method is then repeated for each individual 1 wearing a measuring device 10 and a wave sensor 20.

[0055] In the example presented here, it is considered that the measuring device 10 comprises three sensors 1 la-1 le and more precisely a heart rate sensor fc, a respiratory rate sensor fr and an acceleration sensor Acc.

[0056] The method firstly comprises a step of measurement at different times and by each sensor 1 la-1 le of the measuring device 10, and following, a step of determining a data vector for each time. The vectors thus measured therefore represent a model of the value of each physical / physiological data measured by a sensor 1 la-1 le of the measuring device 10 at the same measurement time, are represented with reference to [Fig.3].

[0057] [Fig. 3] represents a plurality of variations of data relating to said individual 1 as a function of time. The variations are represented in a simplified manner for the sake of clarity. The first variation represents the heart rate fc, the second variation represents the respiratory rate fr and the third variation represents the acceleration Acc.

[0058] Each vector thus measured and identified as corresponding to said measuring device 10 (and therefore to said individual 1), is then sent to the analysis device 30. Thus, the analysis device 10 receives both physical data and physiological data.

[0059] In addition, upon detection of a pulse wave by the wave sensor 20, the method comprises a step of sending an alert by the wave sensor 20 to the device. analysis 30 indicating the time of detection.

[0060] When the analysis device 30 receives said alert, the method comprises a step of comparing at least one first vector with at least one second vector in order to obtain an evolution vector, the vectors being chosen in the time windows determined from the detection instant t0.

[0061] For example, with reference to [Fig.3], each first vector comprises a heart rate value fc, a respiratory rate value fr and an acceleration value Acc, measured at the same time defined in the first time window fl. Each second vector comprises a heart rate value fc, a respiratory rate value fr and an acceleration value Acc measured at a time defined in the second time window f2.

[0062] According to a first embodiment, a single first vector is compared with a single second vector so as to deduce therefrom an evolution vector. The evolution vector comprises any abnormal variation of physiological / physical data in order to characterize a state of health. As presented previously, an abnormal variation is preferably determined by learning and / or classification.

[0063] In this example, the evolution vector includes: • any abnormal variation between the value of the heart rate fc of the first vector measured at the first instant tl and the value of the heart rate fc of the second vector measured at the second instant t2, • any abnormal variation between the value of the respiratory frequency fr of the first vector measured at the first instant tl and the value of the respiratory frequency fr of the second vector measured at the second instant t2, • any abnormal variation between the acceleration value Acc of the first vector measured at the first instant tl and the acceleration value Acc of the second vector measured at the second instant t2.

[0064] According to a second embodiment, several first vectors, each measured at a first instant specific to it, and several second vectors, each measured at a second instant specific to it, are compared. The use of several first vectors and several second vectors makes it possible to increase the reliability of the diagnosis of the state of health. Preferably, the data vectors are measured continuously.

[0065] Finally, the method comprises a step of consulting the database and determining the state of health of the individual 1. This determination is carried out by selecting, in the database, the state of health whose reference evolution vector corresponds to the determined evolution vector.

[0066] More precisely, each abnormal variation present in the evolution vector is compared to the abnormal variations of each reference evolution vector in stored in the memory area so as to determine the closest reference evolution vector and deduce the soldier's state of health. Advantageously, a classification and / or learning method makes it possible to determine the soldier's state of health from the evolution vector.

[0067] The method makes it possible to reliably determine the state of health of each individual 1 following an explosion or a firearm shot, by correlating data from different sensors relating to said individual 1. Thus, in the case of the use of this system by soldiers in operation, the command center reliably knows the situation of each individual 1 on the ground, in particular their position and their state of health, in order to determine the best way to react and act to accompany, rescue or protect the soldiers in operation.

Claims

Claims

1. Method for monitoring the state of health of an individual (1), implemented by a monitoring system (2) comprising: • a measuring device (10) worn by the individual (1) and comprising at least one physiological sensor (11a-11b), each physiological sensor (11a-11b) being capable of measuring physiological data (Da-Db) characterizing the individual (1) over time, the measuring device (10) being configured to measure at different times a data vector grouping the physiological data measurements measured at a given time, • at least one wave sensor (20), worn by the individual (1), configured to detect a pulse wave, • a device (30) for analyzing the state of health of the individual, comprising a memory area in which a database is recorded, the database associating reference evolution vectors (VE) with predetermined health states,• the method comprising the steps of: • Detecting a pulse wave at a detection time (t0), • Determining at least one evolution vector by comparing a first data vector measured at a first time (tl), prior to the detection time (t0), with a second data vector measured at a second time (t2), subsequent to the detection time (t0), the second time (t2) belonging to a second time window [t0+d2a; t0+d2b] in which the duration d2a is greater than or equal to 5 seconds and the duration d2b is less than or equal to 10 seconds.,

2. Method according to claim 1, in which the first instant (tl) belongs to a first time window [to-dla; to] in which the duration dla is greater than or equal to 10 seconds.

3. Method according to claim 2, in which the first instant (tl) belongs to a first time window [tO-dla; tO-dlb] in which the duration dlb is greater than or equal to 0 seconds.

4. Method according to one of claims 1 to 3, in which the step of determining at least one evolution vector consists of determining one or more abnormal variations by comparing the first data vector to the second data vector.

5. Method according to claim 4, wherein at least one abnormal variation is determined by a computer learning method, in particular, by a classification method.

6. Method according to any one of claims 1 to 5, in which the monitoring system comprises at least one physical sensor (11c), each physical sensor (11c) being capable of measuring physical data (De) over time, the measuring device (10) being configured to measure at different times a data vector grouping together the measurements of at least one physiological data item and at least one physical data item at a given time.

7. The method of claim 6 wherein the physical sensor is selected from an acceleration sensor, an individual position sensor, a sound sensor, an outdoor temperature sensor, a visible or infrared imaging sensor, a weather data sensor, or a radio wave sensor.

8. System (2) for monitoring the state of health of an individual (1), comprising: • a measuring device (10) worn by the individual (1) and comprising at least one physiological sensor (11a-11b), each physiological sensor (11a-11b) being capable of measuring physiological data (Da-Db) characterizing the individual (1) over time, the measuring device (10) being configured to measure at different times a data vector grouping the physiological data measurements measured at a given time, • at least one wave sensor (20), worn by the individual (1), configured to detect a pulse wave, • a device (30) for analyzing the state of health of the individual, comprising a memory area in which a database is recorded, the database associating reference evolution vectors (VE) with predetermined health states, the analyzing device (30) being configured to: Determine at least one evolution vector by comparing a first data vector measured at a first instant (tl), prior to the detection instant (tO), to a second data vector measured at a second instant (t2), subsequent to the detection instant (tO), the second instant (t2) belonging to a second time window [t0+d2a; t0+d2b] in which the duration d2a is greater than or equal to 5 seconds and the duration d2b is less than or equal to 10 seconds. Determine the individual's health status (1) by consulting the database from the determined evolution vector.

Images