Device and method for monitoring a subject's posture

Abstract

Device and Method for Posture Tracking of a Subject. The invention relates to a method and a processing device (T1) for performing posture tracking of a subject (UR1). The method comprises: obtaining sensor data (DT1) representative of the subject (UR1); generating, from the sensor data, a model (ML1) defining anatomical points (PT1) of the subject over time; determining, from the model (ML1), kinematic variables (V1) representative of the evolution of the subject's posture; determining, from the kinematic variables (V1), at least one nonlinear variability indicator (V2) representative of a temporal structure of the kinematic variables (V1); analyzing (S10) said at least one nonlinear variability indicator (V2) by comparison with a respective reference value (REF2); and generating a posture command (CMD1) based on a result of the analysis. Figure for the abstract: Fig. 1.

Description

Title of the invention: Device and method for tracking the posture of a subject technical field

[0001] The invention relates to the field of postural analysis and monitoring of a subject and more particularly to a device and processing method for monitoring a subject's posture, notably for the purpose of evaluating and / or correcting the subject's posture. Prior art

[0002] It is now generally accepted that movement is highly beneficial for people, not only physically but also psychologically, cognitively, immunologically, etc., and this is true at any age. Physical activity, and movement more generally, therefore constitutes a major public health issue and now occupies a central place in supporting individuals, particularly for athletes and people with specific needs, or for the management of injuries, illnesses, post-operative follow-up, etc.

[0003] In the field of sport, both professional and amateur, injury is a major concern for athletes and participants. There is therefore a significant need today for comprehensive sports care to prevent injuries (preparation, rehabilitation, etc.), assess their physical and psychological condition, and optimize their performance. Injury prevention and management are key challenges in the monitoring and care of athletes at all levels.

[0004] In addition to supporting athletes, movement is a key element in supporting individuals, particularly in a preventive or curative context (for example, after an injury). Healthcare services are therefore increasingly geared towards an active approach to health. Many public programs aim to promote physical activity in all its forms (sport, play, work, etc.).

[0005] Monitoring a person's posture, that is, how they stand or move in an environment, can provide a wealth of information about their health or condition (for example, in the case of injury, pathology, etc.) or about their functioning (for example, their performance, habits, etc.). There is therefore a growing need today, particularly among professionals but also among the general public, to assess the quality of people's movement in various contexts, notably for the purposes of assisting with physical activity, play, performance monitoring, rehabilitation, or athletic reconditioning. In this context, movement analysis techniques have seen significant growth in recent years, but current solutions still have limitations and constraints. which remain an obstacle to the development of this sector.

[0006] Thus, current posture (or movement) analysis techniques generally do not offer satisfactory results, particularly in terms of reliability and accuracy. For example, applications currently available for smartphones for posture tracking do not allow for a detailed analysis of posture, whether static or dynamic. These solutions often introduce errors in the positioning of the subject's anatomical parts and are generally limited to comparing captured movements to reference models without being able to assess the quality of the movement in all its complexity.

[0007] In addition to limitations in terms of accuracy and reliability, some current solutions require complex, expensive, and / or time-consuming equipment. For example, some current techniques require the deployment of a significant number of sensors, or the placement of reflective targets or markers positioned on the target subject itself. The complexity of installing, adjusting, and using the necessary equipment can then prove problematic, or even prohibitive, in practice. For instance, placing markers on the subject can be tedious and carries the risk of leading to measurement errors in the event of unexpected marker movement. The economic, human, and / or time costs of implementing these techniques, as well as the necessary user training, limit the deployment of these solutions. Description of the invention

[0008] One of the objects of the present invention is to solve at least one of the problems or deficiencies of the technological background described above.

[0009] Another object of the present invention is to enable efficient monitoring of a subject's posture, particularly in terms of accuracy and reliability, with a view to optimizing the health and / or performance of individuals, for example for the purposes of assisting with physical activity, physical preparation, play, performance monitoring, rehabilitation and / or re-athleticization.

[0010] Another object of the present invention is to enable a precise and reliable analysis of the posture, static or dynamic, of a subject, with limited cost and complexity of implementation.

[0011] Another object of the present invention is to help effectively in correcting the posture of a subject.

[0012] To this end, a first aspect of the invention relates to a processing method (or tracking method), implemented by a processing device, for tracking the posture of a subject, said method comprising: a) obtaining sensor data representative of the subject; b) generation, from sensor data, of a model defining anatomical points of the subject over time; c) determination, from the model, of kinematic variables representative of the evolution of the subject's posture over time; d) determination, from the kinematic variables, of at least one non-linear indicator of variability representative of a temporal structure of the kinematic variables; (e) analysis of said at least one non-linear indicator of variability by comparison with a respective reference value; and f) generation of a posture instruction based on an analysis result.

[0013] The method according to the invention may include other features which may be taken separately or in combination, in particular among the following embodiments which are presented by way of illustration only and may be combined or associated unless otherwise stipulated.

[0014] According to a particular example, the sensor data obtained in a) includes image data captured by a plurality of image capture devices, the image data of each image capture device defining a respective 2D view of the subject over time; generation b) comprising: - estimation of the position of the subject's anatomical points over time in the 2D view from each image acquisition device; and - generation of the 3D model by triangulation of the estimated positions of the anatomical points in each 2D view.

[0015] According to a particular example, generation b) comprises: - attribution, to each estimated position of the subject's anatomical points in the 2D views, of a respective confidence score representing a level of confidence in said estimated position; the model being generated from the estimated positions of anatomical points from which are excluded, by a first filtering, each estimated position of an anatomical point whose confidence score does not reach at least a first threshold value.

[0016] According to a particular example, generation b) of the model comprises: - comparison of first and second confidence scores associated respectively with first and second positions estimated for the same anatomical point in 2D views from image acquisition devices; the model being generated from the estimated positions of the anatomical points to which a second filtering is applied according to which, if the difference between said first and second confidence scores reaches at least a second threshold value, The one among the first and second estimated positions with the lowest confidence score is excluded.

[0017] According to a particular example, said at least one nonlinear indicator determined in d) comprises a first sequence of values ​​over time; the analysis e) comprising: - comparison of the first sequence of values ​​of said at least one non-linear indicator over time with respectively a second sequence of reference values; the result of analysis e) being a function of a result of said comparison of the first and second sequences.

[0018] According to a particular example, generation f) comprises at least one of: - the provision, as the first posture instruction, via a user interface, of posture information specifying how to correct the subject's posture; and - sending, as a second posture instruction, a command causing a setting of a movement assistance device to correct the subject's posture.

[0019] According to a particular example, analysis e) comprises: - verification of whether said at least one non-linear indicator of variability meets a conformity criterion with respect to the respective reference value; in which, if the conformity criterion is not met, the posture instruction generated in f) is configured to correct the subject's posture.

[0020] According to a particular example, if a non-linear variability indicator – referred to as a non-conforming indicator – does not meet the conformity criterion, analysis e) includes: - determination of at least one kinematic variable, called the target kinematic variable, associated with the non-compliant indicator and of a correction to be applied to said at least one target kinematic variable to improve the non-compliant indicator relative to the reference value according to the compliance criterion; said posture instruction being generated in f) as a function of the correction to be applied to said at least one target kinematic variable to correct the posture of the subject.

[0021] According to a particular example, the determination of said at least one target kinematic variable and the correction to be applied is carried out by an optimization method taking as input: - a non-compliant indicator trend target; - a function defining a relationship between the non-conforming indicator and a plurality of associated kinematic variables; and - constraint data defining motion constraints that the model must respect; the optimization method delivering as output setpoint data defining said at least one target kinematic variable and the correction to be applied to modify the non-compliant indicator according to the trend objective; the posture instruction being generated in f) from the instruction data.

[0022] According to a particular example, the process includes, prior to calculation c): - recognition, from the model, of a movement performed by the subject; and - selection, based on the recognized movement, of the kinematic variables to be calculated in c).

[0023] In a particular embodiment, the different steps of the process according to the first aspect of the invention are determined by computer program instructions.

[0024] Consequently, a second aspect of the invention relates to a computer program on an information carrier (or recording medium), this program being capable of being implemented in a processing device, or more generally in a computer, this program comprising instructions adapted to the implementation of the steps of the processing method according to the first aspect of the invention. In particular, the second aspect of the invention concerns a computer program comprising instructions for executing the steps of the process according to the first aspect of the invention when said program is executed by a computer.

[0025] Thus, the method of the invention can be implemented by means of a non-volatile memory storing computer program instructions and by means of a processor executing these instructions.

[0026] This program may use any programming language, and be in the form of source code, object code, or code intermediate between source code and object code, such as in a partially compiled form, or in any other desirable form.

[0027] A third aspect of the invention relates to a computer-readable information medium (or recording medium) and comprising computer program instructions according to the second aspect of the invention.

[0028] The information medium can be any entity or device capable of storing the program. For example, the medium can include a storage means, such as a rewritable non-volatile memory or ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording means, for example a hard drive or a memory in the form of a USB key or other.

[0029] On the other hand, the information medium can be a transmissible medium such as an electrical or optical signal, which can be transmitted via an electrical or optical cable, by radio, or by other means. The program according to the invention can, in particular, be downloaded onto an Internet-type network.

[0030] Alternatively, the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the process in question.

[0031] A fourth aspect of the invention relates to a processing device (or tracking device) configured to implement the method according to the first aspect of the invention. The fourth aspect of the invention relates in particular to a processing device configured to perform posture tracking of a subject, said device comprising a memory associated with a processor configured to implement the steps of the processing method according to the first aspect of the present invention.

[0032] According to one example, the invention according to the fourth aspect provides a processing device to assist a subject in performing a posture, said device comprising: - acquisition module configured to obtain sensor data representative of the subject; - first generation module configured to generate, from sensor data, a model defining anatomical points of the subject over time; - first determination module configured to calculate, from the model, kinematic variables representative of the evolution of the subject's posture over time; - second determination module configured to determine, from the kinematic variables, at least one non-linear indicator of variability representative of a temporal structure of the kinematic variables; - an analysis module configured to analyze said at least one non-linear indicator of variability by comparison with a respective reference value; and - second generation module configured to generate a posture instruction based on an analysis result.

[0033] It should be noted that the various embodiments mentioned above (as well as those described below) in relation to the treatment process of the invention and the associated advantages apply in a similar way to the treatment device of the invention.

[0034] In particular, for each step of the treatment process, the treatment device of the invention may include a corresponding module configured to carry out said step.

[0035] According to one embodiment, the invention is implemented by means of software and / or hardware components. In this context, the term "module" may refer in this document to a software component, a hardware component, or a set of hardware and software components.

[0036] A software component corresponds to one or more computer programs, one or Several subroutines of a program, or more generally any element of a program or software capable of implementing a function or set of functions, as described below for the module in question. Such a software component can be executed by a data processor of a physical entity (terminal, server, gateway, router, etc.) and is capable of accessing the hardware resources of that physical entity (memory, storage media, communication buses, input / output electronic cards, user interfaces, etc.).

[0037] Similarly, a hardware component corresponds to any element of a hardware assembly capable of implementing a function or a set of functions, as described below for the module concerned. It may be a programmable hardware component or one with an integrated processor for software execution, for example an integrated circuit, an electronic board, etc.

[0038] The present invention, as defined above and described below in specific examples, advantageously allows for the precise and reliable monitoring of a subject's posture. In particular, it is possible to assess the quality of a subject's static or dynamic posture over time and to guide or improve this posture by generating a postural instruction adapted to the subject's behavior. Such monitoring makes it possible, in particular, to assess and improve the subject's health and / or performance for various purposes, such as, for example, to assist with physical activity, physical preparation, play, physical rehabilitation, and / or athletic reconditioning. Furthermore, the invention allows for precise and reliable monitoring of a subject's static or dynamic posture with limited cost and complexity of implementation. The present invention can also advantageously provide diagnostic assistance.

[0039] In particular, the analysis of at least one non-linear variability indicator determined from kinematic variables advantageously allows for a detailed evaluation of a subject's movement quality in order to take into account complex aspects not perceptible on the sole basis of kinematic variables. Depending on the result of this analysis, a postural instruction can advantageously be generated, for example, to adapt, improve, and / or correct the subject's posture and thus help improve the subject's health and / or performance.

[0040] The invention therefore offers a powerful tool, both for movement professionals such as health and / or sports professionals (doctors, therapists, physiotherapists, physical trainers, coaches / sports trainers, etc.) and for the general public, to analyze and improve the posture of a subject. Brief description of the drawings

[0041] Other features and advantages of the present invention will become apparent from the description of the particular and non-limiting embodiments of the present invention below, with reference to the attached Figures 1 to 18, in which:

[0042] [Fig-1] Fig.1 schematically represents a cooperating processing device with at least one associated device, according to at least one embodiment of the invention;

[0043] [Fig.2] Fig.2 represents an example of implementation of the processing device of Fig.1 according to at least one embodiment of the invention;

[0044] [Fig.3] The [Fig.3] schematically represents the processing device (or monitoring device) of the [Fig.1], according to at least one embodiment of the invention;

[0045] [Fig.4] The [Fig.4] represents, in the form of a diagram, the steps of a treatment process (or monitoring process), according to at least one embodiment of the invention;

[0046] [Fig.5] The [Fig.5] schematically represents the acquisition of sensor data (image data) during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0047] [Fig.6] Fig.6 schematically represents the acquisition of sensor data (inertial data) during the processing of Fig.4, according to at least one embodiment of the invention;

[0048] [Fig.7] The [Fig.7] schematically represents the acquisition of sensor data (EMG data) during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0049] [Fig.8] The [Fig.8] schematically represents the generation of a model defining anatomical points during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0050] [Fig.9] The [Fig.9] schematically represents the generation of a model defining anatomical points during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0051] [Fig. 10] The [Fig. 10] schematically represents the calculation of kinematic variables during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0052] [Fig. 11] The [Fig. 11] schematically represents the calculation of kinematic variables during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0053] [Fig. 12] The [Fig. 12] schematically illustrates different degrees of posture complexity that can be represented by a non-linear indicator of variability determined during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0054] [Fig. 13] Fig. 13 schematically illustrates different degrees of complexity of postures likely to be represented by a non-linear indicator of variability determined during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0055] [Fig. 14] The [Fig. 14] schematically illustrates different degrees of posture complexity that can be represented by a non-linear indicator of variability determined during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0056] [Fig. 15] The [Fig. 15] schematically represents the determination of at least one non-linear variability indicator (entropic approximation indicator) during the processing of the [Fig.4], according to at least one embodiment of the invention;

[0057] [Fig. 16] The [Fig. 16] schematically represents the determination of at least one non-linear variability indicator (GEM type indicator for "Goal Equivalent Manifold" in English) during the treatment process of the [Fig.4], according to at least one embodiment of the invention;

[0058] [Fig. 17] [Fig. 17] schematically represents the generation of a posture instruction during the processing of [Fig. 4], according to at least one embodiment of the invention; and

[0059] [Fig. 18] The [Fig. 18] schematically represents the generation of a posture instruction during the processing of the [Fig.4], according to at least one embodiment of the invention. Description of implementation methods

[0060] Examples of implementations of the invention will now be described in the following with joint reference to Figures 1-18. Unless otherwise indicated, common or similar elements in several figures bear the same reference symbols and have identical or similar characteristics, so that these common elements are generally not described again for the sake of simplicity.

[0061] The terms "first(s)" (or first(s)), "second(s)", etc.) are used in this document by arbitrary convention to allow identification and distinction of different elements (such as operations, threshold values, etc.) implemented in the embodiments described below.

[0062] The invention proposes to track and adapt the posture of a subject. In this context, a posture refers, in this disclosure, to a positioning or movement of the subject in question, or more precisely, of all or part of the subject's body. In other words, the posture being tracked can be a static posture (body position at a given moment) or a dynamic posture (body movement). Generally, depending on the accuracy of the motion sensor(s) used, a posture may therefore appear static or dynamic in certain cases. A posture can To provide information on the position of one or more anatomical points of a subject at a given moment or over time. Insofar as a given individual is normally subject to variations, even minute ones, in the positioning of the different parts of their body due at least to natural or involuntary movements (muscle tension, respiration, tremors, joint repositioning, etc.), a posture can generally be considered dynamic over time provided that the sensor data used is sufficiently precise to represent the movements in question.

[0063] The quality of a subject's posture can be influenced by various factors, including physical, psychological, and environmental ones. It has been found that a good understanding and therefore good monitoring of posture is desirable to effectively guide a subject in performing a posture, movement, or physical activity.

[0064] In this disclosure, the subject of posture monitoring may be any person who may have various physical conditions or states (child, adult, elderly person, injured person, person with a disability, etc.), or more generally any living being, including an animal, such as a dog or a horse for example.

[0065] The invention relates in particular to a processing method (or monitoring method), and a corresponding processing device (or monitoring device), for monitoring the posture of a subject. The method is based in particular on determining kinematic variables from a model defining anatomical points of the subject, and on analyzing at least one non-linear variability indicator determined from the kinematic variables. A posture instruction can thus be generated based on a result of this analysis, in order to adapt, correct, and / or improve the subject's posture.

[0066] The invention, according to its various embodiments, thus implements a processing method, implemented by a processing device, for postural tracking of a subject, said method comprising: a) obtaining sensor data representative of the subject; b) generation, from sensor data, of a model defining anatomical points of the subject over time; c) determination, from the model, of kinematic variables representative of the evolution of the subject's posture over time; d) determination, from the kinematic variables, of at least one non-linear indicator of variability representative of a temporal structure of the kinematic variables; e) analysis of said at least one non-linear indicator of variability by comparison with a respective reference value; and f) generation of a posture instruction based on an analysis result.

[0067] The invention, according to its various embodiments, also relates to a computer program, and an information medium containing such a program, for the execution of the steps of the processing method according to the invention.

[0068] Other aspects and advantages of the present invention will become apparent from the embodiments described below with reference to the drawings mentioned above.

[0069] Figure 1 schematically represents a processing device T1 (also called a tracking device or device) capable of cooperating with at least one associated device 20, also called a control device, to perform posture tracking of a subject denoted UR1 of the invention according to particular embodiments. The processing device T1 and said at least one control device 20 together form a processing system (or tracking system) denoted SY1.

[0070] For the sake of example, it is assumed that the associated device(s) 20 are separate from the processing device T1, although variants are possible in which the associated device(s) are part of the processing device TL

[0071] By way of example, it is hereafter assumed that the processing device T1, also referred to as the device, is configured to perform posture tracking of the user UR1 in a context that may vary depending on the case. The subject UR1 may be, for example, an athlete, an injured or convalescent person, or more generally any person whose posture (or movement) one wishes to assess and improve, for purposes such as physical maintenance, health improvement, injury prevention, play and / or performance.

[0072] As illustrated in [Fig. 1], the processing device T1 can use at least one posture sensor (or motion sensor) 12 to obtain DTI sensor data representative of the subject UR1, or more precisely, of the PSI posture of all or part of the subject URL. The sensor(s) 12 may be separate from the device T1 or part of it, as appropriate. The nature of the DTI sensor data retrieved by the processing device T1 may vary, depending in particular on the type of sensor 12 used. Examples of such sensors, denoted 12a, 12b, and 12c, are described below.

[0073] According to one example, the processing device Tl is configured to receive, from at least one image acquisition device 12a, image data DTla representative of the subject's PSI posture URL

[0074] According to one example, the processing device Tl is configured to receive, from at least one inertial sensor 12b positioned on the subject UR1, inertial data DTlb representative of the subject's posture PSI URL

[0075] According to one example, the processing device Tl is configured to receive, from at least one 12c EMG type sensor (for "electromyogram") positioned on the subject UR1, EMG DTlc data representative of the PSI posture of the subject UR1.

[0076] The processing device Tl can thus be configured to receive, as DTI sensor data, at least one of the image data DTla, inertial data DTlb, and EMG data DTlc as previously described, or any combination of at least two of these data types. The configuration and use of these sensors 12a-12c, as well as the associated sensor data, are described in more detail in the following description of the tracking method according to various embodiments. The processing device Tl can be configured to receive and process in real time all or part of the DTI sensor data received from the sensor(s) 12.

[0077] According to one example, the processing device Tl is configured to obtain all or part of the DTI sensor data by accessing a memory (for example, memory 6) in which the data is stored. For example, it is possible to record DTI sensor data and then configure the processing device Tl to retrieve and process this data later.

[0078] In what follows, it is assumed that the DTI sensor data defines a PSI posture of the subject UR1 over time, which makes it possible in particular to analyze a movement or an evolution of the PSI posture over time, although variants are possible where the DTI sensor data defines the PSI posture of the user UR1 only at a given instant, for example to track the spatial positioning (or disposition) of the subject at a given instant.

[0079] As illustrated, the processing device T1 in this example comprises at least one processor 4 and a memory 6. This memory 6 may include various types of memory, including volatile memory (such as RAM) and non-volatile memory. The non-volatile memory may include read-only memory (such as ROM) and / or rewritable non-volatile memory. This memory 6 may include, in particular, an operating system 10 executable by the processor 4 to operate the processing device TL

[0080] The memory 6 constitutes a recording medium (or information medium) according to particular embodiments, readable by the processing device T1, and on which a computer program PG1 conforming to various particular embodiments is stored. This computer program PG1 includes instructions for executing the steps of the monitoring method of the invention according to particular embodiments. The steps of this method are represented, in a particular embodiment, in [Fig. 4] described later.

[0081] Thus, the processor 4 is configured to execute the instructions of the computer program PG1 in order to carry out steps of the tracking method of the invention according to specific implementation examples. To this end, the processor 4 may include integrated memory, an input / output interface, and various circuits known to those skilled in the art. In particular, the processor 4 may use volatile memory internal to the processing device T1 (this memory may be part of the memory 6 or be a separate memory) to perform the various operations and functions necessary for the operation of the processing device T1, including executing the computer program PG1 during the implementation of the processing method of the invention.

[0082] As shown in [Fig. 1], memory 6 is capable of storing various data that may be used during the execution of the tracking process. Thus, memory 6 can store a model ML1 defining anatomical points PT1 of the subject UR1. Positions PN1 can be predicted (or estimated) and stored for each anatomical point PT1 of the model ML1, possibly in association with respective SCI confidence scores. Memory 6 can also be used to store kinematic variables VI, one or more nonlinear variability indicators V2, and a reference value REF2. According to one particular example, memory 6 can also be used to store a linear variability indicator V3 and a reference value REF3. The nature and use of the aforementioned data are described in more detail later in specific examples.

[0083] The processing device Tl may have various forms depending on the case, and may in particular be a computer, a server, a smartphone, a tablet, or more generally a processing device comprising the means configured to carry out the steps of the monitoring process of the invention.

[0084] As shown in [Fig. 1], the processing device T1 is further configured to cooperate with, or control, at least one control device 20. In particular, the processing device T1 is capable of sending a posture command CMD1 to one or more control devices 20. The number and type of control devices 20 used may vary. A posture command CMD1 is a command (or instruction) intended to control a control device 20 in order to guide or assist the subject UR1 in performing a posture, movement, or physical activity. For example, this posture command CMD1 allows the subject UR1's PSI posture to be adapted, corrected, and / or improved through the control device(s) 20.

[0085] The type of each posture instruction CMD1 depends in particular on the type of control device 20 concerned and the objective sought. Examples of such control devices, denoted 20a, 20b, 20c and 20d ([Fig. 1]), are described below.

[0086] According to one example, the processing device Tl is configured to send a first posture instruction CMD1 to a user interface 20a-20c to cause the restitution of posture information IF1 specifying how to adapt, correct and / or improve the PSI posture of subject UR1. As illustrated, this user interface 20 can take various forms, the way in which the IF1 information is returned being a function of the type of user interface used.

[0087] According to one example, the processing device Tl is configured to send at least one first posture instruction, more precisely denoted CMDla, to a display device 20a to cause the display of posture information IF1 (in visual form). According to another example, the processing device Tl can send at least one first posture instruction CMDla to a sound device 20b to cause the emission of posture information IF1 in audible form. According to another example, the processing device Tl can send at least one first posture instruction CMDla to a light device 20c to cause the emission of posture information IF1 in visual form (for example, by activating an indicator light or by changing the color of an emitted light). Several of the user interfaces 20 described can also be used in combination.

[0088] By way of example, a control device 20 cooperating with the processing device Tl can be a smartphone, tablet, computer or equivalent, including a display device 20a as indicated above.

[0089] According to one example, the processing device Tl is configured to send at least one second posture instruction, more precisely denoted CMDlb, to a control device 20, here taking the form of a movement assistance device 20d (also called a fitness device). This second posture instruction CMDlb then causes a parameterization (or configuration) of the movement assistance device 20d to adapt, correct, and / or improve the PSI posture of the subject UR1.

[0090] The assistance device 20d can be any device configured, under the control of the processing device Tl, to facilitate, guide, and / or intensify the performance of postures, movements, and / or physical activities by a subject or individual. For example, the assistance device 20d can be or include a weight training machine, a treadmill, an elliptical trainer, an exercise bike, a rowing machine, a multi-gym station, etc. This type of device aims, for example, at improving cardiovascular fitness, muscle strengthening, weight loss, developing proprioception, or more generally at physical preparation or training or rehabilitation.

[0091] The assistance device 20d can be more or less complex depending on the case. In particular, it can be configured according to at least one operating parameter (for example, a belt speed for a treadmill, a resistance level for an exercise bike or elliptical trainer, the duration of a training program, etc.). The processing device Tl can thus control the configuration configuration of such an assistance device 20d by sending a second posture instruction CMDlb.

[0092] According to one example, the posture sensor(s) 12 used to obtain DTI sensor data are part of one or more control devices 20.

[0093] By way of example, the processing device Tl cooperates with at least one sensor 12 and at least one control device 20, these elements Tl, 12 and 20 all being part of (or forming) the same device or system. By way of example, a terminal such as a smartphone or other device may comprise the processing device Tl, at least one sensor 12 and a display device 20a as previously described.

[0094] Figure 2 represents an example of an embodiment of the SY1 system as described with reference to Figure 1. In the example of Figure 2, a plurality of cameras 12a are arranged around the subject UR1 to capture its posture PSI over time. In this case, the processing device T1 obtains or retrieves image data DT1 comprising images representing the subject UR1 over time.

[0095] The processing device Tl can thus be configured to send at least one of a first posture instruction CMDla and a second posture instruction CMDlb, or any combination of the different examples of posture instruction envisaged.

[0096] According to one example, the movement assistance device 20d may include at least one of the user interfaces 20a-20c. In this case, a first posture instruction CMDla and a second posture instruction CMDlb may be sent by the processing device Tl to cause both the restitution of information IF1 by the assistance device 20d and the configuration of said assistance device 20d.

[0097] The processing device Tl can be configured to generate and send one or more CMD1 posture commands in real time to control the control device(s) 20. Alternatively, the CMD1 posture command(s) can be generated and sent during a phase subsequent to that of capturing the subject's posture URL

[0098] It should be noted that the processing device Tl shown in [Fig. 1] is only one example of an embodiment; other implementations are possible within the scope of the invention. Those skilled in the art will also understand that certain elements of the processing device Tl are described here only to facilitate understanding of the invention, as implementations of the invention are possible without these elements.

[0099] As shown in [Fig. 3] according to a particular embodiment, the processor 4 controlled by the computer program PG1 ([Fig. 1]) implements a number of modules, namely: an acquisition module MD2, a first generation module MD4, a first determination module MD6, a second determination module MD8 mination, an MD10 analysis module and a second MD12 generation module.

[0100] More specifically, the MD2 acquisition module can be configured to obtain DTI sensor data representative of the UR1 subject.

[0101] The first MD4 generation module can be configured to generate, from DTI sensor data, an ML1 model defining anatomical points PT1 of the subject UR1 over time.

[0102] The first MD6 determination module can be configured to calculate, from the ML1 model, kinematic variables VI representative of the evolution of the PSI posture of the subject UR1 over time.

[0103] The second determination module MD8 can be configured to determine, from the kinematic variables VI, at least one nonlinear variability indicator V2 representative of a time structure of the kinematic variables VL

[0104] The analysis module MD10 can be configured to analyze said at least one nonlinear variability indicator V2 by comparison with a respective reference value REF2.

[0105] The second generation module MD12 can be configured to generate a posture instruction CMD1 based on a result of the analysis performed by the analysis module MD 10.

[0106] The configuration and operation of the MD2-MD12 modules of the Tl device will appear more precisely in the embodiment examples described below.

[0107] The monitoring process implemented by the processing device Tl as previously described with reference to Figures 1-2 is now described according to particular embodiment examples together with Figures 3-18. To do this, the device Tl performs steps S2-S12 ([Fig.4]) by executing the computer program PG1.

[0108] It is assumed by way of example that the processing device Tl is used to monitor and improve the PSI posture of the subject UR1, the latter being able to stand on any surface, or possibly, he may use a movement assistance device 20 as appropriate.

[0109] During a retrieval step S2, the device T1 obtains DTI sensor data representative of the subject URL. This DTI sensor data can be generated by various types of sensors 12 capable of capturing the posture PSI of the subject UR1 over time. The DTI sensor data can be received from one or more posture sensors 12 or retrieved from a memory, for example memory 6 ([Fig. 1]).

[0110] According to one example, the DTI sensor data received by the processing device T1 includes image data (or video data) DT1 representative of the subject URL. As an example, [Fig. 5] illustrates the acquisition S2 of image data DTla represents a subject UR1 performing a squatting movement, although other postures are possible. The image data DTla may include images IG1 representing all or part of the subject UR1 over time. This image data DTla may be generated by at least one image sensor 12a, i.e., at least one image acquisition device such as a camera or equivalent. The processing device Tl may obtain (receive, determine, etc.) this image data DTla in various ways, depending on the case.

[0111] Obtaining DTla image data during step S2 is advantageous in that it allows us to deduce an accurate or faithful representation of the PSI posture of all or part of the subject's body URL From DTla image data, we can generate a model of what is visible in the image, which provides a more complete representation than that provided by other types of sensor, such as those described below.

[0112] The use of a plurality of cameras 12a (as for example in [Fig. 2]) advantageously allows the PSI posture of all or part of the subject UR1 to be captured in space over time. This makes it possible to obtain a detailed visual representation and minimize the risk of occluding an area of ​​interest on the subject's body URL. By using multiple cameras 12a positioned appropriately in space, images (or video) can be acquired capturing the part(s) of the subject UR1 that one wishes to track and analyze, such as, for example, a joint (knee, ankle, etc.) or a limb (lower or upper limb). In particular, a plurality of cameras 12a can be used to perform a 3D acquisition of the subject URL.

[0113] According to an example illustrated in [Fig. 6], the DTI sensor data received by the processing device T1 includes inertial data DTlb representative of the subject UR1 over time. To this end, one or more inertial sensors 12b ([Fig. 1]) can be positioned on the subject UR1 (on its body or clothing) in order to acquire this inertial data, which is then received and processed by the processing device TL. This inertial data DTlb can include, for example, acceleration data 30 acquired by at least one accelerometer 12b and / or gyroscopic data 32 acquired by at least one gyroscope 12b. In other words, the inertial data DTlb is, for example, acquired by means of at least one of an accelerometer-type sensor 12b and a gyroscope-type sensor 12b, said at least one sensor being positioned on the subject UR1.

[0114] As illustrated in [Fig. 6], at least one accelerometer 12b can be used to generate acceleration data 30, for example, linear acceleration data representative of acceleration along one or more axes (e.g., along two axes X and Y, or along three axes X, Y, and Z) as a function of time t. It is possible to use three accelerometers 12b to measure the acceleration in the space of Anatomical points PT1 of subject UR1 over time. Similarly, at least one gyroscope 12b can be used to generate gyroscopic data 32, for example angular velocity data representative of an angular velocity (e.g. in degrees per second) as a function of time t.

[0115] According to one example, the DTI sensor data received by the processing device Tl includes EMG DTlc data (electromyogram data) representative of subject UR1 over time. More specifically, this DTlc data defines the EMG activity of subject UR1 over time. To achieve this, one or more EMG sensors 12c can be positioned on subject UR1 to acquire this EMG DTlc data, which is then received and processed by the processing device Tl.

[0116] Electromyography (EMG) allows the measurement of electrical activity in the muscles of the subject UR1. This electrical activity can be captured by means of electrodes (EMG 20c sensors) placed in contact with the subject, for example, surface electrodes placed on the skin or needle electrodes inserted into the subject's muscle tissue. The EMG DTlc data thus collected can, in particular, be used to analyze the electrical response of muscles to nerve stimulation, providing relevant information on the PSI posture or the condition of the subject UR1, for the monitoring and management of various physical or medical conditions.

[0117] According to one example, the DTI sensor data obtained (S2, [Fig.4]) by the processing device Tl includes at least one of the DT la, DTlb and DTlc sensor data described above, or any combination of at least two of these types of sensor data.

[0118] According to one example, inertial DTlb data and / or EMG DTlc data are obtained in S2 ([Fig. 4]) in addition to DTla image data in order to further improve the capture of the subject's PSI posture. The DTlb-DTlc sensor data produced by the 12b-12c inertial and EMG sensors can be useful for posture analysis but, unlike DTla image data, provide a fragmented representation of the subject's state in certain localized areas. It may be necessary to position multiple 12b and / or 12c sensors on the subject to obtain a sufficiently complete representation of the body area to be analyzed.

[0119] It is subsequently considered, by way of example, that DTla image data are obtained during the acquisition step S2 by means of a plurality of image capture devices 12a (for example cameras), this DTla data being used subsequently during the tracking process as described below.

[0120] During a generation step S4 ([Fig.4]), the processing device T1 generates, from the DTI sensor data obtained in S2, a model ML1 ending PT1 anatomical points (or parts) of the UR1 subject over time. Figures 8 and 9 represent examples of ML1 models generated during the S4 generation step.

[0121] According to one example, the image data DTla produced by each camera 12a include IG1 images defining respective 2D views (or representations) of the subject UR1 over time. Thus, during generation S4 ([Fig. 4]), the processing device Tl estimates (or predicts) the position PN1 of anatomical points PT1 of the subject UR1 over time in the 2D views (IG1 image) from each camera 12a. Each position prediction constitutes a hypothesis of the position at which a respective anatomical point PT1 is located. A 2D model (for example, taking the form of a 2D skeleton as illustrated in Figures 8-9) can thus be generated from each IG1 image produced by a camera 12a. The Tl device can then generate a 3D ML1 model by triangulating the estimated PN1 positions of the anatomical points PT1 in each 2D view (each IG1 image). In other words, a 3D model is constructed from the 2D captures of each camera 12a.This 3D model allows for the precise representation of the PSI posture of subject UR1 in space.

[0122] Alternatively, it is possible to generate a two-dimensional, or even one-dimensional, ML1 model to represent the URL subject. A 1D or 2D model allows the UR1 subject to be modeled in only one or two dimensions, which may be sufficient and advantageous in some cases, for example because of the more limited resource requirement than for the generation of a 3D model.

[0123] The PT1 anatomical points used in S4 to construct the ML1 model can be adapted as needed, depending on the posture to be analyzed and the objective sought. These PT1 anatomical points can, for example, define joints (elbow, knee, ankle, etc.) and / or characteristic points of the subject's body (for example, facial features or different parts of a limb).

[0124] The prediction of the PN1 positions of anatomical points PT1 can be performed by executing a prediction algorithm based on the IG1 images defined in the DTla image data. This algorithm can use an artificial intelligence-based prediction model to accurately estimate the PN1 position of each anatomical point PT1 of interest.

[0125] The prediction of the PN1 positions may also take into account the relative position of the cameras 12a with respect to each other. A calibration step may be necessary prior to the S2-S4 acquisition steps to ensure satisfactory accuracy in the prediction of the PN1 positions, and therefore the generation of a high-quality ML1 model.

[0126] Thus, in S4 ([Fig.4]), a 3D ML1 model can be constructed, for example, in the form of a 3D skeleton representing the subject's PSI posture. as shown in figures 8-9. This ML1 model can be used to model the evolution of the respective PN1 position of the anatomical points PT1 over time.

[0127] According to one example, the device T1 assigns, to each estimated position PN1 of the anatomical points PT1 of the subject UR1 in the 2D views, a respective SCI confidence score that is representative of a level of confidence in said estimated position PN1. In other words, a respective SCI confidence score is associated with each estimate of position PN1. The model ML1 is then generated (S4, [Fig. 4]) taking into account the SCI confidence scores. In particular, the model ML1 can be generated (S4) from the estimated positions PN1 of the anatomical points PT1, among which are excluded, by a first filtering, each estimated position PN1 of an anatomical point PT1 whose SCI confidence score does not reach at least a first threshold value SL1. In other words, this filtering based on SCI confidence scores makes it possible to generate a good quality ML1 model using only the estimated PN1 positions whose confidence score reaches at least the first threshold value SL1.This SL1 value can be adapted on a case-by-case basis. By avoiding the use of estimated PN1 positions, whose reliability is considered insufficient, a precise and reliable ML1 model can be advantageously generated.

[0128] It may be preferable to reach a compromise by limiting the number of anatomical points in the ML1 model to retain only the most reliable points to ensure good model quality. If an anatomical point PT1 has an insufficient confidence score in all 2D views, for example because it is not visible to any camera, this point is then absent (excluded) from the final 3D model (partial model).

[0129] It can indeed be difficult for a camera 12a to capture all parts of the subject corresponding to the PT1 anatomical points of interest. In some cases, troublesome reflection problems or limited image quality can make it difficult to detect a PT1 anatomical point from an IG1 image. The greater the number of cameras 12a used, the lower the risk of occlusion, but the greater the resource cost and implementation complexity. Predicting PN1 positions advantageously allows the construction of a high-quality ML1 model regardless of the subject's posture UR1, including in cases where PT1 anatomical parts of the subject UR1 are occluded or difficult to identify in some IGL images. However, the quality of position predictions varies depending on several factors.Depending on the number and position of the 12a cameras, some PT1 anatomical parts may not be visible in the 2D views produced by the 12a cameras, which can lead to more or less reliable estimates of certain anatomical points. Using SCI confidence scores as previously described allows for the generation of a complete and high-quality ML1 model.

[0130] According to one example, to generate the ML1 model (S4, [Fig. 4]), the device T1 may retain, for a given anatomical point PT1 at a given time, only a single position estimate PN1 from the 2D view of a respective camera 12a, i.e., the best estimate among those available, in order to further improve the quality of the ML1 model. For example, if a first camera 12a visualizes an anatomical point PT1 better than a second camera 12a at a given time (in the case of occultation, for example), it may be preferable to retain only the estimated position PN1 obtained from the image IG1 of the first camera 12a.

[0131] According to one example, during generation S4 ([Fig. 4]), the processing device T1 compares first and second SCI confidence scores associated respectively with first and second estimated positions PN1 for the same anatomical point PT1 in 2D views from cameras 12a. The ML1 model is then generated from the estimated positions PN1 of the anatomical points PT1, to which a second filtering is applied. If the difference between said first and second SCI confidence scores reaches at least a second threshold value SL2, the first and second estimated positions PN1 with the lowest confidence score is excluded (by being taken into account to generate the ML1 model). In other words, only the estimated position PN1 with the highest SCI confidence score is used to generate the ML1 model. This second filtering can be repeated over time to improve the quality of the ML1 model.

[0132] The respective SCI confidence scores obtained for PN1 position estimates of the same anatomical point PT1 (representing, for example, a knee or other point) in several 2D views, or in several 2D skeletons from several cameras 12a, can thus be compared, and the estimated PN1 position with the lowest confidence score can be eliminated, but only in cases of significant disparity between the respective SCI confidence scores, which may indicate a significant risk of error. In this way, the number of PN1 position estimates taken into account during S4 generation can be maximized, thereby further improving the quality of the ML1 model. The second threshold value SL1 can, for example, be defined in the model of the prediction algorithm used to generate the ML1 model.

[0133] During a determination or calculation step S6 ([Fig. 4]), the processing device T1 determines (or calculates), from the ML1 model generated in S4, kinematic variables (or data) VI representative of the evolution of the subject UR1's posture PSI over time. The number and type of kinematic variables VI determined in S6 can be adapted as needed. In general, the kinematic variables VI characterize the dynamics of the subject UR1's posture PSI over time.

[0134] According to one example, the kinematic variables VI determined in S6 include at least one of the velocity and acceleration data calculated from the estimated positions PN1 of the anatomical points PT1 of the ML1 model over time. In other words, the kinematic variables VI can characterize velocities and / or accelerations of points in the ML1 model, for example, anatomical points PT1. These may be angular and / or translational velocities and accelerations, as appropriate.

[0135] A kinematic variable V1 can thus define, for example, a velocity component or an acceleration component along a given direction, for example along an X, Y or Z axis.

[0136] According to one example, the kinematic variables VI determined in S6 include at least one of the following variables, or a combination of at least two of these variables: a position, an angle, a velocity, and an acceleration. In other words, the kinematic variables V1 can characterize positions, angles, velocities, and / or accelerations of points in the model ML1, for example, anatomical points PT1. These may be, for example, angular or translational positions, as appropriate, or angular and / or translational velocities and accelerations, as appropriate.

[0137] According to one example, at least some of the kinematic variables VI determined in S6 are DTI sensor data obtained in S2. In other words, it is possible to determine a kinematic variable VI from DTI sensor data without any intermediate processing, using the ML1 model generated in S4. Alternatively, a kinematic variable V1 is determined in S6 by processing the DTI sensor data defined in the ML1 model.

[0138] By way of illustration, [Fig. 10] schematically represents kinematic variables VI calculated in S6 over time, these variables representing a displacement (in meters) of a center of mass of the ML1 model in space. In this example, the three curves represent translational motion components of the center of mass in 3 dimensions, namely along the X, Y, and Z axes (lateral, vertical, and anteroposterior planes, respectively). In the example considered, the center of mass corresponds to the pelvis of subject UR1 while the latter performs a squatting movement. These kinematic variables Via show in this case that: - the movement of the center of mass along Z is negligible over time t, which means that the subject UR1 is laterally centered; - the center of mass of subject UR1 moves along the Y-axis during time t between a reference position "0" and a low position corresponding to -0.7 meters; and - during the squat movement, the center of mass moves back slightly (by 0.1 meters) along X due to the backward displacement of the buttock of subject UR1 (slight backward movement) then returns to its initial position along X.

[0139] By way of illustration, [Fig. 11] schematically represents a kinematic variable VI determined at S6 over time, this variable representing an angular displacement (in degrees) along a given direction of a joint, namely the knee of subject UR1. In this example, the curve therefore represents the evolution of an angular motion component as a function of time (as a percentage of a gait cycle) while subject UR1 performs a walking motion. As this figure shows, the evolution of the knee flexion angle exhibits two peaks during a complete gait cycle. Indeed, at the beginning of the cycle, the heel touches the ground, then the knee passes under the subject's body (1st angular peak).Following this initial flexion, the knee extends into a posterior position (almost zero flexion) before reaching a second, larger angular peak during which the subject's toe leaves the ground and the knee is brought forward until it returns to the initial position.

[0140] During a determination step S8 ([Fig.4]), the processing device T1 determines, from the kinematic variables V1 obtained in S6, at least one nonlinear variability indicator V2 representative of a temporal structure of the kinematic variables VI (or of at least one variable VI among those previously calculated in S6). One or a plurality of these indicators V2 can be determined as appropriate.

[0141] A nonlinear variability indicator V2 is a statistical or mathematical indicator that characterizes in a complex way how one or more kinematic indicators VI vary over time. Unlike a linear indicator, which would assume proportional and constant relationships, a nonlinear variability indicator captures more complex dynamics such as chaos, nonproportional interdependence, or emergent behaviors. These nonlinear indicators can include, for example, sampling entropy, the Liapunov exponent, the fractal dimension, and other methods from chaos theory and dynamical systems theory. These indicators have been found to be particularly effective for analyzing time series or data where traditional linear models fail to capture the full range of underlying dynamics, thus providing a deeper understanding of the systems under study.

[0142] According to one example, each kinematic variable VI defines a motion component (e.g., position, velocity, or acceleration, angular or translational) in one dimension over time. The device T1 thus determines at least one nonlinear variability indicator V2 in S8 ([Fig. 4]) by projecting the kinematic variables VI into an N-dimensional space, where N > 1. From this indicator or these indicators V2, kinematic variables can advantageously be reconstructed in an N-dimensional space (N > 1).

[0143] By way of illustration, [Fig. 12] includes various C1-C6 representations of the evolution of signals over time, as well as the respective values ​​of a Liapunov exponent representative of the variability of the signal under consideration. This example shows several signals with different signal time structures (periodic, non-periodic, or random). Signals C1, C3, and C5, on the one hand, and signals C2, C4, and C6, on the other, vary within the same range of values ​​(30 and 80, respectively) and have the same mean (equal to 0) for both groups. However, their time structures may or may not differ within the same group, which is reflected in the values ​​of the Liapunov exponents LyE indicated by the reference sign 34 (ranging from -0.001 to 0.564 in this example).For example, although varying within different value ranges, signals C1 and C2 have the same Lyapunov exponent value because they are characterized by the same temporal structure (purely periodic signals, LyE = 0). Signals C3 and C4 (chaotic) each exhibit a certain degree of repetition (repetition of a variation pattern), which is reflected by an increase in their Lyapunov exponent values ​​(close to 0.1). Signals C5 and C6 are completely random, so that each point in time is uncorrelated with other points, resulting in even higher Lyapunov exponent values ​​(close to 0.5).

[0144] By way of illustration, [Fig.13] illustrates further examples of signals (A), (B) and (C) (left) representative of kinematic variables V1 in one dimension and respectively periodic, chaotic and random, as well as their corresponding spatial representations in a 3-dimensional space.

[0145] Figure 14 reproduces the right-hand column of Figure 13 in two dimensions to place the elements on a plane representing the predictability (x-axis) and complexity (y-axis) of the signal. It shows that random signals (C) and periodic signals (A) are respectively not very and very predictable, although only slightly complex, unlike chaotic signals (B), which are complex with a variable predictability falling between the first two.

[0146] According to one example, the nonlinear variability indicator(s) V2 determined in S8 ([Fig.4]) include at least one of the following: - an entropy value or an entropy indicator; - a Liapounov exponent; - a result of a fractal analysis (for example a fractal analysis value or statistical self-affinity value or a DFA rectified fluctuations analysis in English); - a surogation indicator (for "surogate" in English); - a geometric approach indicator, for example UCM for "Uncontrolled Manifold”, TNC for “Tolerance-Noise-Covariation”, MIP for “Minimum Intervention Principle” or GEM for “Goal Equivalent Manifold”.

[0147] Each type of nonlinear variability indicator provides a type of information on how the kinematic data V1 vary over time. In particular, they can provide information on the regularity or predictability of a motion.

[0148] Entropy is a mathematical tool used to quantify the regularity and unpredictability of fluctuations in time series data. In practice, the presence of repetitive fluctuation patterns in a time series makes it more predictable than a time series in which such patterns are absent. Entropy thus reflects the probability that similar observations will not be followed by other similar observations. A time series containing many repetitive patterns exhibits relatively low entropy; conversely, a less predictable process exhibits higher entropy. Using an entropy value is advantageous for characterizing a level of disorganization, or unpredictability, in the variations of kinematic variables VI, which in turn allows for the identification of relevant posture changes and thus improves the subject's posture analysis.

[0149] By way of illustration, [Fig. 15] represents an example in which an approximate entropy indicator V2-1 (also called ApEn or "approximate entropy") is determined in S8 ([Fig. 4]) as a nonlinear indicator of variability V2 from a time series of knee angular flexion VL. As shown in (A), the time series is first divided into short vectors of similar length m. One of these vectors is represented in [Fig. 15](A) by the arrow between points u(44) and u(45). Then, for each vector thus determined, the processing device T1 determines the number of other vectors that are similar to said vector. Vectors are considered similar to the original vector when their tails and tails are contained within a band of width r above and below those of the original vector, here u(44) ± r and u(45) ± r.The vectors that were found to be similar to the original vector are represented by the other arrows. As illustrated in [Fig. 15](B), this operation can then be repeated for vectors that are one unit longer than the shorter vectors (for example, for a vector extending between u44 and u46 as illustrated in 15(B)), i.e., of a length of m+1. The ApEn is then obtained by calculating the natural logarithm of the relative prevalence of repeating patterns of length m compared to those of length m+1.

[0150] The largest Lyapunov exponent, also called more simply the "Liapunov exponent", is a measure of sensitivity to initial conditions in a dynamical system. It can be used to quantify the chaotic nature of a system Nonlinear dynamics. It allows us to measure how the trajectories of a dynamic system diverge from slightly different initial conditions. In a chaotic system, small initial variations can lead to very different trajectories, meaning that the system is sensitive to initial conditions. Using a Lyapunov exponent (or function) is advantageous for estimating the stability of an equilibrium point, that is, the stability of the kinematic variables.

[0151] To calculate the Lyapunov exponent, two very close initial trajectories in state space (generally, very close points in the time series) can be selected. The evolution of these trajectories can then be tracked over time. The Lyapunov exponent is then calculated as the limit (for example, by taking the average over all possible trajectories) of the exponential growth of the differences between these trajectories. As a result: a positive Lyapunov exponent indicates sensitivity to initial conditions and suggests that the system is chaotic. In this case, small initial perturbations can lead to exponential divergences in the trajectories, making long-term prediction difficult. A zero exponent indicates that the trajectories remain close over time, meaning that the system is stable and predictable.Conversely, a negative exponent indicates that the trajectories converge over time, meaning that the system is stable and tends to return to an equilibrium state. The result RS1 determined at S10 ([Fig.4]) by the processing device T1 can therefore include an indication of the sign of the obtained Liapunov exponent, or an indication of whether the obtained Liapunov exponent is positive (> 0), negative (> 0) or zero (= 0).

[0152] Detrended Fluctuation Analysis (DFA) is a fractal analysis technique used to quantify the fluctuations of a time series. This technique involves, in particular, decomposing a signal into small time scales and then calculating the relationship between the fluctuations and the size of the time scales. The alpha coefficient of DFA can then be used to assess the presence of fractality in a time series. An alpha coefficient value between 0.5 and 1 indicates positive autocorrelation (or persistence) in the time series, suggesting a fractal structure. The closer the alpha value is to 1, the more the series is considered fractal and therefore exhibits similarities or repeating patterns.In practice, an autocorrelated time series is characterized by a positive correlation between successive values; that is, high values ​​are likely to be followed by high values, and low values ​​by low values. Conversely, an alpha coefficient between 0 and 0.5 indicates an anticorrelation (or antipersistence) in the time series. This is characterized by an inverse relationship between successive values; that is, when the value... If the previous value is high, it is likely that the next value will be low, and vice versa.

[0153] The GEM is a method, or indicator, that allows a movement to be represented in a multidimensional space. This indicator makes it possible, after defining a task objective, to see how one or more kinematic variables are associated with the success of this objective by looking at the "good" variability (beneficial to the success of the task objective) and the "bad" variability (leading to a deterioration of performance).

[0154] By way of illustration, [Fig. 16] shows an example in which a GEM V2-2 indicator is determined in S8 ([Fig. 4]) as a nonlinear V2 variability indicator. Part (A) of [Fig. 16] illustrates the GEM indicator for maintaining a constant walking speed (v). The central point represents the preferred mean operating point. Each triangle represents a combination of stride time and stride length. Points that lie exactly on the diagonal line of the GEM indicator have achieved the same speed and satisfy the objective, namely, in this example, maintaining a constant walking speed v. Part (B) of [Fig. 16] illustrates the time series of deviations ôT (tangential deviation from the GEM) and ôP (perpendicular deviation from the GEM) for the dataset shown in (A).

[0155] During an analysis step S10 ([Fig.4]), the processing device T1 analyzes said at least one nonlinear variability indicator V2 (determined in S8) by comparing it with a respective reference value REF2. A result RS1 is produced from this S10 analysis.

[0156] According to one example, the device Tl determines a single non-linear indicator of variability V2 in S8 ([Fig.4]) and compares this to a reference value REF2 to determine the result RS1.

[0157] Alternatively, the device T1 determines a plurality of nonlinear V2 variability indicators in S8 ([Fig. 4]) and compares each of them to a respective reference value REF2 to obtain the RS1 result. The use of a combination of different types of nonlinear V2 variability indicators can, for example, advantageously allow for a better characterization of the variability of the V1 kinematic values ​​in their complexity and thus improve the RS1 result of the S10 analysis. Each nonlinear V2 variability indicator can provide a type of information so as to obtain a more complete analysis result covering various aspects of variability, such as the regularity and predictability of posture, for example, with regard to a task objective.

[0158] According to one example, the device T1 checks (S 10, [Fig. 4]) whether said at least one nonlinear variability indicator V2 determined in S8 meets one (or at least one) conformity criterion CRI with respect to its respective reference value REF2. For example, it is checked whether a nonlinear variability indicator V2 is greater than or less than its respective value REF2 or whether the difference between these two values ​​is greater than or less than a threshold value. The RS1 result determined in S10 then depends on whether the conformity criterion is met or not.

[0159] According to one example, the verification of the conformity of a nonlinear variability indicator V2 to a plurality of CRI conformity criteria can be carried out in order to obtain the result RS1. As an example, it can be verified whether a Liapunov exponent representative of the variability meets a plurality of CRI conformity criteria.

[0160] The result RS1 can be a function of whether a plurality of non-linear variability indicators V2 meet a respective conformity criterion CRI with respect to a respective reference value REF2. In other words, the device T1 can perform successive comparisons over time to verify that the evolution of a non-linear variability indicator V2 meets CRI criteria over time.

[0161] By way of example, the T1 device can verify whether a Lyapunov exponent LyE determined at S8 [Fig. 4]) as a nonlinear indicator of variability V2 meets the following two CRI conformity criteria: (1) LyE > 0 (indicating that the movement of subject UR1 is not purely periodic or stereotyped) and (2) LyE < 0.5 (indicating that the movement of subject UR1 is not too random). It has indeed been observed that a LyE value that is too low suggests that subject UR1 is not in good condition (injury, pathology, etc.), which induces stereotyped behavior.Conversely, an excessively high LyE value is undesirable, as the random nature of a movement suggests certain disorders or disturbances in the subject (URL). High variability is necessary for an athlete to adapt to all situations in a sport, for example, rugby, but excessive variability can occur in cases of uncontrollable or disorganized movements, which may result from a disorder. By verifying the conformity of this Lyapunov exponent (LyE) to the two criteria above (relative to the reference values ​​REF2 of 0 and 0.5), one can thus advantageously ensure that the subject's movement (UR1) exhibits satisfactory variability.

[0162] According to one example, the processing device T1 determines in S8 ([Fig. 4]) a plurality of nonlinear variability indicators V2, each of which is compared in S10 with a respective reference value REF2 to obtain the result RS1 of said analysis S10. It has notably been observed that the combination of an entropy indicator V2 with a Liapunov exponent LyE allows for the efficient analysis of variability due to their complementarity. Indeed, these two indicators exhibit different time series characteristics. For example, a signal may have high regularity (low entropy) but may exhibit diverse values ​​in terms of sensitivity to initial conditions (Liapunov exponent). These two indicators can therefore provide valuable information on different and particularly complementary aspects (regularity and sensitivity to initial conditions) of the subject's posture, and thus further improve posture analysis.

[0163] According to one example, the processing device T1 determines the GEM indicator (step S8, [Fig. 4]), which provides access to the time series of tangential and perpendicular deviations from the task objective (as illustrated in [Fig. 16](B)). It is then possible to apply one or more nonlinear analysis techniques, such as DFA, to these deviation time series. The GEM indicator indicates which variability is beneficial or detrimental to the achievement of the task objective, while the other nonlinear analysis techniques provide relevant information on the temporal structure of the variability. The relevance of the posture analysis can thus be further improved.

[0164] According to one example, a nonlinear indicator V2 determined in step S8 ([Fig. 4]) comprises a first sequence of values ​​over time. Thus, the device T1 can determine a first sequence of values ​​defining the evolution of a nonlinear variability indicator V2 over time. In this case, the device T1 compares, for example, this first sequence of values ​​of said at least one nonlinear indicator V2 over time with a second sequence of reference values ​​REF2. Thus, the values ​​of this nonlinear indicator V2 over time can be compared to the respective reference values ​​REF2. The result RS1 of the analysis S10 can then be determined based on the result of the comparison of the first and second sequences.In other words, by comparing the values ​​of a non-linear variability indicator V2 with respective reference values ​​REF2 over time, one can advantageously analyze in detail the temporal structure of the kinematic data VI, and therefore how they vary over time. During these comparisons, the device T1 can, for example, verify the conformity of each value of the first sequence with a respective conformity criterion CRI with respect to the corresponding reference value REF2.

[0165] As an example, the device T1 can verify whether a Lyapunov exponent (LyE) determined at S8 [Fig. 4] as a non-linear indicator of variability (V2) increases over time, or even increases at a minimum growth rate. It has been observed that an injured person typically exhibits stereotyped behavior with a low level of variability. Thus, it is possible to verify, for example, whether a convalescent patient's LyE increases over time, indicating an increase in variability, synonymous with recovery.

[0166] According to one example, during the monitoring process, the processing device T1 also determines at least one linear indicator V3 of variability from the kinematic variables VI and compares this linear indicator of variability with a The second reference value is REF3. The RS1 result of the S10 analysis is then also a function of this comparison. Taking into account at least one non-linear indicator V2 combined with at least one linear indicator V3 further improves the quality of the S10 analysis. Linear indicators can thus be used in addition to non-linear indicators to further enhance posture analysis.

[0167] According to a particular example, at least one nonlinear V2 variability indicator is determined in S8 ([Fig. 4]) and an RS1 result is determined in S10 from this or these V2 indicators. The processing device T1 can then apply one or more linear analyses to the RS1 result (and vice versa). This RS1 result can then be used during step S12 ([Fig. 4]) described below.

[0168] With further reference to [Fig. 4], during a generation step S12, the processing device T1 generates a posture instruction CMD1 based on the result RS1 of the analysis S10. This posture instruction CMD1 can then be sent to a control device 20 ([Fig. 1]) to guide the subject UR1 in a posture, movement, or physical activity. Sending such an instruction UR1 can, for example, advantageously allow for the adaptation, improvement, and / or correction of the subject's posture UR1. The nature and number of posture instructions CMD1 can vary depending on the case.

[0169] The CMD1 posture instruction is intended, for example, for the subject UR1 and / or any other user to assist the subject in their postures or movements. This instruction can, for example, be given to the subject UR1 themselves or to a professional to assist them in their care of the subject UR1 (for example, in developing a training program). The CMD1 posture instruction can, for example, instruct the subject UR1 to increase or decrease the intensity of a physical exercise, lengthen or shorten a physical exercise, or adapt a posture or movement.

[0170] According to one example, during the generation step S12, the device Tl renders (or makes available), as the first posture instruction CMDla, by means of a user interface, a posture information IF1 specifying how to correct or improve the posture of the subject URL. In other words, the device Tl can send a first posture instruction CMDla to a user interface to trigger the rendering, in an appropriate form, of a posture information IF1 specifying an adaptation, correction or improvement of the posture of the subject URL. According to a variant, the posture information IF1 specifies a posture, other than the PSI posture, to be achieved by the subject UR1, for example to improve or maintain his physical condition.

[0171] As already indicated, the posture information IF1 can be restored, under the control of the device Tl, in various forms, the way in which the IF1 information is restored being a function of the type of user interface used.

[0172] According to one example, the processing device Tl sends at least one first posture instruction, more precisely denoted CMDla ([Fig. 1]), to a display device 20a to cause the display of posture information IF1 (in visual form). According to another example, the processing device Tl can send a first posture instruction CMDla to a sound device 20b to cause the emission of posture information IF1 in audible form (for example, by beeps or via a synthesized voice). According to another example, the processing device Tl can send a first posture instruction CMDla to a light device 20c to cause the emission of posture information IF1 in visual form (for example, by activating an indicator light or by changing the color of an emitted light). Several of the user interfaces 20 described can also be used in combination.

[0173] By way of example, the processing device Tl can send a first instruction command CMDla to cause the display of information IF1 containing an indication to correct or improve the subject UR1's posture PSI. Information IF1 can, for example, indicate to subject UR1 and / or a third party (a healthcare professional or other) that a part of subject UR1's body must be moved or repositioned in a given direction for corrective purposes in order to perform a given task (e.g., jumping, running, squatting, etc.). Information IF1 can, for example, be displayed on a screen or in holographic form.

[0174] Fig. 17 represents a particular example of an embodiment in which the device Tl sends (S12, Fig. 4) at least one posture instruction CMDla to cause the display of a first posture information IF1 on a screen (or display device) 20a and / or to cause the rendering of a second posture instruction IF1 in the form of light indications by means of light sources 20c arranged around the subject URL. The screen 20a and / or the light sources 20c can thus transmit instructions to the user UR1 so that he adapts, improves and / or corrects his posture PSL. Reference 24 in Fig. 18 indicates, for example, the position of the feet of the subject UR1 in an initial default position.

[0175] Fig. 18 represents an example of an embodiment in which a screen 20a is used to display (S12, Fig. 4) posture information IF1 and a conveyor belt system is configured (S12) according to a CFI configuration in response to a CMD1 command sent as a posture instruction.

[0176] According to one example, during the S12 generation step (Figures 1 and 4), the device Tl sends, as a second posture instruction CMDlb, a command causing a parameterization (or configuration) of a movement assistance device 20d (also called a fitness device) to adapt, correct or improve the Subject's PSI posture. This command can trigger a setting to allow for a posture other than the PSI posture.

[0177] As already indicated, the assistance device 20d can be any device configured, under the control of the processing device Tl, to facilitate, guide, and / or intensify the performance of postures, movements, and / or physical activities by a subject or individual (weight training machine, treadmill, elliptical trainer, exercise bike, rowing machine, multi-gym station, etc.). The assistance device 20d can be more or less complex depending on the case. In particular, it can be configured according to at least one operating parameter (for example, a belt speed for a treadmill, a resistance level for an exercise bike or elliptical trainer, the duration of a training program, etc.). The processing device Tl can thus control the configuration of at least one parameter of the assistance device 20d by sending a second posture instruction CMDlb.

[0178] As an example, the device Tl sends a posture instruction CMDlb to a treadmill-type device 20d in S12 (figures 1 and 4) to adapt an operating parameter such as the speed of movement of the treadmill or an angle of inclination of the treadmill.

[0179] As already indicated, the device Tl can check in S10 ([Fig. 4]) whether a non-linear variability indicator V2 determined in S8 meets a conformity criterion CRI with respect to its respective reference value REF2. For example, if the conformity criterion is not met, the posture instruction generated in S12 ([Fig. 4]) is configured to correct the subject's posture PSI. In other words, the posture instruction PSI specifies a posture correction to be performed in the form of a posture information IF1 and / or a configuration command CMDlb.

[0180] The way in which the PSI posture should be corrected can be determined in various ways by the device Tl, depending on the case. For example, during the S10 analysis ([Fig. 4]), the processing device Tl determines whether a nonlinear variability indicator V2, referred to as the non-conforming indicator V2a, does not meet a conforming criterion CRI. If so, the device Tl determines at least one kinematic variable VI, referred to as the target kinematic variable Via, associated with the non-conforming indicator V2a and determines a correction to be applied to said at least one target kinematic variable Via to improve the non-conforming indicator V2a relative to the conforming criterion CRI (i.e., relative to the reference value REF2).In other words, the Tl device identifies at least one target kinematic variable Via on which to act to correct the non-compliant indicator V2a, as well as a correction to apply to this target kinematic variable Via in order to improve the non-compliant indicator V2a according to the considered compliance criterion CRI. The posture instruction CMD1 generated in S12 ([Fig.4]) can then be a function of the . correction to be applied to said at least one target kinematic variable Via to correct the posture of subject UR1.

[0181] The non-compliant indicator V2a can indeed be linked to a plurality of kinematic variables VI. It is therefore necessary to identify the kinematic variable(s) VI that need to be adjusted to correct the non-compliant indicator V2a according to the compliance criterion CRI and thus improve the subject's PSI posture URL

[0182] According to an example, during the S10 analysis step ([Fig.4]), the processing device Tl determines said at least one target kinematic variable Via and the correction to be applied to it by an optimization method (for example, of the "optimal control" type) taking as input: - a trend objective for the non-compliant indicator V2a (by an objective of increase or decrease, or an objective of reaching a target value or range of values); - a function defining a relationship between the non-conforming indicator V2a and a plurality of associated kinematic variables V2; and - constraint data defining movement constraints that the ML1 model must respect.

[0183] According to one example, the processing device T1 determines a GEM indicator as a nonlinear V2 indicator of variability as previously described. The trend objective can therefore aim to minimize "bad" variability (variability detrimental to the success of the task objective) because it is deemed too high (non-compliant indicator). The definition of the GEM provides information on the correlation (relationship function) of this indicator with several kinematic variables VI, such as the velocity and position of anatomical parts of the subject. The constraint data to be respected can include, for example, biomechanical rules, such as angular amplitudes or speeds of movement that the human body cannot (or should not) exceed. By combining these three elements in an optimization tool, it is advantageously possible to obtain optimal posture options that respect these criteria and thus generate a relevant posture instruction.

[0184] The optimization method thus used in S10 ([Fig. 4]) can output setpoint data defining at least one target kinematic variable Via and the correction to be applied to it to modify the non-conforming indicator V2a according to the trend objective in order to reach or approach the trend objective. The posture setpoint CMD1 can then be generated in S12 ([Fig. 4]) from the setpoint data thus determined.

[0185] According to one example, the trend objective provides, for example, for the maximization or minimization of the non-conforming indicator V2a, or for the modification of this indicator so that it falls within a range of values. The function defining a The relationship between the non-conforming indicator V2a and a plurality of associated kinematic variables V2 can result from, or be deduced from, the very definition of the non-conforming indicator V2a. Furthermore, the constraint data considered during the optimization method can vary depending on the case: they may, for example, define movements that certain parts of the subject's body UR1 or certain of their anatomical points PT1 cannot perform (e.g., an angular displacement beyond which a knee cannot move, etc.). A non-conforming indicator V2a can thus be effectively corrected, thereby advantageously improving posture or movement. By correcting posture to move towards a target, the user UR1 can be advantageously assisted in their posture or movement.

[0186] According to an example, prior to the S6 calculation step ([Fig. 4]), the processing device Tl performs a recognition, based on the ML1 model generated in S4, of a movement performed by the subject UR1. This movement recognition can be carried out, for example, using a prediction model based on artificial intelligence. The device Tl then selects, based on the recognized movement, kinematic variables VI to be calculated during the S6 calculation step.

[0187] Indeed, it can be useful not to calculate the same kinematic variables V1 depending on the posture or movement performed by the user URL. In other words, the relevant kinematic variables may differ depending on the posture considered. The kinematic variables VI calculated in S6 can thus be adapted (or selected) according to the situation in which the subject is placed URL. By calculating only the relevant kinematic variables, superfluous processing can be avoided, resources (time, processing) can be saved, and the performance of the tracking process can be improved.

[0188] According to one example, the device Tl further determines a task objective to be achieved and selects the kinematic variables VI to be calculated as a function of both the recognized movement and the task objective, in order to further improve the performance of the process.

[0189] In general, the present invention advantageously allows for the precise and reliable monitoring and adaptation of a subject's posture. In particular, it is possible to assess the quality of a subject's static or dynamic posture over time and to guide or improve this posture by generating a posture instruction adapted to the subject's behavior. Such monitoring makes it possible, in particular, to assess and improve the subject's health and / or performance for various purposes, such as, for example, to assist with physical activity, physical preparation, play, physical rehabilitation, and / or athletic reconditioning. The invention also allows for precise and reliable monitoring of a subject's static or dynamic posture at a low cost and limited implementation complexity. The present invention can also provide diagnostic assistance.

[0190] It has indeed been observed that if a person is injured, immobilized, or exhibits certain disorders or impairments, they may display stereotyped behavior, resulting in a low level of variability in their kinematic variables. Conversely, excessive variability in kinematic variables suggests other disorders or impairments, resulting in disorganized or uncontrollable movements. In theory, a relatively high degree of variability in kinematic variables is desirable to allow the subject to adapt to all situations during physical exercise, but within certain limits to maintain good control of movements.

[0191] By analyzing at least one non-linear variability indicator based on kinematic variables, it is advantageous to evaluate the quality of a subject's movement in a detailed and reliable manner, taking into account complex aspects not perceptible on the basis of kinematic variables alone. Depending on the result of this analysis, a postural instruction can be advantageously generated, for example, to adapt, improve, or correct the subject's posture and thus help improve the subject's health and / or performance.

[0192] The invention therefore offers a powerful tool, both for movement professionals such as health and / or sports professionals (doctors, therapists, physiotherapists, physical trainers, coaches / sports trainers, etc.) and for the general public, to analyze and improve the posture of a subject.

[0193] As those skilled in the art will understand, all the embodiments and variations described above, some of which have been intentionally simplified for ease of explanation, are merely non-limiting examples of implementation of this disclosure. In particular, those skilled in the art may consider any adaptation or combination of the embodiments and variations described above to meet a specific need.

[0194] The present invention is therefore not limited to the embodiments described above but extends in particular to a monitoring method that would include secondary steps without departing from the scope of the present invention. The same would apply to a processing device, or more generally to a processing system, for implementing such a method.

Claims

Demands

1. A processing method, implemented by a processing device (T1), for postural tracking of a subject (UR1), said method comprising: a) obtaining (S2) sensor data (DTI) representative of the subject; b) generating (S4), from the sensor data, a model (ML1) defining anatomical points (PT1) of the subject over time; c) determining (S6), from the model, kinematic variables (VI) representative of the evolution of the subject's posture over time; d) determining (S8), from the kinematic variables, at least one nonlinear variability indicator (V2) representative of a temporal structure of the kinematic variables (VI); e) analyzing (S10) said at least one nonlinear variability indicator by comparison with a respective reference value (REF2); and f) generating (S12) a postural setpoint (CMD1) as a function of a result of the analysis.

2. A method according to claim 1, wherein the sensor data obtained in a) comprises image data (DTla) captured by a plurality of image capture devices (12a), the image data from each image capture device defining a respective 2D view of the subject over time; the generation b) comprising: - estimation of a position (PN1) of the anatomical points (PT1) of the subject over time in the 2D view from each image acquisition device; and - generation of the 3D model by triangularization of the estimated positions of the anatomical points in each 2D view.

3. Method according to claim 2, wherein generation b) comprises: - assigning, to each estimated position (PN1) of the anatomical points of the subject in the 2D views, a respective confidence score (SCI) representative of a level of confidence in said estimated position; the model (ML1) being generated from the estimated positions of the anatomical points among which are excluded, by a first filtering, each estimated position of an anatomical point whose confidence score does not reach at least a first threshold value (SL1).

4. Method according to claim 3, wherein the generation b) of the model comprises: - comparison of first and second confidence scores associated respectively with first and second estimated positions for the same anatomical point in 2D views from image acquisition devices; the model being generated from the estimated positions (PN1) of the anatomical points (PT1) to which a second filtering is applied according to which, if the difference between said first and second confidence scores reaches at least a second threshold value, the one among the first and second estimated positions having the lowest confidence score is excluded.

5. A method according to any one of the preceding claims, wherein said at least one nonlinear indicator (V2) determined in d) comprises a first sequence of values ​​over time; the analysis e) comprising: - comparison of the first sequence of values ​​of said at least one nonlinear indicator over time with respectively a second sequence of reference values; the result of the analysis e) being a function of a result of said comparison of the first and second sequences.

6. A method according to any one of the preceding claims, wherein generation f) comprises at least one of: - rendering, as first posture instruction (CMD1a), by means of a user interface (20a-20c), of posture information (IF1) specifying how to correct the subject's posture; and - sending, as second posture instruction (CMD1b), a command causing a setting of a movement assistance device (20d) to correct the subject's posture.

7. A method according to any one of the preceding claims, wherein analysis e) comprises: - checking whether said at least one nonlinear variability indicator (V2) meets a conformity criterion (CRI) with respect to the respective reference value; wherein, if the conformity criterion is not met, the posture instruction generated in f) is configured to correct the subject's posture.

8. A method according to claim 7 wherein, if a non-linear variability indicator (V2a) – referred to as a non-conforming indicator – does not comply not the conformity criterion (CRI), analysis e) includes: - determination of at least one kinematic variable (Via), called target kinematic variable, associated with the non-conforming indicator (V2a) and a correction to be applied to said at least one target kinematic variable to improve the non-conforming indicator relative to the reference value according to the conformity criterion; said posture instruction being generated in f) as a function of the correction to be applied to said at least one target kinematic variable to correct the subject's posture.

9. A method according to claim 8, wherein the determination of said at least one target kinematic variable (Vla) and the correction to be applied is carried out by an optimization method taking as input: - a target trend of the non-conforming indicator; - a function defining a relationship of the non-conforming indicator with a plurality of associated kinematic variables; and - constraint data defining motion constraints that the model must respect; the optimization method delivering as output setpoint data defining said at least one target kinematic variable and the correction to be applied to modify the non-conforming indicator according to the target trend; the posture setpoint being generated in f) from the setpoint data.

10. A method according to any one of the preceding claims, wherein the method comprises, prior to calculation c): - recognition, from the model, of a movement performed by the subject; and - selection, as a function of the recognized movement, of the kinematic variables to be calculated in c).

11. Computer program (PG1) comprising instructions for carrying out the steps of a processing method according to any one of the preceding claims when said program is executed by a computer.

12. A processing device (T1) to assist a subject (UR1) in performing a posture (PSI), said device comprising: - an acquisition module (MD2) configured to obtain sensor data (DTI) representative of the subject; - a first generation module (MD4) configured to generate, from sensor data, from a model (ML1) defining anatomical points (PT1) of the subject over time; - first determination module (MD6) configured to calculate, from the model, kinematic variables (VI) representative of the evolution of the subject's posture over time; - second determination module (MD8) configured to determine, from the kinematic variables, at least one non-linear variability indicator (V2) representative of a temporal structure of the kinematic variables; - analysis module (MD10) configured to analyze said at least one non-linear variability indicator by comparison with a respective reference value (REF2); and - second generation module (MD 12) configured to generate a posture instruction (CMD1) based on a result (RS1) of the analysis.

Images