EP4554471A1 - Procédé et dispositif de détection de chute par capture inertielle - Google Patents
Procédé et dispositif de détection de chute par capture inertielleInfo
- Publication number
- EP4554471A1 EP4554471A1 EP23750655.5A EP23750655A EP4554471A1 EP 4554471 A1 EP4554471 A1 EP 4554471A1 EP 23750655 A EP23750655 A EP 23750655A EP 4554471 A1 EP4554471 A1 EP 4554471A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- data
- event
- accelerometer
- barometer
- gyrometer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
- A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
- A61B5/1116—Determining posture transitions
- A61B5/1117—Fall detection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
- A61B5/7282—Event detection, e.g. detecting unique waveforms indicative of a medical condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/74—Details of notification to user or communication with user or patient; User input means
- A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
- A61B2562/02—Details of sensors specially adapted for in-vivo measurements
- A61B2562/0219—Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
Definitions
- TITLE Method and device for detecting falls by inertia capture Ile
- the invention relates to an automatic fall detection method and device based on one or more sensors worn by a natural person.
- the accelerometer is the sensor most often used for fall detection, including in recent studies.
- the gyrometer is rarer, because its contribution in terms of performance does not always justify the increase in electrical consumption induced.
- the barometer is very rarely used due to its significant noise and possible redundancy with the accelerometer.
- the first detection models were based on one or more thresholds on characteristics extracted from signals such as the norm of acceleration, the speed at the point of impact, or the angular change. Despite correct announced detection rates, the very high number of false alerts (several per day) did not make them suitable for daily use. In institutions for people requiring care (for example: EHPAD), these devices were abandoned by the healthcare teams due to these false positives which caused additional work on teams already in high demand elsewhere. Raising the detection thresholds solves these problems of false alarms (better specificity), but deteriorates performance in terms of good detection (less good sensitivity).
- the invention firstly concerns a fall detection device comprising at least one accelerometer and means, or digital means, programmed for:
- - identify for example by a filter, among the data from said accelerometer which exceed a certain threshold, data of at least one signal likely to constitute a fall event; - identify, among said data likely to constitute a drop event and in a time window around, or including, this event, at least the longest duration, or 1st duration, during which the signal is, uninterrupted, under the average value of the detected signal;
- - classify preferably according to said greatest duration, said data or the event, in fall or in non-fall, by regression of a probability or using a classifier based for example on at least one tree decision and/or on a gradient descent (XGB, or “Gradient Boosting tree”), and/or an additional tree (or
- the invention implements a drop predictor (the longest duration during which the signal is, uninterrupted, below the average value, also called a drop predictor (the longest duration during which the signal is, uninterrupted, below the average value, also called a drop predictor (the longest duration during which the signal is, uninterrupted, below the average value, also called a drop predictor (the longest duration during which the signal is, uninterrupted, below the average value, also called a drop predictor (the longest duration during which the signal is, uninterrupted, below the average value, also called
- “longest_strike_below_mean”) which, in combination with a classification, based for example on at least one decision tree and/or on a gradient descent (XGB, or a “Gradient Boosting tree”), and/or on an additional tree ( or "Extremely Randomized Tree", or ETC, or “Extra Trees”), turns out to be very relevant for the detection of falls, in particular because it is very differentiating between, on the one hand, falls and, on the other hand , the activities of daily life, while being very fast and low energy consuming.
- XGB or a “Gradient Boosting tree”
- ETC Extremely Randomized Tree
- This predictor corresponds to the duration of the longest “strike” (continuous segment) located below the average of a signal, when this is likely to be a signal reflecting a fall. According to one embodiment, It is calculated over a 1st time window of, for example, 1 second before the main event and over a 2nd time window of, for example, 2 seconds, after the main event.
- This predictor is similar to the duration of the fall. It was identified, among around 1000 temporal and frequency predictors, as being more relevant, for fall detection, than other commonly used predictors (speed at the point of impact in particular). It is even more effective when the sensor is worn on the wrist, which is the solution favored by seniors.
- the means, or the digital means can be programmed to calculate, from the data received from the accelerometer, the dynamic component of the acceleration norm, which is then used to identify data of at least one signal capable of to constitute a fall event.
- the classifier used may be:
- Random Forest type with for example between 2 and 20 trees, each with a depth which can be between 2 and 12;
- RFT Limited Random Forest
- the event can be classified as falling or non-falling by random forest regression.
- a device may further comprise means, or digital means, for calculating, for the data of the accelerometer signal, or accelerometer data, one or more of the other values among the maximum value of these data d accelerometer, the minimum value of this accelerometer data, the median value of this accelerometer data, the standard deviation of this accelerometer data, the variance of this accelerometer data, the root mean square value of this data accelerometer, the sum of the values of this accelerometer data, and to enter at least one of said other calculated values into said classifier.
- a device may further comprise at least one other sensor chosen from a gyrometer and/or a barometer.
- the means, or the digital means can be programmed to calculate, from the data received respectively from a gyrometer and/or a pressure sensor, the standard of the angular speed and/or the short-term variation of the local atmospheric pressure, which is/are then used to identify data from at least one signal likely to constitute a fall event.
- a device according to the invention may further comprise means, or digital means, programmed to calculate:
- gyrometer signal data for the gyrometer signal data, or gyrometer data, one or more of the values among the greatest period during which the signal is, continuously, below the average value of the detected signal, the maximum value of this gyrometer data, the median value of this gyrometer data, the minimum value of this gyrometer data, the root mean square value of this gyrometer data, the standard deviation of this gyrometer data, the variance of this gyrometer data, and to enter at least one of these values calculated for the gyro signal data in said classifier;
- the barometer signal data or barometer data, one or more of the values among the greatest period during which the signal is, continuously, below the average value of the detected signal, the maximum value of these data of barometer, the average value of this barometer data, the median value of this barometer data, the minimum value of this barometer data, the root mean square value of this barometer data, the standard deviation of this barometer data, the sum values of this barometer data, the variance of this barometer data and to enter at least one of these calculated values for the barometer signal data into said classifier.
- the time window around, or including, the event can be a maximum of 1.5 seconds before the event and/or a maximum of 2.5 seconds after the event.
- the means, or the digital means, for identifying whether an event is likely to constitute a fall can be programmed to identify whether an event is followed, for a certain duration, or 2nd duration, for example between 2 and 4 seconds, d another event, or is not followed by another event: repetitive events can thus be eliminated from events likely to constitute a fall.
- the means, or digital means, for identifying whether an event is likely to constitute a fall may implement a state machine.
- a device may further comprise means, or digital means, for, or programmed to, send an alert on the basis of the results generated by the means, on the basis of the classification result.
- the invention also relates to a method for detecting a fall using at least one accelerometer, this method comprising at least the following steps: a) - collecting at least acceleration data coming from said accelerometer; b) - identify, among the data from said accelerometer, data likely to constitute a fall; c) - identify, using said data likely to constitute a fall event and in a time window around, or including, this event, at least the longest duration, or 1st duration, during which the signal is, uninterruptedly, below the average value of the detected signal; d) - classify, preferably according to said greatest duration, said data or the event, in fall or in non-fall, by regression of a probability or using a classifier based for example on at least a decision tree and/or on a gradient descent (XGB, or “Gradient Boosting tree”), and/or on an additional tree (or
- the classifier is for example of the Random Forest type, with for example between 2 and 20 trees, each with a depth which can be between 2 and 12.
- Said classifier can implement at least one Limited Random Forest (RFT), for example limited to 6 trees (rfjight, or “Random Forest limited”).
- RFT Limited Random Forest
- the event can be classified as falling or non-falling by random forest regression.
- Such a method may further comprise a step of calculating, for the accelerometer signal data, or accelerometer data, one or more of the other values among the maximum value, the minimum value of this accelerometer data, the median value of this accelerometer data, the standard deviation of this accelerometer data, the variance of this accelerometer data, the root mean square value of this accelerometer data, the sum of the values of this accelerometer data , and the entry of at least one of these other values, or calculated values, into said classifier.
- Such a method may further comprise a measurement of data from said movement using at least one other sensor chosen from a gyrometer and/or a barometer and/or the collection of at least one data from at least one other chosen sensor. among a gyrometer and/or a barometer.
- step b) Before step b), it is possible to calculate, from the data received respectively from a gyrometer and/or a pressure sensor, the standard of the angular velocity and/or the short-term variation of the pressure local atmosphere, which is/are then used during steps b) and following.
- Such a method may also include a calculation step:
- gyrometer signal data for the gyrometer signal data, or gyrometer data, one or more of the values of the greatest period during which the signal is, continuously, below the average value of the detected signal, the maximum value of such gyrometer data , the median value of this gyrometer data, the minimum value of this gyrometer data, the root mean square value of this gyrometer data, the standard deviation of this gyrometer data, the variance of this gyrometer data, and the input at least one of these values calculated for the gyrometer signal data in said classifier;
- the barometer signal data one or more of the values among the greatest period during which the signal is, continuously, below the average value of the detected signal, the maximum value of this barometer data, the average value of this barometer data, the median value of this barometer data, the value minimum of this barometer data, the root mean square value of this barometer data, the standard deviation of this barometer data, the sum of the values of this barometer data, the variance of this barometer data, and the entry of at least one of these values calculated for the barometer signal data in said classifier.
- said time window around the event is for example a maximum of 1.5 seconds before the event and/or a maximum of 2.5 seconds after the event.
- Step b) can be implemented by identifying if an event is followed, for a certain duration, or 2nd duration, for example between 2 and 4 seconds, by another event, or is not followed by another event: repetitive events can thus be eliminated from events likely to constitute a fall.
- step b) implements a state machine.
- a method according to the invention may include a step of sending an alert based on the results of classification step d): if an event is classified as a fall, it is possible to send an alert to another person, for example to a caregiver.
- a method according to the invention can be implemented using digital means, for example a microcontroller or a processor or a microprocessor, programmed for this purpose.
- a device or a method according to the invention can be implemented in a bracelet or in a pendant.
- FIG 1 schematically illustrates an embodiment of a device according to the present invention.
- FIG 2 represents an example of carrying out, by a state machine, a predetection of events.
- FIG 3 represents an example of temporal windowing.
- FIG 4 represents the time interval corresponding to the predictor with the longest duration below the mean
- FIG 5A and [FIG 5B] represent the classification of the various predictors studied, carried out by a Random Forest type classifier, for a bracelet (figure 5A) and a pendant (figure 5B).
- FIG 6A and FIG 6B represent performances obtained, with a Random Forest type classifier, for a bracelet (figure 6A) and a pendant (figure 6B);
- FIG. 7 illustrates the contribution of the different sensors to the overall performance
- FIG 8 compares the respective performances of different classifiers
- FIG 9 schematically illustrates an embodiment of a method according to the present invention.
- FIG 10A and [FIG 10B] represent the influence of the hyper-parameters of a random forest on the performance (measured in average recall).
- a portable fall detection system 10 is shown schematically in Figure 1 and comprises at least one sensor making it possible to detect movements of a person, for example at least one accelerometer 2 (preferably 3 axes), and optionally at least one gyrometer 4 (preferably 3-axis) and/or at least one pressure sensor, or barometer, 6.
- Figure 1 represents a system comprising an accelerometer 2, a gyrometer 4 and a barometer 6, but other configurations can be made, as explained below.
- Such a system preferably comprises digital means 8, for example a microcontroller or a processor or a microprocessor, making it possible to receive the data transmitted or measured by the sensor(s) of the device and programmed to implement an analysis and/or or data processing as described below, preferably in real time.
- these means 8 can trigger the emission of an alert signal 12.
- These means 8 can also include storage means for storing data or instructions implementing a analysis and/or processing of data as described below and possibly storage means for storing all or part of the data measured by at least one of said sensor(s) and/or data which result from one or more of the treatments described below.
- Such a system can be incorporated into an everyday object worn by a natural person, for example a bracelet or a pendant.
- an accelerometer 2 a gyrometer 4 and a barometer 6, the signals that they produce during movements, and in particular during a fall, d 'a person to whom they are linked, are respectively an acceleration (of components (accN_x), (accN_y), (accN_z)), an angular velocity and a measure of local atmospheric pressure.
- acceleration of components
- accN_y acceleration
- accN_z angular velocity
- the means 8 can for example be programmed to calculate the dynamic component of the acceleration norm, the norm being given by the formula:
- the means 8 can be programmed to calculate the gyrN_mg standard of the angular speed, by a formula identical or similar to that given above.
- the means 8 are preferably programmed to retain the short-term variation in local atmospheric pressure (and eliminate slow variations due to meteorology).
- the barometer signals arrive, for example, at 50Hz; they can be resampled to, for example, 100Hz to simplify temporal windowing (in order to have a common frequency for the three sensors). Variations due to atmospheric pressure can be compensated for by applying a high-pass filter, for example with a cutoff frequency of 0.2 Hz.
- the corrected pressure is noted pressuredyn_mbar.
- the same filter is used for the accelerometer and the barometer, which is a resource-saving solution for the processing means (for example in RAM and CPU resources).
- a first processing, or pre-processing called “event detection” of at least part of the data produced by the sensor(s). Indeed certain events, which have nothing to do with a fall, for example a movement of the hand, can nevertheless cause at least one sensor to produce a signal of intensity greater than a certain threshold value, a signal that must be therefore a priori take into account.
- preprocessing is preferably carried out to eliminate signals of this type which cannot result from a drop.
- an event identified by a signal which exceeds a certain threshold (set by the operator in the means 8) is repetitive if it is repeated within a certain time window, for example of a duration of order of 2 to 4 s, for example another 3.5 s.
- a certain threshold set by the operator in the means 8
- an event not followed by another event during this duration is likely to constitute a fall.
- an analysis is triggered in the event of repetitive movements, but preferably only once, that is to say at the end of the succession of movements: if there was no preprocessing, for example by a state machine as explained below, an analysis would be triggered at each movement, thus increasing the energy consumption of the microcontroller.
- This pre-processing implements for example a state machine using as input the signals of the dynamic component of the acceleration accNdyn_mg (obtained as explained above). It can be programmed in means 8 and is illustrated in Figure 2.
- the first state (2-1) “WAIT 0 ” corresponds to waiting for an event.
- the dynamic component of the acceleration accNdyn_mg is calculated and compared to a threshold H_THR (which is defined previously by an operator in the means 8; for example it was defined by systematic search according to a regular grid, the criterion being the average recall ( “recall_mean”). If this threshold is exceeded by the dynamic component, the machine goes to state (2-2) IMPACT_DETECTE and then initializes a counter (“impact_detected_timer”).
- H_THR which is defined previously by an operator in the means 8; for example it was defined by systematic search according to a regular grid, the criterion being the average recall ( “recall_mean”). If this threshold is exceeded by the dynamic component, the machine goes to state (2-2) IMPACT_DETECTE and then initializes a counter (“impact_detected_timer”).
- IMPACT_DETECTE if a new crossing of the threshold by accNdyn is detected, the counter is reset to zero. Otherwise, we increment it by +1 for each new sample (for example at 104 Hz).
- IMPACT_DETECTED_TIMOUT which may for example, here again, have been defined by systematic search according to a regular grid, the criterion being the average recall (“recall_mean”)
- recall_mean the average recall
- the machine passes into the state 2-3 “FALL_LIKE_EVENT” (“FALL_LIKE_EVENT”) and triggers the rest of the analysis process (windowing, extraction of predictors and classification, see below). It then returns to the 2-1 “WAIT” state.
- Table 1 specifies examples of threshold values H_THR and IMPACT_DETECTED_TIMOUT, for a bracelet and for a pendant. [Table 1]
- Temporal windowing consists of selecting a range of samples on either side of the last threshold violation recorded by the state machine.
- This last threshold crossing is therefore located at the instant defined by the “IMPACT_DETECTED_TIMOUT” duration of samples before the transition to the “FALL_LIKE_EVENT” state (which is the literal translation of the equation above).
- temporal windowing Another step, called temporal windowing, can therefore then be implemented by means 8 to select before and/or after the event (which may have been identified in the manner explained above), the data of interest to be analyzed more precisely in order to confirm, or not, that this event is a fall.
- a time window F of a predetermined total duration is identified (by an operator in the means 8) , for example a few seconds, for example between 3 and 4s, around the last transition to the IMPACT-DETECTED state: a first part of the window is located before the last transition to the IMPACT-DETECTED state, a second part of the window is located after it; according to an exemplary embodiment with a window of a total duration of 3s, this is distributed in ls (corresponding for example to 100 samples) before the last transition to the IMPACT-DETECTED state and in 2s (corresponding for example to 200 samples) after the last transition to the IMPACT-DETECTED state.
- ls corresponding for example to 100 samples
- 2s corresponding for example to 200 samples
- This windowing step can be applied to the signals from at least one of the sensors mentioned above, preferably to the accelerometer signals; more particularly, it can be applied to one or more of the 3 quantities previously calculated during the conditioning phase: to the dynamic component of the norm of the accelerometer accNdyn_mg, to the norm of the gyrometer gyroN_mdps, to the dynamic component of the pressure pressuredyn_mbar.
- a device 1 can also only implement an accelerometer 2, or an accelerometer 2 and a gyrometer 4, or an accelerometer 2 and a barometer 6, in which case the explanations above are transposed to the signals of the sensor(s) implemented.
- This tool makes it possible to extract around 1000 predictors most commonly used in the analysis of time series, and to select those which are relevant to separate the two classes of events (falls and non-falls). This selection is based on a calculation of p-value, or probability, which can be seen as the smallest threshold of significance for which the null hypothesis is accepted, as explained in the article by JD Gibbons et al., “P- Values: Interpretation and Methodology,” Am. Stat., vol. 29, no. 1, p. 20-25, 1975, doi: 10.2307/2683674, with control of the false discovery rate by the method of Benjamini Yekutieli (Yoav Benjamini, Daniel Yekutieli.
- the filing order is simply alphabetical order.
- the units appear in the name of the predictor (for the gyrometer, the unit is millidegree per second (mdps));
- predictors are therefore the “longest strike below mean”, the maximum value, the average value (but not for the gyroscope data), the median value, the minimum value, the root mean square value, the sum of the values (but not for gyroscope data), the variance.
- the “longest_strike_below_mean” can for example be achieved (in means 8) using the algorithm below:
- Figure 4 (the dynamic component of the acceleration, expressed in milliG (mg)) represents a time window, around a transition to the IMPACT-DETECTED state, which can possibly be that of a fall, which represents the longer duration below average.
- the relevance of these predictors for separating falls from non-falls was evaluated using the Benjamini and Yekutieli method (see the article by these authors already cited above). Here we are mainly interested in the importance attributed to them by the classifier used (random forest as explained below). The degree of relevance is shown in Figures 5A and 5B, respectively for a bracelet and a pendant.
- a classification, or fusion, of the data can then be carried out, for example by a “random forest” type classifier, selected after a comparison of the performances of different classifiers.
- a classification algorithm can be implemented or programmed in the digital means 8. The input data of this classifier are one or more of the predictors mentioned above and/or in Table 2 above.
- this random forest has 5 decision trees, each making a decision independently of the others. The final decision is obtained by vote.
- the prediction of the random forest can then be a simple majority vote (Ensemble learning).
- the parameters of this classifier can be optimized using a search grid with constant steps.
- max_depth the depth (or maximum number of nodes) of each tree
- Tables 4 (for a bracelet) and 5 (for a pendant) below bring together precision (column 1) and specificity (column 2) data; the 3rd column (fl score) is the harmonic mean between precision and recall.
- the “support” column corresponds to the number of observations on which the calculation of the indicators is based.
- the macro average corresponds to the average of the two upper lines.
- the weighted average is the average of the two lines weighted by the support.
- - or the accelerometer can be used in combination with the gyroscope or with the barometer; - or the accelerometer can, as already explained above, be used with both the gyroscope and the barometer.
- FIG. 8 The abbreviations shown in Figure 8 correspond to: Etc: ExtraTrees Classifier (Extremely Randomized Trees) Trees, RF: Random Forest, xgb: Gradient Boosting Trees, rfjight: Random Forest limited to 5 trees, dt: Decision Trees, svm: Support Vector Machines knn: k-Nearest Neighbors, gnb: Gaussian Naive Bayes, mlp: MultiLayer Perception (Neural Network).
- Etc ExtraTrees Classifier (Extremely Randomized Trees) Trees
- RF Random Forest
- xgb Gradient Boosting Trees
- rfjight Random Forest limited to 5 trees
- dt Decision Trees
- svm Support Vector Machines
- knn k-Nearest Neighbors
- gnb Gaussian Naive Bayes
- mlp MultiLayer Perception (Neural Network).
- the 4 best classifiers (RF, ETC, XGB, RF_LIGHT), in particular the random forest, are of the set type, that is to say that they aggregate the decisions of several classifiers. This makes it possible to make the decision more robust for the application targeted in this application and to optimize the bias/variance compromise.
- a probability regression can be used, for example a Random Forest regression instead of a random forest classification. Unlike a classifier which provides a binary value (1 for fall or 0 for non-fall, or the opposite), regression provides a continuous value between 0 and 1 akin to a probability of fall. The decision to classify as a fall or not a fall is then made by setting a threshold on this probability. If the probability is greater than the threshold, then we trigger an alert, otherwise, we ignore the event.
- the advantage is that this allows the user to set the sensitivity threshold for fall detection themselves, depending on the intended application.
- the threshold can be programmed in means 8. It can be adjusted by an operator.
- the trigger threshold in a context of high risk of falling (very dependent people), we can set the trigger threshold very low to detect all falls. In the case of a low risk of falling (independent and healthy people), the threshold can be set higher to reduce the number of false alarms.
- the invention finds applications in particular:
- Figure 9 represents an embodiment of a method according to the invention, comprising 5 steps executed chronologically:
- steps S1 and S2 comprising steps S1 and S2
- steps S3 - S5 are preferably only activated in the event of a positive response from event detection in order to save system energy.
- step S1 (which allows the standard of the signal to be calculated) allows you to work with data independent of orientation; in certain applications, it is preferable to implement this step, in particular for senior people, who do not always wear an object, for example a pendant or a bracelet, which implements the invention in the same direction or on the same hand. Calculating the standard allows us to overcome this problem;
- step S2 contributes to a saving in terms of electrical power (battery in particular), because without it the calculation of the predictors and the classification must operate permanently.
- step S5 can trigger the emission of an alert signal 12 (figure 1).
- the “longest_strike_below_mean” predictor can be implemented using the following code:
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Physiology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Pathology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Dentistry (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Emergency Alarm Devices (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FR2207288A FR3137828B1 (fr) | 2022-07-15 | 2022-07-15 | Procédé et dispositif de détection de chute par capture inertielle |
| PCT/FR2023/051047 WO2024013447A1 (fr) | 2022-07-15 | 2023-07-07 | Procédé et dispositif de détection de chute par capture inertielle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| EP4554471A1 true EP4554471A1 (fr) | 2025-05-21 |
Family
ID=84569238
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| EP23750655.5A Pending EP4554471A1 (fr) | 2022-07-15 | 2023-07-07 | Procédé et dispositif de détection de chute par capture inertielle |
Country Status (3)
| Country | Link |
|---|---|
| EP (1) | EP4554471A1 (fr) |
| FR (1) | FR3137828B1 (fr) |
| WO (1) | WO2024013447A1 (fr) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN121524765B (zh) * | 2026-01-15 | 2026-03-31 | 成都信息工程大学 | 时序气压轨迹与惯性数据交叉验证的跌倒检测方法及系统 |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10629048B2 (en) * | 2017-09-29 | 2020-04-21 | Apple Inc. | Detecting falls using a mobile device |
| JP7418614B2 (ja) * | 2020-05-20 | 2024-01-19 | コーニンクレッカ フィリップス エヌ ヴェ | 生理学的感知を組み込んだ転倒検出器 |
-
2022
- 2022-07-15 FR FR2207288A patent/FR3137828B1/fr active Active
-
2023
- 2023-07-07 EP EP23750655.5A patent/EP4554471A1/fr active Pending
- 2023-07-07 WO PCT/FR2023/051047 patent/WO2024013447A1/fr not_active Ceased
Also Published As
| Publication number | Publication date |
|---|---|
| FR3137828A1 (fr) | 2024-01-19 |
| WO2024013447A1 (fr) | 2024-01-18 |
| FR3137828B1 (fr) | 2024-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP0589791B1 (fr) | Procédé et dispositif de mesure de ressemblance entre échantillons sonores | |
| EP2696344B1 (fr) | Procede et systeme pour detecter des evenements sonores dans un environnement donne | |
| EP3846087A1 (fr) | Procede et systeme de selection d'un modele d'apprentissage au sein d'une pluralite de modeles d'apprentissage | |
| FR2981561A1 (fr) | Procede de detection d'activite a capteur de mouvements, dispositif et programme d'ordinateur correspondants | |
| WO2011039283A1 (fr) | Systeme et procede de reconnaissance de gestes | |
| WO2010122056A2 (fr) | Systeme et methode pour detecter des evenements audio anormaux | |
| EP3245946B1 (fr) | Procédé de tri non supervisé en temps réel de potentiels d'action d'une pluralité de neurones biologiques | |
| EP3252563B1 (fr) | Détermination d'un contexte de mobilité d'un utilisateur porteur d'un équipement muni de capteurs inertiels | |
| WO2024013447A1 (fr) | Procédé et dispositif de détection de chute par capture inertielle | |
| EP3489831A1 (fr) | Procédé et dispositif de surveillance d'un processus générateur de données d'une métrique pour la prédiction d'anomalies | |
| EP4266281A1 (fr) | Dispositif électronique de surveillance d'un état de conscience d'un opérateur dans un aéronef, procédé et programme d'ordinateur associés | |
| WO2013030508A1 (fr) | Procédé de configuration d'un dispositif de détection à capteur, programme d'ordinateur et dispositif adaptatif correspondants | |
| CA2940380C (fr) | Determiner la severite d'une perturbation geomagnetique sur un reseau electrique a l'aide de mesures de similarite | |
| WO2016198786A1 (fr) | Procédé de traitement des mesures d'au moins un capteur électronique logé dans un appareil portable | |
| EP3773161B1 (fr) | Dispositif informatique de détection de troubles du rythme cardiaque | |
| EP2196821A1 (fr) | Procédé de filtrage cinématique temporel multidimensionel de plots radar, de tour d'antenne à tour d'antenne | |
| EP3543904A1 (fr) | Procédé de contrôle de détection de scènes et appareil correspondant | |
| WO2012062997A1 (fr) | Dispositif de détection à capteur, procédé de détection et programme d'ordinateur correspondants | |
| FR3089648A1 (fr) | Procede de detection non supervise d’attaques internes et systeme associe | |
| EP3505957A1 (fr) | Procede de traitement pour la detection, la caracterisation et le suivi multi-cibles et dispositif associe | |
| EP1979854A1 (fr) | Procede d'identification de postures d'une personne | |
| EP2881841A1 (fr) | Procede de reconnaissance continue de gestes d'un utilisateur d'un terminal mobile prehensible muni d'un ensemble capteur de mouvement, et dispositif associe | |
| FR2875626A1 (fr) | Procede de detection et de pistage de cibles ponctuelles, dans un systeme de surveillance optronique | |
| EP3737254A1 (fr) | Chaussure comprenant un accelerometre, ensemble et procede de detection de chute correspondant | |
| FR3066296A1 (fr) | Procede de controle de detection, avec filtrage, de scenes par un appareil, par exemple un appareil de communication sans fil, et appareil correspondant |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: UNKNOWN |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE |
|
| PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE |
|
| 17P | Request for examination filed |
Effective date: 20250117 |
|
| AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC ME MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
| DAV | Request for validation of the european patent (deleted) | ||
| DAX | Request for extension of the european patent (deleted) | ||
| STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: EXAMINATION IS IN PROGRESS |
|
| 17Q | First examination report despatched |
Effective date: 20260507 |