US20120095722A1 - Fall prevention - Google Patents

Fall prevention Download PDF

Info

Publication number
US20120095722A1
US20120095722A1 US13/377,419 US201013377419A US2012095722A1 US 20120095722 A1 US20120095722 A1 US 20120095722A1 US 201013377419 A US201013377419 A US 201013377419A US 2012095722 A1 US2012095722 A1 US 2012095722A1
Authority
US
United States
Prior art keywords
user
measurements
value
gait
estimating
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.)
Abandoned
Application number
US13/377,419
Inventor
Warner Rudolph Theophile Ten Kate
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Koninklijke Philips NV
Original Assignee
Koninklijke Philips Electronics NV
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Koninklijke Philips Electronics NV filed Critical Koninklijke Philips Electronics NV
Assigned to KONINKLIJKE PHILIPS ELECTRONICS N.V. reassignment KONINKLIJKE PHILIPS ELECTRONICS N.V. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: TEN KATE, WARNER RUDOLPH THEOPHILE
Publication of US20120095722A1 publication Critical patent/US20120095722A1/en
Abandoned legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01PMEASURING LINEAR OR ANGULAR SPEED, ACCELERATION, DECELERATION, OR SHOCK; INDICATING PRESENCE, ABSENCE, OR DIRECTION, OF MOVEMENT
    • G01P15/00Measuring acceleration; Measuring deceleration; Measuring shock, i.e. sudden change of acceleration
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/103Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/1116Determining posture transitions
    • A61B5/1117Fall detection
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/72Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235Details of waveform analysis
    • A61B5/7264Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00Measuring for diagnostic purposes; Identification of persons
    • A61B5/72Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7271Specific aspects of physiological measurement analysis
    • A61B5/7282Event detection, e.g. detecting unique waveforms indicative of a medical condition
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C22/00Measuring distance traversed on the ground by vehicles, persons, animals or other moving solid bodies, e.g. using odometers, using pedometers
    • G01C22/006Pedometers
    • GPHYSICS
    • G08SIGNALLING
    • G08BSIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/02Alarms for ensuring the safety of persons
    • G08B21/04Alarms for ensuring the safety of persons responsive to non-activity, e.g. of elderly persons
    • G08B21/0438Sensor means for detecting
    • G08B21/0446Sensor means for detecting worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems

Definitions

  • the invention relates to a method and device for monitoring the motion of a user, and in particular to a method and device for determining a fall risk for a user.
  • a fall is defined as a sudden, uncontrolled and unintentional downward displacement of the body to the ground.
  • Exemplary fall detectors can comprise personal help buttons (PHBs) or worn and/or environment-based automatic detectors.
  • Automatic fall detectors comprise one or a set of sensors that measure the movement of the user, and a processor that compares the measured or processed signals with predetermined thresholds in order to detect a fall.
  • automatic fall detectors store a set of predetermined threshold values and/or parameter sets. When the detector is activated, movement data obtained from the sensors (such as, for example, an accelerometer) will be continuously transformed and processed, and then compared with those parameter sets to determine if a fall event occurs.
  • fall detectors are useful, they do not actually prevent falling, and only provide a warning or alarm in the event that a user already has fallen.
  • users that are insecure during walking for example caused or enhanced by a fear of falling, by fatigue in the muscles, by frequently multi-tasking (i.e. they are carrying items when walking, talking to their grandchild, etc, or that move in places where there is dim lighting, a wet or irregular ground surface—such as loose carpet, electricity wires, toys, tools, and other hazards) or that are under medication that may affect balance or concentration, can be assisted by a device for fall prevention that decreases the actual risk of falling, or at least alerts the user that they are at a higher risk of falling at a particular time, and makes them feel more safe.
  • a device for fall prevention that decreases the actual risk of falling, or at least alerts the user that they are at a higher risk of falling at a particular time, and makes them feel more safe.
  • a method of determining a fall risk of a user comprising collecting measurements of the motion of the user; estimating a value for a parameter related to the gait of the user from the measurements; and determining a fall risk for the user from a comparison of the estimated value with a normal value for the parameter determined from motion of the user in which the user is at their normal risk of falling.
  • the step of determining a fall risk comprises weighting the comparison between the estimated value and the normal value according to a standard deviation of the normal value.
  • the estimated value is determined from motion of the user over a period of time that is shorter than the period of time over which the normal value is determined.
  • the step of estimating comprises identifying a step boundary in the collected measurements.
  • the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold.
  • the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold, apart from a subset of the measurements whose magnitude is less than the threshold, provided that the subset covers a time period less than a time threshold.
  • the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements, wherein the first collected measurement in the collected measurements whose magnitude exceeds a first threshold denotes the first measurement in a cluster and wherein the first collected measurement after the first measurement in the cluster whose magnitude falls below a second threshold denotes the last measurement in the cluster, provided that the last measurement is more than a minimum period after the first measurement.
  • the step of identifying step boundaries can further comprise identifying the step boundary as the measurement in each cluster with the highest magnitude.
  • the parameter related to the gait of the user comprises a step size and the step of estimating a value for the parameter comprises integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries.
  • the step of estimating a value for the parameter comprises omitting the average velocity of the user from the integration such that the step size is determined based on the variation of the velocity.
  • the parameter related to the gait of the user comprises, or additionally comprises, a forward step size and the step of estimating a value for the parameter comprises integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries to give a start and end position for a step; and determining the forward step size as the norm of the vector connecting the start and end positions.
  • the parameter related to the gait of the user additionally comprises a lateral step size and the step of estimating a value for the parameter further comprises defining a straight line between the start and end positions; integrating collected measurements occurring during the step to give a series of positions during the step; determining the distance between each position and the straight line; and determining the lateral step size as the maximum distance in this series.
  • the method further comprises a calibration step that includes collecting measurements of the motion of the user when the user is at their normal risk of falling; and estimating the normal value for the parameter related to the gait of the user from the collected measurements.
  • the step of estimating comprises estimating values for a plurality of parameters related to the gait of the user from the measurements
  • the step of determining a fall risk comprises comparing the estimated values with values of the parameters determined from motion of the user in which the user is at a low risk of falling.
  • the parameter or parameters related to the gait of the user is/are selected from step size, step width, step time, double support time, gait velocity, cadence, average step size, average step time, average double support time, average gait velocity and average cadence.
  • a fall prevention device comprising at least one sensor for collecting measurements of the motion of a user of the device; and a processor for estimating a value for a parameter related to the gait of the user from the measurements, and for determining a fall risk for the user from a comparison of the estimated value with a value of the parameter determined from motion of the user in which the user is at their normal risk of falling.
  • a computer program product comprising computer-readable code that, when executed on a suitable computer or processor, is configured to cause the computer or processor to perform the steps in the method described above.
  • determining gait parameters including step boundaries, a step size, a forward step size and/or a lateral step size as described above and in the following Detailed Description.
  • FIG. 1 shows a fall prevention device according to the invention being worn by a user
  • FIG. 2 shows the fall prevention device of FIG. 1 in more detail
  • FIG. 3 is a flow chart illustrating the steps in a method according to the invention.
  • FIG. 4 is a plot illustrating measurements from an accelerometer that is in a pendant worn around a user's neck
  • FIG. 5 is a plot illustrating measurements from an accelerometer that is worn on a user's ankle.
  • FIG. 6 is a plot illustrating the derivative of measurements from an accelerometer that is worn on a user's ankle.
  • the fall prevention device comprises a single unit that is worn by a user.
  • FIG. 1 shows a fall prevention device 2 in the form of a pendant being worn around the neck of a user 4 and FIG. 2 shows the fall prevention device 2 in more detail.
  • the fall prevention device 2 comprises three sensors, an accelerometer 6 , a magnetometer 8 and a gyroscope 10 , which are connected to a processor 12 .
  • the processor 12 receives measurements from the sensors 6 , 8 , 10 and processes the measurements to determine if the user 4 of the fall prevention device 2 is at a higher risk of falling.
  • the processor 12 processes the measurements from the accelerometer 6 , magnetometer 8 and gyroscope 10 to determine the orientation of the fall prevention device 2 , and processes the measurements from the accelerometer 6 (using the determined orientation) to determine parameters relating to the motion (and specifically the gait) of the user 2 .
  • the three sensors are included by way of example only. Using sensor fusion, the measurements of the three sensors can complement each other for an optimal estimate of the orientation, as is known to those skilled in the art. Other sensors, for example a barometer and GPS receiver, can be added to further improve the accuracy of the parameters estimated by the device 2 . It will be further appreciated that sensors can be omitted to reduce the power consumption and cost of the device 2 —for example in a minimal device 2 , only the accelerometer 6 is present, in which case it is only possible to estimate the inclination of the device 2 .
  • the fall prevention device 2 also comprises an alarm 14 that can be activated by the processor 12 to warn or notify the user 4 that they are at an increased risk of falling.
  • the alarm 14 can also comprise (or be replaced by) some means that provides the user 4 with an indication of their current risk of falling, even if they are not currently at a high risk of falling.
  • the fall prevention device 2 could be provided with means for providing vibrotactile or auditory feedback, or with a series of lights (or a light that can show different colors) which can be illuminated to indicate the current fall risk (for example there could be lights associated with the user 4 being at low, medium and high risk of falling).
  • the fall prevention device 2 also comprises transmitter circuitry 16 that allows the fall prevention device 2 to transmit an alarm or warning signal to a base station associated with the fall prevention device 2 (which can then issue an alarm or summon help from a healthcare provider or the emergency services).
  • the base station can also carry out further processing of the sensor measurements, and/or store the sensor measurements for later analysis.
  • the transmitter circuitry 16 may be omitted if the fall prevention device 2 does not need to contact a base station to issue an alarm or summon help from a healthcare provider (for example if the fall prevention device 2 can contact the healthcare provider by using sound).
  • the fall prevention device 2 can include a memory unit for storing the sensor measurements for later analysis.
  • gait parameters can include measures such as step size, step width, step time, double support time (i.e. the time that both feet are in contact with the ground), gait velocity, and cadence.
  • the gait parameters can also include averages of the above measures over a few steps, so for example an average step size, average step width, average step time, average double support time, average gait velocity and average cadence.
  • the fall prevention device 2 determines values for gait parameters from sensor measurements covering a short period of time and compares these values with normal values for the user 4 (i.e. values that are obtained when the user 4 is at their normal (i.e. preferably a low or minimal) risk of falling).
  • the fall prevention device 2 can determine various gait parameters from sensor measurements covering, say, 12 steps (so 6 strides).
  • the normal values for the user 4 can be obtained by collecting measurements while the user 4 is walking steadily for a period of time (for example a minute) or for a certain number of steps (for example 40 steps).
  • step 101 the sensors 6 , 8 , 10 take measurements of the motion of the user 4 and in step 103 , the processor 12 estimates values for the required gait parameters from the measurements. In step 105 , the processor 12 then compares these estimates to usual values of these gait parameters for the user 4 to determine the user's dynamic risk of falling. As indicated above, the usual values correspond to those observed when the user 4 is at their normal (i.e. preferably a minimal or low) risk of falling.
  • the normal or usual values for the gait parameters can be obtained during a calibration session before the fall prevention device 2 is used (for example the user 4 can wear the fall prevention device 2 while it is in a calibration mode, and the fall prevention device 2 can determine values for each gait parameter while the user 4 is walking normally).
  • the comparison between the estimated values and the normal values is weighted according to the standard deviation of the normal gait parameter values.
  • ⁇ e represents the calibration mean (i.e. the mean of the normal values for a particular parameter)
  • represents the standard deviation in that calibration mean
  • a represents the currently observed parameter value
  • the processor 12 estimates values for an average step size and an average step width.
  • other combinations of gait parameters can be used.
  • the processor 12 in determining estimates for specific gait parameters will be described further below.
  • the way in which the averages a and ⁇ are calculated relates them to the variance in step size rather than to the mean, as will be explained further below.
  • the invention effectively compares current and normal variances. It should be noted that a is typically averaged over a shorter period than ⁇ .
  • the fall prevention device 2 is in the form of a pendant to be worn around the neck of a user 4
  • the invention can be implemented in alternative forms that are to be worn on different parts of the body of the user 4 , such as at the waist or on the ankle of the user 4 .
  • the gyroscope 10 , the magnetometer 8 or both the magnetometer 8 and gyroscope 10 can be omitted from the device 2 .
  • the fall prevention device 2 is a pendant, it has been found that the gait variation estimate is less sensitive without the gyroscope 10 , but it is still capable of detecting deviations from the normal gait.
  • step size In order to obtain an estimate of the step size, a number of processing steps are required. In particular, it is necessary to estimate the step boundaries and the orientation of the accelerometer (so that the accelerometer measurements can be transformed to Earth coordinates) in order to estimate the step or stride size.
  • step boundaries are defined as the moment of heel strike (HS), i.e. when the swing leg is touching the ground again, starting the stance phase), since missing a boundary will cause a significant deviation in a, hence leading to an alarm.
  • HS moment of heel strike
  • Step boundaries can also be used to estimate stepping time, which is another gait parameter, and, when combined with step size, allows an estimate of walking velocity to be determined.
  • step boundaries are usually found by observing the ‘zero’ crossing of the vertical acceleration.
  • the actual “crossing” will be through 1 g (9.81 ms ⁇ 2 ) since gravity is always acting on the user.
  • the vertical acceleration will be known after the orientation of the accelerometer is determined—although for rigidly mounted accelerometers, the accelerometer reading along the coordinate axis corresponding to vertical usually provides a sufficient approximation.
  • the fall prevention device 2 in the form of a pendant, the fall prevention device 2 (and therefore the accelerometer 6 ) is free to move relative to the user 4 , which means that the accelerometer coordinate system also moves relative to the user 4 .
  • step boundaries i.e. heel strikes
  • the norm of the signal from the accelerometer 6 is calculated.
  • the peaks in this signal serve as boundary markers for each step, which closely link to the heel strike (HS), as shown in FIG. 4 .
  • the peaks are found using a two-step procedure. First, so-called clusters are identified. Second, the maximum value in each cluster is identified as the step boundary. The clusters are found as the range of samples that are above a certain threshold (typically 2 ms ⁇ 2 above gravity, i.e. ⁇ 12 ms ⁇ 2 ), where a small gap of samples not surpassing that threshold is permitted (typically 0.3 times a typical step time (which is around 0.5 seconds), i.e. 0.15 sec).
  • a certain threshold typically 2 ms ⁇ 2 above gravity, i.e. ⁇ 12 ms ⁇ 2
  • a small gap of samples not surpassing that threshold is permitted (typically 0.3 times a typical step time (which is around 0.5 seconds), i.e. 0.15 sec).
  • This algorithm works for accelerometers 6 that are attached to the upper body of the user 4 . If the accelerometer 6 is attached at or to the lower body of the user 4 (for example the ankle), then usually two clusters appear per step, or actually per stride, as shown in FIG. 5 . Since the accelerometer signal is observed at the ankle, the period of a stride (i.e. a step with both the left foot and the right foot) is seen.
  • One cluster corresponds to the lifting of the foot, and the other corresponds to the heel strike. It can be difficult to decide which of the two is the heel strike. However, it can also be observed that a single minimum occurs per step, and this can be used instead for detecting the step boundaries.
  • FIG. 5 also shows another way to identify the clusters. Instead of a single threshold (with a small gap being permitted), two thresholds are used. They introduce a hysteresis—surpassing one threshold indicates the start of a cluster, and falling below the second threshold indicates the end of a cluster, provided that the measurements fall below the second threshold after a minimum duration from the first measurement in the cluster.
  • the minimum duration is derived from the step time, as described above.
  • the unique minimum shown for each stride in FIG. 5 corresponds to the swing phase and it is less favorable to use this as the step boundary. Instead, it is preferable to use the heel strike as the step boundary for several reasons. Firstly, this is a clearly defined event. Secondly, when measuring at the feet or ankle, at heel strike the velocity relative to the ground is zero, which can be used in estimating the movement. Thirdly, at heel strike, the acceleration in the horizontal direction is low, which leads to lower errors when (double) integrating the acceleration for estimating the step size. The sample values at the beginning of the integration have a large influence on the outcome, so large values can cause a bias in the total outcome.
  • a better algorithm for detecting step boundaries when the accelerometer 6 is located on the lower body of the user 4 is found by observing the derivative of the acceleration, as shown in FIG. 6 .
  • the boundaries are more clear, and they associate with the heel strike event.
  • the processor 12 in the fall prevention device 2 needs to know the location used.
  • the user 4 may be able to select an appropriate location from a list of locations offered by the processor 12 , or, alternatively, the processor 12 can execute a classifier algorithm to detect the location (and therefore the algorithm to use to detect the step boundary) based on the particular patterns in the measurements of the accelerometer 6 .
  • Estimating accelerometer orientation The orientation of the accelerometer 6 (and therefore the fall prevention device 2 ) can be estimated from the direction that gravity appears in the accelerometer coordinate system.
  • the orientation follows through the vector dot product of the measured gravity direction and the z-axis, i.e. through the z-component of the measured gravity (as is known, when using normalized values, the dot product equals the cosine of the enclosed angle).
  • the accelerometer 6 Since the accelerometer 6 is sensitive to both acceleration due to gravity and acceleration due to movement, a filter is needed in order to estimate which component is due to gravity. Usually, a gyroscope 10 is added to measure the angular rotation speed and to correct the measured acceleration correspondingly. However, if the movements of the user 4 do not cause fast rotations of the fall prevention device 2 , the gyroscope 10 can be omitted.
  • the gravity component can be found as the low-pass filtered component of the signal from the accelerometer 6 . Since causality constraints in the filter design introduce a delay, the filtered acceleration signal needs to be corrected for this delay.
  • the inclination or orientation of the fall prevention device 2 can be estimated. However, this inclination or orientation does not provide information about the horizontal direction of the device 2 (i.e. which way the device 2 is facing), so the magnetometer 8 can be used to determine the horizontal orientation of the device 2 . Gyroscopes can also be used in the estimation of horizontal orientation.
  • Orientation can be expressed in different ways, of which Euler angles and Euler parameters are the most commonly used. They are usually implemented through matrices or quaternions. The algebras are isomorphic, and provide a way to transform the measured values (acceleration etc.) as expressed in the accelerometer's local coordinate system to the global (Earth) coordinate system.
  • Estimating step or stride sizes can be computed by double integrating (with respect to time) the horizontal components of these transformed signals.
  • the boundaries for the integrations are given by the estimated step boundaries.
  • the integration constants, velocity and position at the beginning of the step, are both set to zero.
  • the step size p follows as the integral of velocity v over step time:
  • HS 0 and HS 1 are the times of subsequent heel strikes (HS) and v and ⁇ v denote the average velocity and its deviation respectively.
  • ⁇ 0 2 ⁇ 0 2 and var(p) ⁇ 2 ⁇ 2 .
  • the values ⁇ 0 and ⁇ 0 are user dependent and need to be set for each user in a calibration phase. Note, however, that since the average velocity v is not accounted for in the calculations, the method is insensitive to the actual walking speed of the user 4 .
  • v ⁇ T in the expression for ⁇ p above has been neglected and introduces an error. It is proportional to the average velocity and the deviation in step time. As a refinement, it could be included in the estimate of ⁇ p.
  • band-pass filtering or high-pass filtering
  • the signals from the accelerometer 6 preferably using a linear phase filter so that the wave form of the signal is maintained.
  • Typical cut-off frequencies are 0.1 Hz and 20-40 Hz (where the actual application of the upper cut-off frequency also depends on the used sampling frequency). This band-pass filtering is not required.
  • a measure that has been found to be particularly effective is the de-trending per integration interval (step or stride) of the acceleration. This amounts to requiring that the average acceleration (from the start heel strike to the end heel strike) is zero. In other words, the velocity after integration equals the velocity at the start of integration (which is zero).
  • the de-trended acceleration is obtained by subtracting the integrated acceleration, divided by the step duration, from the measured acceleration. The de-trended acceleration is used to compute the step size as described above.
  • the start position (which is the origin, by definition) and the end position are computed as explained above. Their difference spans a 2D vector in the horizontal plane. Then, the forward step size is defined as the norm of this 2D vector, i.e. as the distance of the end position from the start position (it should be noted that, as a consequence, the above ⁇ p are always positive). Subsequently, a straight line is defined between the start position and the end position and the distance is determined between this line and each point from the double integration of the accelerometer measurements during the step. The maximum distance in this series is taken as the lateral step size.
  • Step time follows as the duration between the estimated step boundaries, and velocity as the ratio between step size and step time (in the event that the device 2 is located on a foot of the user 4 ).
  • step size can be computed when the device 2 is on the upper body of the user 4 , yielding twice as fast an average.
  • step size can be used that do not require the double integration of the horizontal acceleration measurements.
  • These methods include the inverted pendulum model [Zijlstra & H of, Displacement of the pelvis during human walking, Gait and Posture 6, 1997, 249-267] and the 4 th power root of the difference between maximum and minimum vertical acceleration [Weinberg, Using the ADXL202 in Pedometer and Personal Navigation Applications, Application Note AN-602, Analog Devices, 2002].
  • the collection and processing of the sensor measurements is performed in a single unit.
  • the processing of the measurements can be performed in a unit that is remote from the sensors, in which case the fall prevention device 2 will comprise a sensor unit to be worn by the user 4 that transmits the sensor measurements to the remote unit.
  • the sensor unit there is no need for the sensor unit to include a dedicated processor.
  • the algorithms described above that are used to determine various gait parameters can be used in applications other than fall prevention. For example, they can be used in activity monitoring and fitness applications such as endurance coaching (for example to support the keeping of cadence during jogging).
  • the algorithms may be used in step counters or devices that support injury prevention, for example during jogging. Further the algorithms may be used in a device or system that uses gait parameters as a biometric for identifying an individual.
  • a computer program may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Abstract

There is provided a method of determining a fall risk of a user, the method comprising collecting measurements of the motion of the user, estimating a value for a parameter related to the gait of the user from the measurements, and determining a fall risk for the user from a comparison of the estimated value with a normal value for the parameter determined from motion of the user in which the user is at their normal risk of falling.

Description

    TECHNICAL FIELD OF THE INVENTION
  • The invention relates to a method and device for monitoring the motion of a user, and in particular to a method and device for determining a fall risk for a user.
  • BACKGROUND TO THE INVENTION
  • Falls affect millions of people each year and result in significant injuries, particularly in the elderly. In fact, it has been estimated that falls are one of the top three causes of death in elderly people.
  • A fall is defined as a sudden, uncontrolled and unintentional downward displacement of the body to the ground. There are currently some fall detection systems available that detect these falls and allow the user to obtain assistance manually or automatically if a fall occurs. Exemplary fall detectors can comprise personal help buttons (PHBs) or worn and/or environment-based automatic detectors.
  • Automatic fall detectors comprise one or a set of sensors that measure the movement of the user, and a processor that compares the measured or processed signals with predetermined thresholds in order to detect a fall. In particular, automatic fall detectors store a set of predetermined threshold values and/or parameter sets. When the detector is activated, movement data obtained from the sensors (such as, for example, an accelerometer) will be continuously transformed and processed, and then compared with those parameter sets to determine if a fall event occurs.
  • Although these fall detectors are useful, they do not actually prevent falling, and only provide a warning or alarm in the event that a user already has fallen.
  • However, users that are insecure during walking, for example caused or enhanced by a fear of falling, by fatigue in the muscles, by frequently multi-tasking (i.e. they are carrying items when walking, talking to their grandchild, etc, or that move in places where there is dim lighting, a wet or irregular ground surface—such as loose carpet, electricity wires, toys, tools, and other hazards) or that are under medication that may affect balance or concentration, can be assisted by a device for fall prevention that decreases the actual risk of falling, or at least alerts the user that they are at a higher risk of falling at a particular time, and makes them feel more safe.
  • There is therefore a need for a method and device that can determine an instantaneous fall risk for a user.
  • SUMMARY OF THE INVENTION
  • According to a first aspect of the invention, there is provided a method of determining a fall risk of a user, the method comprising collecting measurements of the motion of the user; estimating a value for a parameter related to the gait of the user from the measurements; and determining a fall risk for the user from a comparison of the estimated value with a normal value for the parameter determined from motion of the user in which the user is at their normal risk of falling.
  • In a preferred embodiment, the step of determining a fall risk comprises weighting the comparison between the estimated value and the normal value according to a standard deviation of the normal value.
  • Preferably, the estimated value is determined from motion of the user over a period of time that is shorter than the period of time over which the normal value is determined.
  • In a preferred embodiment, the step of estimating comprises identifying a step boundary in the collected measurements.
  • In one embodiment, the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold.
  • In an alternative embodiment, the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold, apart from a subset of the measurements whose magnitude is less than the threshold, provided that the subset covers a time period less than a time threshold.
  • In another alternative embodiment, the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements, wherein the first collected measurement in the collected measurements whose magnitude exceeds a first threshold denotes the first measurement in a cluster and wherein the first collected measurement after the first measurement in the cluster whose magnitude falls below a second threshold denotes the last measurement in the cluster, provided that the last measurement is more than a minimum period after the first measurement.
  • In any of these alternative embodiments, the step of identifying step boundaries can further comprise identifying the step boundary as the measurement in each cluster with the highest magnitude.
  • In a preferred embodiment, the parameter related to the gait of the user comprises a step size and the step of estimating a value for the parameter comprises integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries.
  • In a preferred embodiment, the step of estimating a value for the parameter comprises omitting the average velocity of the user from the integration such that the step size is determined based on the variation of the velocity.
  • In one embodiment, the parameter related to the gait of the user comprises, or additionally comprises, a forward step size and the step of estimating a value for the parameter comprises integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries to give a start and end position for a step; and determining the forward step size as the norm of the vector connecting the start and end positions.
  • Preferably, the parameter related to the gait of the user additionally comprises a lateral step size and the step of estimating a value for the parameter further comprises defining a straight line between the start and end positions; integrating collected measurements occurring during the step to give a series of positions during the step; determining the distance between each position and the straight line; and determining the lateral step size as the maximum distance in this series.
  • In preferred embodiments, the method further comprises a calibration step that includes collecting measurements of the motion of the user when the user is at their normal risk of falling; and estimating the normal value for the parameter related to the gait of the user from the collected measurements.
  • Preferably, the step of estimating comprises estimating values for a plurality of parameters related to the gait of the user from the measurements, and the step of determining a fall risk comprises comparing the estimated values with values of the parameters determined from motion of the user in which the user is at a low risk of falling.
  • Preferably, the parameter or parameters related to the gait of the user is/are selected from step size, step width, step time, double support time, gait velocity, cadence, average step size, average step time, average double support time, average gait velocity and average cadence.
  • In accordance with a second aspect of the invention, there is provided a fall prevention device, comprising at least one sensor for collecting measurements of the motion of a user of the device; and a processor for estimating a value for a parameter related to the gait of the user from the measurements, and for determining a fall risk for the user from a comparison of the estimated value with a value of the parameter determined from motion of the user in which the user is at their normal risk of falling.
  • In accordance with a third aspect of the invention, there is provided a computer program product comprising computer-readable code that, when executed on a suitable computer or processor, is configured to cause the computer or processor to perform the steps in the method described above.
  • In accordance with alternative aspects of the invention, there are provided methods for determining gait parameters, including step boundaries, a step size, a forward step size and/or a lateral step size as described above and in the following Detailed Description.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • Embodiments of the invention will be described, by way of example only, with reference to the following drawings, in which:
  • FIG. 1 shows a fall prevention device according to the invention being worn by a user;
  • FIG. 2 shows the fall prevention device of FIG. 1 in more detail;
  • FIG. 3 is a flow chart illustrating the steps in a method according to the invention;
  • FIG. 4 is a plot illustrating measurements from an accelerometer that is in a pendant worn around a user's neck;
  • FIG. 5 is a plot illustrating measurements from an accelerometer that is worn on a user's ankle; and
  • FIG. 6 is a plot illustrating the derivative of measurements from an accelerometer that is worn on a user's ankle.
  • DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
  • Although the invention will be described in terms of a method and device that is used for fall prevention, it will be appreciated that the method and device can be provided with additional functionality so that the device can also be used as a fall detector.
  • In a preferred embodiment of the invention, the fall prevention device comprises a single unit that is worn by a user.
  • FIG. 1 shows a fall prevention device 2 in the form of a pendant being worn around the neck of a user 4 and FIG. 2 shows the fall prevention device 2 in more detail.
  • The fall prevention device 2 comprises three sensors, an accelerometer 6, a magnetometer 8 and a gyroscope 10, which are connected to a processor 12. The processor 12 receives measurements from the sensors 6, 8, 10 and processes the measurements to determine if the user 4 of the fall prevention device 2 is at a higher risk of falling.
  • In particular, the processor 12 processes the measurements from the accelerometer 6, magnetometer 8 and gyroscope 10 to determine the orientation of the fall prevention device 2, and processes the measurements from the accelerometer 6 (using the determined orientation) to determine parameters relating to the motion (and specifically the gait) of the user 2.
  • It should be understood that the three sensors are included by way of example only. Using sensor fusion, the measurements of the three sensors can complement each other for an optimal estimate of the orientation, as is known to those skilled in the art. Other sensors, for example a barometer and GPS receiver, can be added to further improve the accuracy of the parameters estimated by the device 2. It will be further appreciated that sensors can be omitted to reduce the power consumption and cost of the device 2—for example in a minimal device 2, only the accelerometer 6 is present, in which case it is only possible to estimate the inclination of the device 2.
  • The fall prevention device 2 also comprises an alarm 14 that can be activated by the processor 12 to warn or notify the user 4 that they are at an increased risk of falling. The alarm 14 can also comprise (or be replaced by) some means that provides the user 4 with an indication of their current risk of falling, even if they are not currently at a high risk of falling. For example, the fall prevention device 2 could be provided with means for providing vibrotactile or auditory feedback, or with a series of lights (or a light that can show different colors) which can be illuminated to indicate the current fall risk (for example there could be lights associated with the user 4 being at low, medium and high risk of falling).
  • The fall prevention device 2 also comprises transmitter circuitry 16 that allows the fall prevention device 2 to transmit an alarm or warning signal to a base station associated with the fall prevention device 2 (which can then issue an alarm or summon help from a healthcare provider or the emergency services). The base station can also carry out further processing of the sensor measurements, and/or store the sensor measurements for later analysis. In alternative embodiments, the transmitter circuitry 16 may be omitted if the fall prevention device 2 does not need to contact a base station to issue an alarm or summon help from a healthcare provider (for example if the fall prevention device 2 can contact the healthcare provider by using sound).
  • In some embodiments of the invention (not represented by the device 2 shown in FIG. 2), the fall prevention device 2 can include a memory unit for storing the sensor measurements for later analysis.
  • It is known that a large variance in gait parameters (i.e. parameters associated with the gait of a user) corresponds to a high risk of falling. However, this is a static relationship; the variance is an average obtained during general walking, and is related to an average risk of falling, which does not necessarily provide any information about the current risk of falling for the user 4. Gait parameters can include measures such as step size, step width, step time, double support time (i.e. the time that both feet are in contact with the ground), gait velocity, and cadence. The gait parameters can also include averages of the above measures over a few steps, so for example an average step size, average step width, average step time, average double support time, average gait velocity and average cadence.
  • In accordance with the invention, the fall prevention device 2 determines values for gait parameters from sensor measurements covering a short period of time and compares these values with normal values for the user 4 (i.e. values that are obtained when the user 4 is at their normal (i.e. preferably a low or minimal) risk of falling). In particular, the fall prevention device 2 can determine various gait parameters from sensor measurements covering, say, 12 steps (so 6 strides). The normal values for the user 4 can be obtained by collecting measurements while the user 4 is walking steadily for a period of time (for example a minute) or for a certain number of steps (for example 40 steps).
  • A method of determining a dynamic risk of falling in accordance with the invention is shown in FIG. 3. In step 101, the sensors 6, 8, 10 take measurements of the motion of the user 4 and in step 103, the processor 12 estimates values for the required gait parameters from the measurements. In step 105, the processor 12 then compares these estimates to usual values of these gait parameters for the user 4 to determine the user's dynamic risk of falling. As indicated above, the usual values correspond to those observed when the user 4 is at their normal (i.e. preferably a minimal or low) risk of falling.
  • The normal or usual values for the gait parameters can be obtained during a calibration session before the fall prevention device 2 is used (for example the user 4 can wear the fall prevention device 2 while it is in a calibration mode, and the fall prevention device 2 can determine values for each gait parameter while the user 4 is walking normally).
  • In a preferred embodiment of the invention, the comparison between the estimated values and the normal values is weighted according to the standard deviation of the normal gait parameter values.
  • For example, if μ epresents the calibration mean (i.e. the mean of the normal values for a particular parameter), σ represents the standard deviation in that calibration mean, and a represents the currently observed parameter value, a deviation is signaled if
  • a - μ σ ( 1 )
  • exceeds a threshold.
  • For example,
  • exp [ - ( a - μ ) 2 2 σ 2 ] ( 2 )
  • maps to a value between 0 and 1, where 1 indicates normal gait (for that user), and a deviation from normal gait is signaled (i.e. the user is at a higher risk of falling) if the result falls below a threshold, for example 0.7.
  • It will be appreciated that the values for μ and σ are usually dependent on the user and will need to be set independently for each user.
  • In one embodiment of the invention, the processor 12 estimates values for an average step size and an average step width. However, in other embodiments of the invention, other combinations of gait parameters can be used.
  • The operation of the processor 12 in determining estimates for specific gait parameters will be described further below. As it will be shown, in the embodiment in which the fall prevention device 2 is a pendant located around the neck of the user 4, the way in which the averages a and μ are calculated relates them to the variance in step size rather than to the mean, as will be explained further below. Thus, the invention effectively compares current and normal variances. It should be noted that a is typically averaged over a shorter period than μ.
  • In addition, although in the preferred embodiment of the invention the fall prevention device 2 is in the form of a pendant to be worn around the neck of a user 4, it will be appreciated that the invention can be implemented in alternative forms that are to be worn on different parts of the body of the user 4, such as at the waist or on the ankle of the user 4. As described further below, in these embodiments, it is necessary to modify the processing used to determine the gait parameters from the sensor measurements.
  • Moreover, depending on the particular gait parameters that the processor 12 determines from the sensor measurements (and if the movements of the user 4 do not cause fast rotations of the fall prevention device 2), the gyroscope 10, the magnetometer 8 or both the magnetometer 8 and gyroscope 10 can be omitted from the device 2. In the embodiment in which the fall prevention device 2 is a pendant, it has been found that the gait variation estimate is less sensitive without the gyroscope 10, but it is still capable of detecting deviations from the normal gait.
  • Estimating the Step Size
  • In order to obtain an estimate of the step size, a number of processing steps are required. In particular, it is necessary to estimate the step boundaries and the orientation of the accelerometer (so that the accelerometer measurements can be transformed to Earth coordinates) in order to estimate the step or stride size.
  • Estimating step boundaries—It is important to accurately estimate step boundaries (which are defined as the moment of heel strike (HS), i.e. when the swing leg is touching the ground again, starting the stance phase), since missing a boundary will cause a significant deviation in a, hence leading to an alarm.
  • Step boundaries can also be used to estimate stepping time, which is another gait parameter, and, when combined with step size, allows an estimate of walking velocity to be determined.
  • For accelerometers that are rigidly attached to the upper body of a user, step boundaries are usually found by observing the ‘zero’ crossing of the vertical acceleration. Of course, it will be appreciated that the actual “crossing” will be through 1 g (9.81 ms−2) since gravity is always acting on the user. The vertical acceleration will be known after the orientation of the accelerometer is determined—although for rigidly mounted accelerometers, the accelerometer reading along the coordinate axis corresponding to vertical usually provides a sufficient approximation.
  • However, in the preferred embodiment of the invention in which the fall prevention device 2 is in the form of a pendant, the fall prevention device 2 (and therefore the accelerometer 6) is free to move relative to the user 4, which means that the accelerometer coordinate system also moves relative to the user 4.
  • Therefore, it is preferred to modify the detection of step boundaries (i.e. heel strikes) as follows. Firstly, the norm of the signal from the accelerometer 6 is calculated. Then, the peaks in this signal serve as boundary markers for each step, which closely link to the heel strike (HS), as shown in FIG. 4.
  • The peaks are found using a two-step procedure. First, so-called clusters are identified. Second, the maximum value in each cluster is identified as the step boundary. The clusters are found as the range of samples that are above a certain threshold (typically 2 ms−2 above gravity, i.e. ˜12 ms−2), where a small gap of samples not surpassing that threshold is permitted (typically 0.3 times a typical step time (which is around 0.5 seconds), i.e. 0.15 sec).
  • This algorithm works for accelerometers 6 that are attached to the upper body of the user 4. If the accelerometer 6 is attached at or to the lower body of the user 4 (for example the ankle), then usually two clusters appear per step, or actually per stride, as shown in FIG. 5. Since the accelerometer signal is observed at the ankle, the period of a stride (i.e. a step with both the left foot and the right foot) is seen.
  • One cluster corresponds to the lifting of the foot, and the other corresponds to the heel strike. It can be difficult to decide which of the two is the heel strike. However, it can also be observed that a single minimum occurs per step, and this can be used instead for detecting the step boundaries.
  • FIG. 5 also shows another way to identify the clusters. Instead of a single threshold (with a small gap being permitted), two thresholds are used. They introduce a hysteresis—surpassing one threshold indicates the start of a cluster, and falling below the second threshold indicates the end of a cluster, provided that the measurements fall below the second threshold after a minimum duration from the first measurement in the cluster. Preferably, the minimum duration is derived from the step time, as described above.
  • The unique minimum shown for each stride in FIG. 5 corresponds to the swing phase and it is less favorable to use this as the step boundary. Instead, it is preferable to use the heel strike as the step boundary for several reasons. Firstly, this is a clearly defined event. Secondly, when measuring at the feet or ankle, at heel strike the velocity relative to the ground is zero, which can be used in estimating the movement. Thirdly, at heel strike, the acceleration in the horizontal direction is low, which leads to lower errors when (double) integrating the acceleration for estimating the step size. The sample values at the beginning of the integration have a large influence on the outcome, so large values can cause a bias in the total outcome.
  • A better algorithm for detecting step boundaries when the accelerometer 6 is located on the lower body of the user 4 is found by observing the derivative of the acceleration, as shown in FIG. 6. The boundaries are more clear, and they associate with the heel strike event.
  • It will be appreciated that as different algorithms can be used for different fall prevention device 2 locations on the user's body, the processor 12 in the fall prevention device 2 needs to know the location used. The user 4 may be able to select an appropriate location from a list of locations offered by the processor 12, or, alternatively, the processor 12 can execute a classifier algorithm to detect the location (and therefore the algorithm to use to detect the step boundary) based on the particular patterns in the measurements of the accelerometer 6.
  • Estimating accelerometer orientation—The orientation of the accelerometer 6 (and therefore the fall prevention device 2) can be estimated from the direction that gravity appears in the accelerometer coordinate system. By defining the z-axis to correspond with the vertical direction when the accelerometer 6 is not tilted, the orientation follows through the vector dot product of the measured gravity direction and the z-axis, i.e. through the z-component of the measured gravity (as is known, when using normalized values, the dot product equals the cosine of the enclosed angle).
  • Since the accelerometer 6 is sensitive to both acceleration due to gravity and acceleration due to movement, a filter is needed in order to estimate which component is due to gravity. Usually, a gyroscope 10 is added to measure the angular rotation speed and to correct the measured acceleration correspondingly. However, if the movements of the user 4 do not cause fast rotations of the fall prevention device 2, the gyroscope 10 can be omitted.
  • Without fast rotations, the gravity component can be found as the low-pass filtered component of the signal from the accelerometer 6. Since causality constraints in the filter design introduce a delay, the filtered acceleration signal needs to be corrected for this delay.
  • From the measured direction of gravity, the inclination or orientation of the fall prevention device 2 can be estimated. However, this inclination or orientation does not provide information about the horizontal direction of the device 2 (i.e. which way the device 2 is facing), so the magnetometer 8 can be used to determine the horizontal orientation of the device 2. Gyroscopes can also be used in the estimation of horizontal orientation.
  • Orientation can be expressed in different ways, of which Euler angles and Euler parameters are the most commonly used. They are usually implemented through matrices or quaternions. The algebras are isomorphic, and provide a way to transform the measured values (acceleration etc.) as expressed in the accelerometer's local coordinate system to the global (Earth) coordinate system.
  • Estimating step or stride sizes—Once the acceleration signals have been transformed to the Earth coordinate system, the step sizes can be computed by double integrating (with respect to time) the horizontal components of these transformed signals. The boundaries for the integrations are given by the estimated step boundaries. The integration constants, velocity and position at the beginning of the step, are both set to zero.
  • For the position this is fine, since the step size is needed, which is the difference between end and start position.
  • For the velocity this is correct in the case that the device 2 is located on a foot of the user 4, since upon heel strike the velocity relative to the Earth is zero. However, when the device 2 is located on the upper body of the user 4, for example when the device 2 is a pendant, there will be a near constant velocity. Therefore, setting the velocity to zero at each step boundary will introduce errors, since a constant velocity yields a larger step size estimate. However, if it is assumed that the velocity is constant during the period of observation, i.e. during the averaging period (typically around 10-12 steps for determining a, and around 30-60 seconds for determining g), the constant velocity component can be omitted. This means that the step size can be computed based on the variation in the velocity, which means that the device 2 (and the estimated fall risk) will become independent of the current walking speed—only the deviations will be observed.
  • The step size p follows as the integral of velocity v over step time:

  • p=∫ HS 0 HS 1 vdt= v (HS 1 −HS 0)+∫HS 0 HS 1 Δvdt  (3)
  • where HS0 and HS1 are the times of subsequent heel strikes (HS) and v and Δv denote the average velocity and its deviation respectively.
  • Defining the step time T=HS1−HS0= T+ΔT, and p= p+Δp, where p= v T, the deviation in step size follows as:

  • Δp= vΔT+∫ HS 0 HS 1 Δvdt  (4)
  • It should be noted that, by definition, the mean of (Δp)2 equals the variance of p. The velocity deviation Δv can be computed from the measured acceleration through:

  • Δv(t)=Δv(HS 0)+∫HS 0 t adt′  (5)
  • At constant walking velocity,

  • HS 0 HS 1 adt≈0  (6)
  • and it is reasonable to assume that Δv(HS0) is roughly the same at each step. Then, by neglecting the first term in equation (4) above, Δp is approximated as:

  • Δp≈∫ HS 0 HS 1 Δv(HS 0)dt+∫ HS 0 HS 1 dt∫ HS 0 t adt′=Δv(HS 0)(HS 1 −HS 0)+∫HS 0 HS 1 dt∫ HS 0 t adt′=μ+Δs  (7)
  • The second term, Δs, is the result of the double integration of the acceleration when using zero integration constants. Since, by definition, E[Δp]=0, it follows that

  • μ=−E[Δs]  (8)
  • So, a first order approximation of g can be obtained by observing the (horizontal) acceleration under stable walking conditions and computing the mean of the double integrated acceleration. This process also yields a standard deviation:

  • μ0=mean(Δs stable)  (9)

  • σ0=std(Δs stable)  (10)
  • So, μ0 estimates −μ, and the variance of p follows as

  • var(p)=E└(Δp)2┘=μ00 20 2  (11)
  • In typical cases, σ0 2<<μ0 2 and var(p)≈2μ2. During operation, a running estimate a=E[Δs] is made over a few steps, and is compared with μ0 relative to σ0, see equation (1). So, although mean values are compared, they basically reflect variance in the step sizes.
  • The values μ0 and σ0 are user dependent and need to be set for each user in a calibration phase. Note, however, that since the average velocity v is not accounted for in the calculations, the method is insensitive to the actual walking speed of the user 4.
  • The first term vΔT in the expression for Δp above has been neglected and introduces an error. It is proportional to the average velocity and the deviation in step time. As a refinement, it could be included in the estimate of Δp.
  • Since the orientation estimate contains errors, there is always some form of gravity leakage into the horizontal accelerations (i.e. the horizontal accelerations will include some component due to gravity) and double integration may induce large errors in the estimated step size. One way to suppress this leakage is by band-pass filtering (or high-pass filtering) the signals from the accelerometer 6, preferably using a linear phase filter so that the wave form of the signal is maintained. Typical cut-off frequencies are 0.1 Hz and 20-40 Hz (where the actual application of the upper cut-off frequency also depends on the used sampling frequency). This band-pass filtering is not required.
  • A measure that has been found to be particularly effective is the de-trending per integration interval (step or stride) of the acceleration. This amounts to requiring that the average acceleration (from the start heel strike to the end heel strike) is zero. In other words, the velocity after integration equals the velocity at the start of integration (which is zero). The de-trended acceleration is obtained by subtracting the integrated acceleration, divided by the step duration, from the measured acceleration. The de-trended acceleration is used to compute the step size as described above.
  • Assuming no errors in the coordinate system transformation, for example due to inhomogeneity in the direction and size of the geomagnetic field, this integration yields the direction of a step in terms of north-south and east-west. However, for gait stability it is necessary to examine the motion in forward and lateral directions. This is solved in the following way.
  • The start position (which is the origin, by definition) and the end position are computed as explained above. Their difference spans a 2D vector in the horizontal plane. Then, the forward step size is defined as the norm of this 2D vector, i.e. as the distance of the end position from the start position (it should be noted that, as a consequence, the above Δp are always positive). Subsequently, a straight line is defined between the start position and the end position and the distance is determined between this line and each point from the double integration of the accelerometer measurements during the step. The maximum distance in this series is taken as the lateral step size.
  • Step time follows as the duration between the estimated step boundaries, and velocity as the ratio between step size and step time (in the event that the device 2 is located on a foot of the user 4).
  • While stride sizes are computed when the device 2 is on the lower body of the user 4, step size can be computed when the device 2 is on the upper body of the user 4, yielding twice as fast an average.
  • It will be appreciated by those skilled in the art that alternative methods for determining step size can be used that do not require the double integration of the horizontal acceleration measurements. These methods include the inverted pendulum model [Zijlstra & H of, Displacement of the pelvis during human walking, Gait and Posture 6, 1997, 249-267] and the 4th power root of the difference between maximum and minimum vertical acceleration [Weinberg, Using the ADXL202 in Pedometer and Personal Navigation Applications, Application Note AN-602, Analog Devices, 2002].
  • Although the invention has been described in terms of a pendant to be worn around the neck of a user 4, it will be appreciated that the invention can be implemented in alternative forms that are to be worn on different parts of the body of the user 4. Of course, in these embodiments, it will be necessary to modify the processing used to determine the gait parameters from the sensor measurements, but these modifications will be readily apparent to a person skilled in the art based on the description provided above.
  • In the embodiment described with reference to FIG. 2, the collection and processing of the sensor measurements is performed in a single unit. However, in alternative embodiments, the processing of the measurements can be performed in a unit that is remote from the sensors, in which case the fall prevention device 2 will comprise a sensor unit to be worn by the user 4 that transmits the sensor measurements to the remote unit. In this embodiment, there is no need for the sensor unit to include a dedicated processor.
  • There is therefore provided a method and device that can determine an instantaneous fall risk for a user.
  • It will be appreciated that the algorithms described above that are used to determine various gait parameters can be used in applications other than fall prevention. For example, they can be used in activity monitoring and fitness applications such as endurance coaching (for example to support the keeping of cadence during jogging). The algorithms may be used in step counters or devices that support injury prevention, for example during jogging. Further the algorithms may be used in a device or system that uses gait parameters as a biometric for identifying an individual.
  • While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments.
  • Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. A computer program may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.

Claims (15)

1. A method of determining a fall risk of a user, the method comprising:
collecting measurements of the motion of the user;
estimating a value for a parameter related to the gait of the user from the measurements; and
determining a fall risk for the user from a comparison of the estimated value with a normal value for the parameter determined from motion of the user in which the user is at their normal risk of falling.
2. A method as claimed in claim 1, wherein the step of determining a fall risk comprises weighting the comparison between the estimated value and the normal value according to a standard deviation of the normal value.
3. A method as claimed in claim 1, wherein the estimated value is determined from motion of the user over a period of time that is shorter than the period of time over which the normal value is determined.
4. A method as claimed in claim 1, wherein the step of estimating comprises identifying a step boundary in the collected measurements.
5. A method as claimed in claim 4, wherein the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold.
6. A method as claimed in claim 4, wherein the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements in which the magnitude of each of the measurements exceeds a threshold, apart from a subset of the measurements whose magnitude is less than the threshold, provided that the subset covers a time period less than a time threshold.
7. A method as claimed in claim 4, wherein the step of identifying a step boundary comprises identifying clusters of contiguous measurements in the collected measurements, wherein the first collected measurement in the collected measurements whose magnitude exceeds a first threshold denotes the first measurement in a cluster and wherein the first collected measurement after the first measurement in the cluster whose magnitude falls below a second threshold denotes the last measurement in the cluster, provided that the last measurement is more than a minimum period after the first measurement.
8. A method as claimed in claim 5, wherein the step of identifying step boundaries further comprises identifying the step boundary as the measurement in each cluster with the highest magnitude.
9. A method as claimed in claim 4, wherein the parameter related to the gait of the user comprises a step size and the step of estimating a value for the parameter comprises integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries.
10. A method as claimed in claim 9, wherein the step of estimating a value for the parameter comprises computing a double integration with respect to time of the horizontal components of the collected measurements relating to acceleration, the integration constants being set to zero at the beginning of the step.
11. A method as claimed in claim 4, wherein the parameter related to the gait of the user comprises, or additionally comprises, a forward step size and the step of estimating a value for the parameter comprises:
integrating horizontal components of the collected measurements with the integral bounds being given by consecutive identified step boundaries to give a start and end position for a step; and
determining the forward step size as the norm of the vector connecting the start and end positions.
12. A method as claimed in claim 10, wherein the parameter related to the gait of the user additionally comprises a lateral step size and the step of estimating a value for the parameter further comprises:
defining a straight line between the start and end positions;
integrating collected measurements occurring during the step to give a series of positions during the step;
determining the distance between each position and the straight line; and
determining the lateral step size as the maximum distance in this series.
13. A method as claimed in claim 1, further comprising a calibration step that includes:
collecting measurements of the motion of the user when the user is at their normal risk of falling; and
estimating the normal value for the parameter related to the gait of the user from the collected measurements.
14. A fall prevention device, comprising:
at least one sensor for collecting measurements of the motion of a user of the device; and
a processor for estimating a value for a parameter related to the gait of the user from the measurements, and for determining a fall risk for the user from a comparison of the estimated value with a value of the parameter determined from motion of the user in which the user is at their normal risk of falling.
15. A computer program product comprising computer-readable code that, when executed on a suitable computer or processor, is configured to cause the computer or processor to perform the steps in the method defined in claim 1.
US13/377,419 2009-07-10 2010-07-06 Fall prevention Abandoned US20120095722A1 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
EP09165127.3 2009-07-10
EP09165127 2009-07-10
PCT/IB2010/053090 WO2011004322A1 (en) 2009-07-10 2010-07-06 Fall prevention

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2010/053090 A-371-Of-International WO2011004322A1 (en) 2009-07-10 2010-07-06 Fall prevention

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US14/693,953 Continuation US10670621B2 (en) 2009-07-10 2015-04-23 Fall prevention

Publications (1)

Publication Number Publication Date
US20120095722A1 true US20120095722A1 (en) 2012-04-19

Family

ID=42830315

Family Applications (2)

Application Number Title Priority Date Filing Date
US13/377,419 Abandoned US20120095722A1 (en) 2009-07-10 2010-07-06 Fall prevention
US14/693,953 Active 2031-03-21 US10670621B2 (en) 2009-07-10 2015-04-23 Fall prevention

Family Applications After (1)

Application Number Title Priority Date Filing Date
US14/693,953 Active 2031-03-21 US10670621B2 (en) 2009-07-10 2015-04-23 Fall prevention

Country Status (9)

Country Link
US (2) US20120095722A1 (en)
EP (1) EP2451351B1 (en)
JP (1) JP5647240B2 (en)
CN (1) CN102469955B (en)
AU (1) AU2010269846B2 (en)
BR (1) BR112012000267A8 (en)
ES (1) ES2716976T3 (en)
RU (1) RU2550934C2 (en)
WO (1) WO2011004322A1 (en)

Cited By (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110152727A1 (en) * 2008-09-04 2011-06-23 Koninklijke Philips Electronics N.V. Fall prevention system
US20120032806A1 (en) * 2010-08-06 2012-02-09 Samsung Electronics Co., Ltd. Detecting apparatus and method, and mobile terminal apparatus having detecting apparatus
US20120083237A1 (en) * 2010-10-04 2012-04-05 Ram David Adva Fish Fall detection system using a combination of accelerometer, audio input and magnetometer
US20120154145A1 (en) * 2010-12-15 2012-06-21 Dell Products L.P. Mobile and automated emergency service provider contact system
US20130120152A1 (en) * 2011-11-14 2013-05-16 Vigilo Networks, Inc. Method and system for fall detection of a user
US20140358041A1 (en) * 2012-04-18 2014-12-04 Matthew Alan Hopcroft Assessing physical stability of a patient using an accelerometer
US9153114B2 (en) * 2014-02-07 2015-10-06 Ge Yi Fall detection method and system
WO2016049688A1 (en) * 2014-10-03 2016-04-07 Neuroscience Research Australia (Neura) Systems and methods to monitor the step performance of a user
US20170119283A1 (en) * 2015-10-28 2017-05-04 Koninklijke Philips N.V. Monitoring activities of daily living of a person
US20170249821A1 (en) * 2010-10-19 2017-08-31 Orthocare Innovations Llc Fall risk assessment device and method
US10024660B2 (en) 2012-08-27 2018-07-17 Universite Du Quebec A Chicoutimi Method to determine physical properties of the ground
US10231651B2 (en) * 2014-09-25 2019-03-19 Bae Systems Information And Electronic Systems Integration Inc. Gait authentication system and method thereof
US10307084B2 (en) 2015-06-30 2019-06-04 Zibrio Inc. Identifying fall risk using machine learning algorithms
US10335059B2 (en) 2013-09-11 2019-07-02 Koninklijke Philips N.V. Fall detection system and method
US10692346B2 (en) 2018-10-22 2020-06-23 Tidi Products, Llc Electronic fall monitoring system
US10888503B1 (en) * 2019-05-17 2021-01-12 Inga Londeree Pacifier with onboard drop sensor and notifications
US11039761B2 (en) * 2018-12-14 2021-06-22 At&T Intellectual Property I, L.P. Fall prediction based on electroencephalography and gait analysis data
US11183304B2 (en) 2019-01-08 2021-11-23 International Business Machines Corporation Personalized smart home recommendations through cognitive load analysis
US11210922B2 (en) 2018-10-22 2021-12-28 Tidi Products, Llc Electronic fall monitoring system
CN114271578A (en) * 2021-12-24 2022-04-05 杭州电子科技大学 Intelligent shoe capable of detecting foot step disorder and detection method
US11504071B2 (en) 2018-04-10 2022-11-22 Hill-Rom Services, Inc. Patient risk assessment based on data from multiple sources in a healthcare facility
US11908581B2 (en) 2018-04-10 2024-02-20 Hill-Rom Services, Inc. Patient risk assessment based on data from multiple sources in a healthcare facility

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012205816A (en) * 2011-03-30 2012-10-25 Omron Healthcare Co Ltd Walking posture determination device
EP2549228A1 (en) * 2011-07-20 2013-01-23 Koninklijke Philips Electronics N.V. A method of enhancing the detectability of a height change with an air pressure sensor and a sensor unit for determining a height change
JP5884623B2 (en) * 2012-04-27 2016-03-15 富士通株式会社 Fall prediction program, portable electronic device, and fall prediction method
US9383202B2 (en) * 2013-03-12 2016-07-05 Google Inc. Barometric pressure sensor based orientation measurement
JP6253076B2 (en) * 2013-03-27 2017-12-27 長崎県公立大学法人 Fall prevention device
JP6448626B2 (en) 2013-06-06 2019-01-09 コーニンクレッカ フィリップス エヌ ヴェKoninklijke Philips N.V. Fall detection system and method
EP3099229B1 (en) 2014-01-30 2019-10-30 Koninklijke Philips N.V. Improvements in the detection of walking in measurements of the movement of a user
CN105091903B (en) * 2015-06-30 2018-04-13 小米科技有限责任公司 Ambulatory status monitoring method and device
CN105286874B (en) * 2015-11-03 2018-06-19 长安大学 A kind of system and method for being used to predict Falls in Old People risk
CN106389074A (en) * 2016-01-27 2017-02-15 北京航空航天大学 Falling process stability predicting device and method based on plantar pressure sensing
CA3014337A1 (en) * 2016-02-16 2017-08-24 Koninklijke Philips N.V. Lanyard device, method and personal lanyard monitoring system
WO2018069262A1 (en) 2016-10-12 2018-04-19 Koninklijke Philips N.V. Method and apparatus for determining a fall risk
CN106388831B (en) * 2016-11-04 2020-01-10 郑州航空工业管理学院 Method for detecting tumbling action based on sample weighting algorithm
US10438136B2 (en) 2017-06-09 2019-10-08 Midea Group Co., Ltd. System and method for care support at home
US11213224B2 (en) * 2018-03-19 2022-01-04 Electronic Caregiver, Inc. Consumer application for mobile assessment of functional capacity and falls risk
EP3627515A1 (en) * 2018-09-19 2020-03-25 Koninklijke Philips N.V. Monitoring a user
EP3711666A1 (en) 2019-03-21 2020-09-23 Koninklijke Philips N.V. Analysing movement of a subject
JP7115423B2 (en) * 2019-06-07 2022-08-09 トヨタ自動車株式会社 Gait training system, display method, and display program
SE1951107A1 (en) 2019-09-30 2021-03-31 Pink Nectarine Health Ab System and method for monitoring an individual
CN110916984B (en) * 2019-12-03 2021-11-26 上海交通大学医学院附属第九人民医院 Wearable device for preventing frozen gait and implementation method thereof
CN112382052A (en) * 2020-11-16 2021-02-19 南通市第一人民医院 Patient falling alarm method and system based on Internet

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5836887A (en) * 1996-09-19 1998-11-17 Colin Corporation Physical information monitor system having means for determining reference range for abnormality determination, based on moving average of previously obtained values
US5919149A (en) * 1996-03-19 1999-07-06 Allum; John H. Method and apparatus for angular position and velocity based determination of body sway for the diagnosis and rehabilitation of balance and gait disorders
US6305221B1 (en) * 1995-12-12 2001-10-23 Aeceleron Technologies, Llc Rotational sensor system
US20050027216A1 (en) * 2003-07-02 2005-02-03 Commissariat A L'energie Atomique Portable detector for measuring movements of a person wearing the detector, and method
US20060195050A1 (en) * 2003-04-03 2006-08-31 University Of Virginia Patent Foundation Method and system for the derivation of human gait characteristics and detecting falls passively from floor vibrations
US20080108913A1 (en) * 2006-11-06 2008-05-08 Colorado Seminary, Which Owns And Operates The University Of Denver Smart apparatus for gait monitoring and fall prevention
US20080146968A1 (en) * 2006-12-14 2008-06-19 Masuo Hanawaka Gait analysis system
US20090240170A1 (en) * 2008-03-20 2009-09-24 Wright State University Systems and methods for determining pre-fall conditions based on the angular orientation of a patient
US7610166B1 (en) * 2006-07-21 2009-10-27 James Solinsky Geolocation system and method for determining mammal locomotion movement
US20100286571A1 (en) * 2006-11-17 2010-11-11 Allum John H J System and Method for Providing Body Sway Feedback to a Body of a Subject
US8206325B1 (en) * 2007-10-12 2012-06-26 Biosensics, L.L.C. Ambulatory system for measuring and monitoring physical activity and risk of falling and for automatic fall detection

Family Cites Families (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SU1258380A1 (en) * 1985-03-19 1986-09-23 Уральский политехнический институт им.С.М.Кирова Device for estimating vestibular sensor system
RU2008056C1 (en) * 1990-03-11 1994-02-28 Государственный институт физической культуры им.П.Ф.Лесгафта Device for measuring statokinetic stimuli
US5485402A (en) * 1994-03-21 1996-01-16 Prosthetics Research Study Gait activity monitor
CN1256752A (en) * 1998-02-25 2000-06-14 皇家菲利浦电子有限公司 Method of and system for measuring performance during exercise activity, and athletic shoe for use in the system
US6626557B1 (en) * 1999-12-29 2003-09-30 Spx Corporation Multi-colored industrial signal device
WO2001088836A1 (en) * 2000-05-18 2001-11-22 Commwell, Inc. Method and apparatus for remote medical monitoring incorporating video processing and system of motor tasks
US6594617B2 (en) * 2000-08-18 2003-07-15 Applanix Corporation Pedometer navigator system
EP1195139A1 (en) * 2000-10-05 2002-04-10 Ecole Polytechnique Féderale de Lausanne (EPFL) Body movement monitoring system and method
US20070038155A1 (en) 2001-01-05 2007-02-15 Kelly Paul B Jr Attitude Indicator And Activity Monitoring Device
ATE427699T1 (en) 2002-02-07 2009-04-15 Ecole Polytech BODY MOVEMENT MONITORING DEVICE
JP4592360B2 (en) 2004-09-02 2010-12-01 公立大学法人会津大学 Physical condition monitoring device
DE102005004086A1 (en) 2005-01-21 2006-07-27 Xybermind Gmbh Device and method for detecting movement
US20070112287A1 (en) * 2005-09-13 2007-05-17 Fancourt Craig L System and method for detecting deviations in nominal gait patterns
WO2007033194A2 (en) 2005-09-13 2007-03-22 Aware Technologies, Inc. Method and system for proactive telemonitor with real-time activity and physiology classification and diary feature
JP4962765B2 (en) * 2005-09-26 2012-06-27 アイシン精機株式会社 Walking analysis device and walking analysis method
US20070073514A1 (en) 2005-09-26 2007-03-29 Aisin Seiki Kabushiki Kaisha Walking analyzer
DE602005014641D1 (en) * 2005-10-03 2009-07-09 St Microelectronics Srl Pedometer device and step detection method by algorithm for self-adaptive calculation of acceleration limits
JP2010508945A (en) * 2006-11-14 2010-03-25 コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ System for preventing fall and method for preventing fall using this system
US7612681B2 (en) * 2007-02-06 2009-11-03 General Electric Company System and method for predicting fall risk for a resident
US7647196B2 (en) * 2007-08-08 2010-01-12 Dp Technologies, Inc. Human activity monitoring device with distance calculation
US9179864B2 (en) 2007-08-15 2015-11-10 Integrity Tracking, Llc Wearable health monitoring device and methods for fall detection
US8120498B2 (en) * 2007-09-24 2012-02-21 Intel-Ge Care Innovations Llc Capturing body movement related to a fixed coordinate system
FR2924847B1 (en) 2007-12-06 2014-08-29 Vigilio METHOD AND EQUIPMENT FOR DETECTING CRITICAL SITUATION OF A SUBJECT
EP2335231B1 (en) 2008-09-04 2012-12-26 Koninklijke Philips Electronics N.V. Fall prevention system

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6305221B1 (en) * 1995-12-12 2001-10-23 Aeceleron Technologies, Llc Rotational sensor system
US5919149A (en) * 1996-03-19 1999-07-06 Allum; John H. Method and apparatus for angular position and velocity based determination of body sway for the diagnosis and rehabilitation of balance and gait disorders
US5836887A (en) * 1996-09-19 1998-11-17 Colin Corporation Physical information monitor system having means for determining reference range for abnormality determination, based on moving average of previously obtained values
US20060195050A1 (en) * 2003-04-03 2006-08-31 University Of Virginia Patent Foundation Method and system for the derivation of human gait characteristics and detecting falls passively from floor vibrations
US20050027216A1 (en) * 2003-07-02 2005-02-03 Commissariat A L'energie Atomique Portable detector for measuring movements of a person wearing the detector, and method
US7610166B1 (en) * 2006-07-21 2009-10-27 James Solinsky Geolocation system and method for determining mammal locomotion movement
US20080108913A1 (en) * 2006-11-06 2008-05-08 Colorado Seminary, Which Owns And Operates The University Of Denver Smart apparatus for gait monitoring and fall prevention
US20100286571A1 (en) * 2006-11-17 2010-11-11 Allum John H J System and Method for Providing Body Sway Feedback to a Body of a Subject
US20080146968A1 (en) * 2006-12-14 2008-06-19 Masuo Hanawaka Gait analysis system
US8206325B1 (en) * 2007-10-12 2012-06-26 Biosensics, L.L.C. Ambulatory system for measuring and monitoring physical activity and risk of falling and for automatic fall detection
US20090240170A1 (en) * 2008-03-20 2009-09-24 Wright State University Systems and methods for determining pre-fall conditions based on the angular orientation of a patient

Cited By (34)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110152727A1 (en) * 2008-09-04 2011-06-23 Koninklijke Philips Electronics N.V. Fall prevention system
US9392966B2 (en) * 2008-09-04 2016-07-19 Koninklijke Philips N.V. Fall prevention system
US20120032806A1 (en) * 2010-08-06 2012-02-09 Samsung Electronics Co., Ltd. Detecting apparatus and method, and mobile terminal apparatus having detecting apparatus
US9761118B2 (en) * 2010-08-06 2017-09-12 Samsung Electronics Co., Ltd. Detecting apparatus and method, and mobile terminal apparatus having detecting apparatus
US10309980B2 (en) 2010-10-04 2019-06-04 Nortek Security & Control Llc Fall detection system using a combination of accelerometer, audio input and magnetometer
US8843101B2 (en) * 2010-10-04 2014-09-23 Numera, Inc. Fall detection system using a combination of accelerometer, audio input and magnetometer
US9462444B1 (en) 2010-10-04 2016-10-04 Nortek Security & Control Llc Cloud based collaborative mobile emergency call initiation and handling distribution system
US9648478B2 (en) 2010-10-04 2017-05-09 Nortek Security & Control Llc Fall detection system using a combination of accelerometer, audio input and magnetometer
US20120083237A1 (en) * 2010-10-04 2012-04-05 Ram David Adva Fish Fall detection system using a combination of accelerometer, audio input and magnetometer
US20170249821A1 (en) * 2010-10-19 2017-08-31 Orthocare Innovations Llc Fall risk assessment device and method
US8633818B2 (en) * 2010-12-15 2014-01-21 Dell Products L.P. Mobile and automated emergency service provider contact system
US20120154145A1 (en) * 2010-12-15 2012-06-21 Dell Products L.P. Mobile and automated emergency service provider contact system
US20130120152A1 (en) * 2011-11-14 2013-05-16 Vigilo Networks, Inc. Method and system for fall detection of a user
US9818281B2 (en) * 2011-11-14 2017-11-14 Vital Connect, Inc. Method and system for fall detection of a user
US20140358041A1 (en) * 2012-04-18 2014-12-04 Matthew Alan Hopcroft Assessing physical stability of a patient using an accelerometer
US10024660B2 (en) 2012-08-27 2018-07-17 Universite Du Quebec A Chicoutimi Method to determine physical properties of the ground
US10335059B2 (en) 2013-09-11 2019-07-02 Koninklijke Philips N.V. Fall detection system and method
US9153114B2 (en) * 2014-02-07 2015-10-06 Ge Yi Fall detection method and system
US10231651B2 (en) * 2014-09-25 2019-03-19 Bae Systems Information And Electronic Systems Integration Inc. Gait authentication system and method thereof
WO2016049688A1 (en) * 2014-10-03 2016-04-07 Neuroscience Research Australia (Neura) Systems and methods to monitor the step performance of a user
US10307084B2 (en) 2015-06-30 2019-06-04 Zibrio Inc. Identifying fall risk using machine learning algorithms
US10542914B2 (en) 2015-06-30 2020-01-28 Zibrio Inc. Identifying fall risk using machine learning algorithms
US10863927B2 (en) 2015-06-30 2020-12-15 Zibrio Inc. Identifying fall risk using machine learning algorithms
US20170119283A1 (en) * 2015-10-28 2017-05-04 Koninklijke Philips N.V. Monitoring activities of daily living of a person
US11504071B2 (en) 2018-04-10 2022-11-22 Hill-Rom Services, Inc. Patient risk assessment based on data from multiple sources in a healthcare facility
US11908581B2 (en) 2018-04-10 2024-02-20 Hill-Rom Services, Inc. Patient risk assessment based on data from multiple sources in a healthcare facility
US10692346B2 (en) 2018-10-22 2020-06-23 Tidi Products, Llc Electronic fall monitoring system
US11776374B2 (en) * 2018-10-22 2023-10-03 Tidi Products, Llc Electronic fall monitoring system
US11210922B2 (en) 2018-10-22 2021-12-28 Tidi Products, Llc Electronic fall monitoring system
US20220101712A1 (en) * 2018-10-22 2022-03-31 Tidi Products, Llc Electronic Fall Monitoring System
US11039761B2 (en) * 2018-12-14 2021-06-22 At&T Intellectual Property I, L.P. Fall prediction based on electroencephalography and gait analysis data
US11183304B2 (en) 2019-01-08 2021-11-23 International Business Machines Corporation Personalized smart home recommendations through cognitive load analysis
US10888503B1 (en) * 2019-05-17 2021-01-12 Inga Londeree Pacifier with onboard drop sensor and notifications
CN114271578A (en) * 2021-12-24 2022-04-05 杭州电子科技大学 Intelligent shoe capable of detecting foot step disorder and detection method

Also Published As

Publication number Publication date
ES2716976T3 (en) 2019-06-18
JP5647240B2 (en) 2014-12-24
JP2012532652A (en) 2012-12-20
EP2451351B1 (en) 2019-01-16
RU2012104679A (en) 2013-08-20
US10670621B2 (en) 2020-06-02
EP2451351A1 (en) 2012-05-16
BR112012000267A8 (en) 2017-12-05
AU2010269846A1 (en) 2012-03-01
RU2550934C2 (en) 2015-05-20
CN102469955B (en) 2015-07-08
WO2011004322A1 (en) 2011-01-13
US20150226764A1 (en) 2015-08-13
CN102469955A (en) 2012-05-23
AU2010269846B2 (en) 2014-10-02

Similar Documents

Publication Publication Date Title
US10670621B2 (en) Fall prevention
US9392966B2 (en) Fall prevention system
JP5695778B2 (en) Fall detection system
US20200046263A1 (en) System and method for characterizing biomechanical activity
US9456771B2 (en) Method for estimating velocities and/or displacements from accelerometer measurement samples
US9427177B2 (en) Fall detection methods and devices
KR100601981B1 (en) Method and apparatus for monitoring human activity pattern
US11497966B2 (en) Automatic coaching system and method for coaching user&#39;s exercise
JP2009525107A (en) Gait analysis
US9159213B1 (en) Motion detection method and device
WO2017088817A1 (en) Crutch walker
JP4214259B2 (en) Human operating state monitoring method and apparatus
Kita et al. Reliable and robust detection of freezing of gait episodes with wearable electronic devices
CN106327799A (en) Trip monitoring method and system in walking state
CN109512435B (en) Walking posture correction instrument capable of detecting in real time
TW201918225A (en) Method and system for analyzing gait
JPWO2016063661A1 (en) Information processing apparatus, information processing method, and program
US20180125394A1 (en) Apparatus and method for determining a sedentary state of a subject

Legal Events

Date Code Title Description
AS Assignment

Owner name: KONINKLIJKE PHILIPS ELECTRONICS N.V., NETHERLANDS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:TEN KATE, WARNER RUDOLPH THEOPHILE;REEL/FRAME:027356/0470

Effective date: 20100709

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION