Classifications

• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
  • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
  • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
  • A61B5/1116—Determining posture transitions
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
  • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
  • A61B5/1118—Determining activity level
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
  • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
  • A61B5/1123—Discriminating type of movement, e.g. walking or running
• • A—HUMAN NECESSITIES
  • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
  • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
  • A61B5/00—Measuring for diagnostic purposes; Identification of persons
  • A61B5/40—Detecting, measuring or recording for evaluating the nervous system
  • A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
  • A61B5/4094—Diagnosing or monitoring seizure diseases, e.g. epilepsy
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
  • G06V40/20—Movements or behaviour, e.g. gesture recognition
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06F—ELECTRIC DIGITAL DATA PROCESSING
  • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing

Definitions

• the present invention relates to a system and method for determining the posture of a person.
• An object of the invention is to improve the accuracy of the determination of the activity of a mobile element, particularly for a living being, human or animal.
• a system for determining the posture of a person comprising at least one motion sensor with at least one measuring axis, provided with fixing means for integrally connecting said motion sensor. to a user.
• the system comprises: a filter for selecting, for each measurement axis of the motion sensor, high frequencies greater than a first threshold, and low frequencies below a second threshold less than or equal to said first threshold;
• calculation means a probability density of said high frequency component and a probability density of said component low frequency, said high frequency component probability density being defined by a Chi-2 law of degree of freedom equal to the number of measurement axes taken into account by the motion sensor, and said probability density of the low frequency component being defined by a Gaussian law;
• a hidden Markov model is defined by two random processes: a first one which is called a "state" in the present application and which is not observed, or, in other words, which is hidden, and a second which is the observation whose probability density at a given moment depends on the value of the state at the same instant.
• the state takes discrete values.
• Such a system makes it possible to determine the activity of a mobile element, particularly for a living being, human or animal, with improved accuracy.
• said determination means are adapted to determine a one-dimensional low frequency component equal to a linear combination of measurements along the measurement axes taken into account by the motion sensor, said high frequency component being defined by a Chi-law. -2 to one degree of freedom.
• the probability density of a pair of values for the low frequency component and the high frequency component comprises the product of a probability density of obtaining the value for the low frequency component and the density. probability of obtaining the value for the high frequency component, said probability densities being defined, for each state i, by the following expressions: [x (n) - ⁇ x J 2
• x (n) is a signal of dimension 1, representing the low frequency component at the sample of index n;
• ⁇ x j represents a vector of the same dimension as the low frequency component, representative of the state i of the hidden Markov model considered;
• ⁇ X t ⁇ represents the square root of the vahance of the low frequency component x, representative of the state of the hidden Markov model i considered;
• y (n) represents the high frequency component at the sample of index n;
• k represents the number of measurement axes taken into account by the motion sensor;
• ⁇ y j is a quantity proportional to the time average of the variable y (n), in state i.
• ⁇ yu is the time average of the variable y (n) divided by k
• the system includes display means.
• said motion sensor comprises an accelerometer, and / or a magnetometer, and / or a gyrometer.
• the system comprises a first accelerometer with a measurement axis and fastening means adapted to fix the first accelerometer at the torso of the user so that the measurement axis coincides with the vertical axis VT of the body when the user is upright.
• said analysis means are adapted to determine a user's posture as a function of time using a Markov model hidden at most four states among the standing or sitting posture, the walking posture, the posture leaning, and lying posture.
• the hidden Markov model is then defined by:
• This variable, or state is a Markov sequence of order 1, and is therefore characterized by the probabilities of passing from one state to another;
• the observed process of the hidden Markov model is the multidimensional signal W " ⁇ ', whose probability density depends on the state (the hidden process) at a given time.
• Pi (x (n), y (n)) represents the probability density associated with state i, at time n, of x (n) and y (n). It corresponds to the product of the previously defined probability densities P x , (! («)) And P y ⁇ ⁇ y ⁇ n)).
• the estimated sequence of states E (0: N) is the one with the highest probability.
• P (E (O)) denotes the probability associated with the initial state E (O).
• One can, for example, choose an equiprobable distribution of each of the possible states when n 0.
• the system further comprises a second accelerometer with a measurement axis and fixing means adapted to fix the second accelerometer at the level of the thigh of the user so that the measurement axis coincides with the vertical axis VT of the body when the user is upright.
• said analysis means is adapted to determine a user's posture as a function of time using a Markov model hidden at most four states among standing posture, seated posture, elongated posture, and posture. walk.
• x ⁇ n) represents the pair of respective low frequency components of said two accelerometers
• y ⁇ n) represents the high frequency component of said second accelerometer, to the sample of index n, the probability density of obtaining the value x ⁇ n), corresponding to the state i, being defined by the following expression:
• ⁇ xi is a diagonal matrix of dimension 2 describing the covariance matrix of the signal x (n) for the state i of the model.
• ⁇ xi represents a two-component column vector, representative of the state i of the model.
• E (O, N) corresponds to the sequence of states E (O), E (I) ... E (N) maximizing the expression:
• ⁇ (n) ⁇ x (n), y (n) ⁇ , x (n) and y (n) are respectively low and high frequency components of the signal S (n) measured by two accelerometers at the instant n.
• a method for determining the posture of a person characterized in that: - one filters to select, for each axis of measurement of a motion sensor, high frequencies above a first threshold, and low frequencies below a second threshold less than or equal to said first threshold;
• a one-dimensional high frequency component equal to the sum of the squares of said high frequencies of the measurement axes taken into account by the motion sensor is determined, and a low frequency component of dimension equal to the number of measurement axes taken into account by the sensor of movement ;
• said high frequency component being defined by a a Chi-2 law of degree of freedom equal to the number of measurement axes taken into account by the motion sensor, and said low-frequency component being defined by a Gaussian law;
• a user's posture is determined as a function of time by using a hidden Markov model with N states corresponding respectively to N postures, this determination being made by combining: - joint probability densities of said low frequency and high frequency components , these densities of probabilities being defined for each posture, and - probabilities of passage between two successive postures.
• FIG. 1 illustrates a system, according to one aspect of the invention
• FIG. 2 illustrates an exemplary recording of a system according to one aspect of the invention.
• FIG. 3 illustrates an exemplary recording of a system according to another aspect of the invention.
• FIG. 1 illustrates a system for determining the posture of a person comprising at least one motion sensor CM with at least one measuring axis, arranged in a housing BT, provided with fixing means comprising for example an elastic element, for solidly bind the CM motion sensor to a user.
• the CM motion sensor can be, an accelerometer, a magnetometer, or a gyrometer, with one, two, or three measurement axes.
• the system comprises a FILT filter for selecting, for each measurement axis of the CM motion sensor, high frequencies higher than a first threshold S1, and low frequencies lower than a second threshold S2 less than or equal to the first threshold S1.
• the system also comprises a determination module DET of a high frequency component HF unidimensional equal to the sum of the squares of said high frequencies of the measurement axes taken into account of the motion sensor CM, and a low frequency component BF unidimensional equal to a linear combination of measurements according to the measurement axes taken into account of the CM motion sensor.
• the system also comprises a calculation module CALC of the square of the variance of the probability P y of said high frequency component HF and the square of the variance of the probability P x of said low frequency component BF, said high frequency component HF being defined by a Chi-2 law with a degree of freedom and said low-frequency component BF being defined by a Gaussian law.
• AN analysis means make it possible to determine a user's posture as a function of time by using a hidden Markov model with N states corresponding respectively to N postures.
• the joint probability probability probability Pj (x (n), y (n)) of obtaining a pair of values (x (n), y (n)) for the low frequency component BF and the high frequency component HF being equal to the product of the probability density P x j of obtaining the value x (n) for the low frequency component BF and the probability density P y j of obtaining the value y (n) for the high frequency component HF
• the probability densities P x j, P yJ are defined for each state i by the following expressions:
• x ⁇ n represents the low frequency component of the sample of index n
• ⁇ x ⁇ represents a vector of the same dimension as the low frequency component, representative of the state i of the hidden Markov model considered
• ⁇ x ⁇ represents the square root of the variance of the low frequency component x, representative of the state of the hidden Markov model considered
• y ⁇ n) represents the high frequency component at the sample of index n
• k represents the number of measurement axes taken into account by the motion sensor
• ⁇ y is a quantity proportional to the time average of the variable y (n), in state i.
• ⁇ yj is the time average of the variable y (n) divided by k
• the system also includes an AFF display screen.
• the system comprises an accelerometer with a measurement axis and a fixing element for fixing the accelerometer at the torso of the user so that the measurement axis coincides with the vertical axis VT of the body. when the user is upright.
• the hidden Markov model used includes four states corresponding to four postures, standing or sitting posture (state 1), walking posture (state 2), tilted posture (state 3), and lying posture (state 4).
• state 1 standing or sitting posture
• state 2 walking posture
• state 3 tilted posture
• state 4 lying posture
• the states of the hidden Markov model are defined as follows:
• E (N) is written p (E) (0: N) ⁇ ⁇ (0: N -I)) which is proportional to: p (E (0)) P (O (O) / E (O)) f [p (E (n) l E ( n - 1)) p ( ⁇ (n) / E (n))
• the estimated sequence of states E (0: N) is the one with the highest probability.
• P (E (O)) denotes the probability associated with the initial state E (O).
• E (O) the probability associated with the initial state E (O).
• the series of states E (O)... E (N) maximizing the expression (1) can be obtained by using, for example, the Viterbi algorithm, well known to those skilled in the art. So,
• ⁇ (n) ⁇ x (n), y (n) ⁇ , x (n) and y (n) are respectively low and high frequency components of the measured signal S (n). by an accelerometer at the moment n.
• the passing probability probabilities P (state / state j ) of a state state corresponding to a posture of the Markov model hidden at another state state j corresponding to a posture of the hidden Markov model are as follows, chosen so as to ensure good stability to the system:
• the analysis module AN determines, from the input signals and the hidden Markov model as defined, the sequence of states (postures) most likely, according to conventional methods, for example by calculating for the whole sequences of possible states the associated probability taking into account the observed signal and keeping the most probable sequence, as described for example in the document "An introduction to hidden Markov models” LR Rabiner and BH Juang, IEEE ASSP Magazine, January 1986, or in the book “Inference in Hidden Markov Models" by Capcher, Moulines and Ryden of Springer, from the series “Springer series in statisctics”.
• the various elements of the system may, for example, be integrated in the same LV box, as shown in Figure 1a, or some outsourced, for example in a laptop OP, as shown in Figure 1b.
• Figure 2 illustrates an example of a system user registration of the first example, on the lower graph, and the result provided by the system that indicates that the user was in the standing or sitting posture (state 1) while 36 seconds, then in the walking posture (state 2) for 16 seconds, then in the standing or sitting posture (state 1) for 8 seconds, then in the leaning posture (state 3) for 18 seconds, then in the upright posture or sitting (state 1) for 6 seconds, then in the walking posture (state 2) for 30 seconds, then in the tilted posture (state 3) for 38 seconds, then in the standing or sitting posture (state 1) for 8 seconds. seconds, then in the walking posture (state 2) for 51 seconds, and finally ends up in standing or sitting posture (state 1).
• the system comprises a first accelerometer with a measurement axis and a first attachment element for fixing the first accelerometer at the torso of the user so that the measurement axis coincides with the vertical axis VT. of the body when the user is standing upright, and a second accelerometer to a measurement axis and a second attachment element to fix the second accelerometer at the level of the thigh of the user so that the measurement axis coincides with the VT vertical axis of the body when the user is standing upright.
• the hidden Markov model used includes four states corresponding to four postures, standing posture (state 1), sitting posture (state 2), elongated posture (state 3), and posture walking (state 4).
• x ⁇ n represents the pair of the respective low frequency components BF of said two accelerometers
• y (n) represents the high frequency component HF of said second accelerometer, with the sample of index n, the probability density P x obtaining the value x (n) being defined by the following expression:
• ⁇ x ⁇ is a diagonal matrix of dimension 2 describing the covariance matrix of the signal x (n) for the state i of the model.
• ⁇ x ⁇ represents a two-component column vector, representative of the state i of the model.
• the probabilities of the variables x (n) and y (n) associated with these states are defined by the probabilities above, with the following parameters: - for the standing posture (state 1), the parameters of the probability densities are defined as follows:
• ⁇ x 2 [l Of and ⁇ .
• ⁇ y [l Of and ⁇ .
• ⁇ y , > , 2 3e "2
• ⁇ x 3 [ ⁇ Of and ⁇ 3
• the instant n N
• E (O, N) corresponds to the sequence of states E (O), E (I) ... E (N) maximizing the expression:
• ⁇ (n) ⁇ x (n), y (n) ⁇ , x (n) and y (n) are respectively low and high frequency components of the signal S (n) measured by two accelerometers at the instant n.
• Transition probability densities P (condition / state]) of a state; corresponding to a posture of the Markov model hidden to another state state j corresponding to a hidden Markov model posture are the following, chosen so as to ensure good stability to the system:
• the analysis module AN determines, from the input signals and the hidden Markov model as defined, the sequence of states (postures) most likely, according to conventional methods, for example by calculating for the whole sequences of possible states the associated probability taking into account the observed signal and keeping the most probable sequence, as described for example in the document "An introduction to hidden Markov models” LR Rabiner and BH Juang, IEEE ASSP Magazine, January 1986, or in the book “Inference in Hidden Markov Models" by Capcher, Moulines and Ryden of Springer, from the series “Springer series in statisctics”.
• Figure 3 illustrates an example of a system user registration of the first example, on the lower graph, and the result provided by the system which indicates that the user has been in the seated posture (state 2) for 50 seconds , then in the walking posture (state 4) for 85 seconds, then in the standing posture (state 1) for 50 seconds, then in the walking posture (state 4) for 61 seconds, then in the sitting posture (state 2 ) for 8 seconds, then in the lying posture (state 3) for 94 seconds, then in the walking posture (state 4) for 54 seconds, and finally finishes in the sitting posture (state 2).
• the present invention makes it possible, at reduced cost and with improved accuracy, to determine in real time or deferred the posture of a person, by accurately determining the changes in posture.

Landscapes

• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Engineering & Computer Science (AREA)
• Physics & Mathematics (AREA)
• General Health & Medical Sciences (AREA)
• Animal Behavior & Ethology (AREA)
• Surgery (AREA)
• Veterinary Medicine (AREA)
• Public Health (AREA)
• Biophysics (AREA)
• Molecular Biology (AREA)
• Physiology (AREA)
• Biomedical Technology (AREA)
• Pathology (AREA)
• Medical Informatics (AREA)
• Heart & Thoracic Surgery (AREA)
• Oral & Maxillofacial Surgery (AREA)
• Dentistry (AREA)
• Neurology (AREA)
• Neurosurgery (AREA)
• Human Computer Interaction (AREA)
• Computer Vision & Pattern Recognition (AREA)
• Theoretical Computer Science (AREA)
• Multimedia (AREA)
• Psychiatry (AREA)
• General Physics & Mathematics (AREA)
• Social Psychology (AREA)
• Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Priority Applications (4)

Applications Claiming Priority (2)

Publications (1)

Family

ID=41200528

Family Applications (1)

Country Status (6)

Cited By (5)

Families Citing this family (20)

Citations (3)

Family Cites Families (4)

• 2010
  • 2010-04-26 CN CN2010800224939A patent/CN102438521A/zh active Pending
  • 2010-04-26 KR KR1020117026765A patent/KR20120027229A/ko not_active Withdrawn
  • 2010-04-26 EP EP10715545A patent/EP2421436A1/fr not_active Withdrawn
  • 2010-04-26 WO PCT/EP2010/055562 patent/WO2010122174A1/fr not_active Ceased
  • 2010-04-26 US US13/264,965 patent/US9445752B2/en active Active
  • 2010-04-26 JP JP2012506530A patent/JP2012524579A/ja active Pending

Patent Citations (3)

Non-Patent Citations (6)

Also Published As

Similar Documents

Legal Events