WO2010107091A1 - 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 - Google Patents
生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 Download PDFInfo
- Publication number
- WO2010107091A1 WO2010107091A1 PCT/JP2010/054701 JP2010054701W WO2010107091A1 WO 2010107091 A1 WO2010107091 A1 WO 2010107091A1 JP 2010054701 W JP2010054701 W JP 2010054701W WO 2010107091 A1 WO2010107091 A1 WO 2010107091A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- signal
- sensor
- sensors
- noise
- monitoring
- Prior art date
Links
Images
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—Detecting, measuring or recording 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, mobility of a limb
- A61B5/1102—Ballistocardiography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6887—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient mounted on external non-worn devices, e.g. non-medical devices
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60K—ARRANGEMENT OR MOUNTING OF PROPULSION UNITS OR OF TRANSMISSIONS IN VEHICLES; ARRANGEMENT OR MOUNTING OF PLURAL DIVERSE PRIME-MOVERS IN VEHICLES; AUXILIARY DRIVES FOR VEHICLES; INSTRUMENTATION OR DASHBOARDS FOR VEHICLES; ARRANGEMENTS IN CONNECTION WITH COOLING, AIR INTAKE, GAS EXHAUST OR FUEL SUPPLY OF PROPULSION UNITS IN VEHICLES
- B60K28/00—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions
- B60K28/02—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver
- B60K28/06—Safety devices for propulsion-unit control, specially adapted for, or arranged in, vehicles, e.g. preventing fuel supply or ignition in the event of potentially dangerous conditions responsive to conditions relating to the driver responsive to incapacity of driver
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2503/00—Evaluating a particular growth phase or type of persons or animals
- A61B2503/20—Workers
- A61B2503/22—Motor vehicles operators, e.g. drivers, pilots, captains
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2560/00—Constructional details of operational features of apparatus; Accessories for medical measuring apparatus
- A61B2560/02—Operational features
- A61B2560/0242—Operational features adapted to measure environmental factors, e.g. temperature, pollution
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
- A61B2562/02—Details of sensors specially adapted for in-vivo measurements
- A61B2562/0247—Pressure sensors
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
- A61B2562/04—Arrangements of multiple sensors of the same type
- A61B2562/046—Arrangements of multiple sensors of the same type in a matrix array
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6887—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient mounted on external non-worn devices, e.g. non-medical devices
- A61B5/6893—Cars
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2540/00—Input parameters relating to occupants
- B60W2540/22—Psychological state; Stress level or workload
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2540/00—Input parameters relating to occupants
- B60W2540/221—Physiology, e.g. weight, heartbeat, health or special needs
Definitions
- the present invention relates to the monitoring of biological parameters of occupants of seat or bed members.
- the present invention relates to the extraction and monitoring of biological parameters of a vehicle occupant, in particular, but not limited to, a driver or a passenger.
- a vehicle occupant in particular, but not limited to, a driver or a passenger.
- it tries to extract the heart beat and breathing signals of a person so that they are not restrictive and if possible under all driving conditions.
- obtaining these biological parameters makes it possible to use a monitoring system to improve road traffic safety.
- Patent Document 1 presents a method for monitoring the heartbeat of a seat occupant.
- the seat includes a piezoelectric sensor and extracts the heart beat signal by analyzing the signal.
- the signal transmitted from each sensor includes noise that is independent of the desired biological signal.
- the above document discloses applying a passband filter and considering only the frequency range including the frequency of the phenomenon under consideration in the signal.
- noise present in the signal coming out of the sensor includes the entire frequency range. Therefore, the passband filter cannot isolate a desired signal by removing this noise. In other words, even if it is limited to a certain frequency range by the passband filter, a desired signal cannot be obtained due to the amount of noise that remains in this range. Moreover, if it is desired to use such a method in a vehicle to monitor the biological parameters of the driver or occupant, a satisfactory result can be obtained by the noise level caused by the vibration on the vehicle. I can't.
- an object of the present invention is to improve the estimation of biological signals in environments exposed to strong noise.
- a method for monitoring at least one biological parameter of a heart beat and / or respiratory signal of an occupant of a seat or bed member, wherein the method is coupled to the member Then, the received signal or each received signal from one or more sensors capable of detecting pressure fluctuation due to contact is processed by non-linear filtering.
- non-linear filtering gives good results as long as the observed phenomenon acts as a non-linear system. Therefore, by non-linear filtering, parameters can be estimated in the presence of noise in the collected signal, and thus signals relating to the parameter to be monitored can be easily extracted.
- the filtering includes a Bayesian recursive estimator such as an extended Kalman filter or an individual filter.
- Bayesian tools is particularly effective when estimating values such as frequencies that cannot be observed directly in such a situation.
- Using an extended Kalman filter in the presence of a nonlinear system gives particularly good results if the nonlinearity can be modeled.
- the individual filter is particularly adapted to better identify noise fluctuations that are not Gaussian noise.
- x k + 1 A ⁇ x k + v k
- x k + 1 a vector indicating the following types of states:
- v k is white Gaussian noise.
- H is a matrix relating the state x k to the measurement y k
- n k is the noise observed by the observer.
- Receiving the signal of each sensor in a group of sensors This signal is input to an atom dictionary (dictionaire d'atomes in French) Select several sensors by input, Monitoring is performed using only selected sensors.
- atom dictionary dictionaire d'atomes in French
- this monitoring method can be adapted in real time to the characteristics of the body of the occupant of the member, its posture, changes in posture, or also the movement of the occupant.
- the method can monitor biological parameters by simultaneously adapting to the monitoring conditions spatially and temporally without the need for operator intervention.
- the method facilitates obtaining the most reliable data regarding the biological parameter to be monitored.
- the vibration level when mounted on a vehicle, the vibration level is configured such that most of the reliability of the present method depends on the type of processing performed in the previous stage in order to consider ambient noise.
- data indicating the biological parameter to be monitored can often be appropriately extracted therefrom.
- This problem is not limited to monitoring performed onboard a vehicle. If monitoring involves, for example, a patient occupying a bed, ambient noise (eg noise generated by appliances) can interfere with this monitoring, so that the reliability of the method is here But similarly, it depends on the quality of the processing performed to remove the noise from the signal used. In a medical bed, this problem occurs when the patient is transported because vibration occurs when the bed moves.
- At least one accelerometer is coupled to the member, Determining a model of a transfer function between at least one signal of an accelerometer at the input or at least one accelerometer of the at least one accelerometer and a signal at one sensor of the sensor or sensors at the output; , The model estimates the noise value, The estimated noise value is removed from the sensor signal.
- this accelerometer or each accelerometer senses vibration by definition. Therefore, a particularly faithful reference signal of ambient noise can be supplied. Furthermore, by determining a transfer function model, the noise value can be estimated with high reliability. Therefore, most of the noise contained in the signal to be used can be removed. This therefore does not directly remove the signal read by this accelerometer or each accelerometer from the signal to be used, i.e. does not subtract the signal.
- the accelerometer signal is used to model the effect of noise on the signal to be used, and this noise estimate is more appropriately extracted from the signal to be used. As a result, a utilization signal that does not contain noise is obtained in the previous stage. The noise still there is not an obstacle to obtaining a signal that is faithfully indicative of one or more biological parameters to be utilized.
- this method is used by being mounted on a vehicle.
- a computer program including code instructions capable of controlling the execution of the method according to the present invention when executed on a computer or a computer is constituted.
- a data recording medium containing such a program in a recorded form.
- it is further configured to use and provide such a program on a remote communication network for downloading.
- a monitoring device of at least one biological parameter of a heart beat and / or respiratory signal of an occupant of a seat or bed member comprising: At least one sensor coupled to the member and capable of detecting pressure fluctuations due to contact; And a sensor or a reception signal from each sensor or processing means for each reception signal by non-linear filtering.
- FIG. 1 is an overall view showing an apparatus according to an embodiment of the present invention. It is a side view of the seat couple
- FIG. 3 shows one of the sensor arrays of the seat of FIG. It is a graph which shows the atom of the dictionary utilized with the method of this invention. It is a figure which shows modeling of the transfer function in this embodiment.
- FIG. 4 is a diagram showing that a transfer function modeled in this way is taken into account when the present method is carried out. It is a graph which shows the amplitude of the signal before extraction of the noise by a transfer function, and after extraction of noise over time. It is a figure which shows the heart beat estimated by this method, and the heart beat of an actual heart with two curves, respectively, and the acceptable tolerance.
- the present embodiment describes an embodiment of the method and related apparatus according to the invention that constitutes a complete system for extracting biological information from the body of a person 2 who is an occupant or driver of a vehicle 4.
- the present invention seeks to monitor a person's biological parameters such as heart beat and / or respiratory rhythm. This monitoring is performed under various driving conditions and is desired to be reliable with respect to body movement. This relates to obtaining and monitoring the above biological parameters. In particular, it is desirable to obtain the above parameters so that no matter what the driving conditions are, the person involved is not restrained. In fact, obtaining information about heart beats and / or respiratory signals provides a monitoring system that reduces road drowsiness or some illness-related accidents and improves road traffic safety.
- the apparatus includes a plurality of piezoelectric sensors 6 supported by a seat 8 occupied by a person 2. These sensors are configured so that a desired signal can always be obtained by the sensor regardless of the position of the occupant or the movement of the occupant.
- the sensor can detect a change in contact pressure, and is a piezoelectric sensor here.
- the sensor is disposed in the vicinity of the main surface above the seat 10 and the backrest 12 of the seat 8, and this surface is provided adjacent to the occupant 2.
- the sensor can also be placed directly on this surface. In this way, the sensor receives body pressure, particularly in the vicinity of the artery, and in particular pressure fluctuations arising from the body.
- the sensor is made of a film or sheet. However, as will be described later, these sensors sense any type of mechanical vibration, and the desired signal cannot be observed directly at the output of such sensors.
- the seat 8 includes about 60 sensors, but it should be understood that this number is not limiting.
- the number of sensors in the seat portion 10 may be 10 to 70, for example, 40.
- the number of sensors in the backrest 12 is 5 to 50, and can be, for example, 20.
- the substrate used during the use of the method comprises, for example, 20 sensors, i.e. 15 in the seat and 5 in the back.
- the substrate includes, for example, about 30% of the sensors present in the seat.
- the sensor signal processing uses an analog part and a digital part.
- the analog portion includes an amplitude collecting step and a knitting step of the output signal of the sensor 6.
- Each of the analog outputs is digitized and then only a few sensors are automatically selected depending on the position of the body 2 in the seat 8.
- the signals sent from these sensors are then processed and fused.
- Such selection and fusion steps of the sensor are performed multiple times during the implementation of the method so as to predict this appropriately according to body movements. It is noted here that the movements of the vehicle occupant and the driver are generally not the same type. The occupant movement is more random, but the number is less than the driver movement. In the first place, the driver's movement must take into account such characteristics as the presence of four drive wheels in the vehicle, hand movement, acceleration, braking, and the position of the legs for gear changes.
- an apparatus for selecting and fusing sensor signals is indicated by a block 14.
- This block can sense body movement and track and predict this movement.
- the block can select the sensor that is most capable of providing a signal that is effective for obtaining a biological parameter.
- This block can accurately detect one motion and predict one motion to obtain a list of candidate sensors to select. Therefore, the method is continuously performed without variation regardless of the occupant's movement.
- the block can also determine if only inertial objects are placed on the member and if so do nothing.
- the seat 8 is further provided with an accelerometer 16 that acts as a reference sensor for ambient noise such as vibration noise caused by the vehicle.
- Each of these accelerometer sensors 16 senses noise in three directions perpendicular to one another, namely the horizontal directions X, Y and the vertical direction Z.
- the transfer function is reestimated each time it turns out that it is necessary to track the current driving conditions of the vehicle and body movements.
- block 20 serves to create a dynamic model that changes over time to estimate the transfer function between the accelerometer and the piezoelectric sensor.
- Various values relating to noise can be predicted by the above-described transfer function estimated as described above, and then the noise transmitted by the signal of the piezoelectric sensor can be reduced.
- the model used at this stage may be linear or non-linear. This is the first step of reducing noise in the signal supplied by the sensor 6.
- block 20 also includes a second estimation and tracking step that can more accurately obtain biological parameters such as heartbeat and / or respiratory signals.
- This step can better track biological parameter variations whatever the driving conditions (body movement, urban driving, high speed driving and highway driving).
- This second step is configured to adapt to the nonlinear system.
- This step can include an extended Kalman filter or an individual filter to better identify any noise variation.
- Such a non-linear processing step is just suitable for extracting and monitoring parameters derived from non-linear models such as heart rhythm and respiratory rhythm in a vibration environment that varies over time.
- block 22 can obtain the desired signal, monitor it and predict its change.
- the present invention is suitable for all types of vehicles. However, the present invention is not limited to vehicles. As such, it can be used with other types of members, such as a seat or bed, eg, a medical bed for monitoring patient parameters.
- the method used is self-adaptive.
- the seat 8 now includes two sensor arrays 6 disposed on a seat portion 10 and a backrest portion 12, respectively.
- Each of these arrays in this case includes a plurality of rows and a plurality of columns.
- the array includes 5 rows and 5 columns to form a grid-like sensor.
- the sensors in each array are substantially in the same plane.
- the sensors are suitably electrically connected to other parts of the device and transmit signals read by these sensors to the means 14 and 20.
- Block 14 serves to select only a few available sensors. This relates to, for example, selecting two sensors at the seat 10 and two sensors at the backrest 12. However, this number can be reduced or increased as needed.
- the vehicle engine is off. Drivers and passengers enter the vehicle.
- the device 7 according to the invention comprising means 14, 20, 22 commands the use of the method.
- the processing member and the calculation member of the apparatus 7 are accommodated in the dashboard 11, for example.
- the device 7 When the use of this method is started, the device 7 basically activates the default substrate of the sensor 6 selected from the sensors of the seat part and the back part.
- This substrate has a memory. These default substrates form part of the basic adjustment of the method.
- signals are transmitted from several sensors 6.
- the means 14 analyzes these signals to estimate whether it is preferable to adjust the substrate in view of various situations, in particular the characteristics of the body 2 and its posture at the seat.
- the means 14 also analyzes whether an object is present on the member instead of a person. If so, no additional processing is performed.
- the means 14 basically identifies the sensor 6 that supplies the most available signal. Sensors that do not supply any signals are not considered.
- the sensor signal that provides the high pressure signal is also not considered. This is because this method examines pressure fluctuations in particular.
- the sensor placed under the buttock of the person 2 receives the weight of most of the person's torso and provides a relatively poor signal in information. These are generally not considered.
- the substrate most relevant to the means 14 must be obtained, in this case the substrate with the sensor most sensitive to slight movement, ie slight pressure fluctuations. Therefore, the base substrate is adapted by adding or removing several sensors. In a substrate adapted in this way, the selection of the sensor is performed at a later stage of the method.
- block 14 is configured to select a new sensor that can be associated therewith.
- block 14 completely refreshes the sensor's substrate so that it can better adapt to the situation.
- the block selects the most appropriate sensor, i.e. the sensor most likely to contain the desired signal related to the biological parameter, in the presence of a given sensor substrate.
- the weight can be estimated using atom decomposition that provides high frequency accuracy, and the position of the component to be tracked can be located. This is an extension by a kind of wavelet packet.
- the signal of one sensor is shown here as a linear combination of the expansion functions f m, n .
- the signal x is a column vector (N ⁇ 1 formant)
- ⁇ is a column vector of an expansion coefficient (N ⁇ 1)
- F is an N ⁇ M matrix whose columns are expansion functions f m, n .
- One signal model is provided by one linear combination of expansion coefficients and various functions. Compact models tend to include extended functions that are highly correlated with the signal.
- Atom dictionaries suitable for a wide range of time-frequency behavior can be prepared, and signals can be decomposed by selecting several appropriate atoms in the atom dictionary.
- This dictionary is constructed as follows.
- the dictionary is composed of sine waveforms and cosine waveforms of various possible frequencies (however, frequencies limited to the monitoring range of the present invention).
- the strongest frequency is 20 Hz, so in this embodiment it is preferable to use a frequency range between 0.2 Hz and 3 Hz for a single respiratory signal.
- a range from 0.7 to 20 hertz is used, and in either case, one pitch is 0.1 hertz.
- each atom in the dictionary is weighted by a Hanning window, thereby avoiding edge effects in particular. for that reason, here
- m indicates the length of the atom. In practice, this length is an important value for frequency resolution. In the absence of such weighting, atoms may have different durations in the dictionary.
- the dictionary consists of N weighted sine atoms and N weighted cosine atoms.
- these atoms form a small support signal (ie, a limited, non-zero finite number of points) that can be identified by analogy with wavelet packets.
- FIG. 4 shows one of the atoms in the dictionary that just includes the concatenation of one sine atom and one cosine atom for one frequency.
- the signal shows time on the horizontal axis and amplitude on the vertical axis.
- the atom length can be measured here between 0 and 2000 points on the horizontal axis, and time is selected as the number of samples (depending on the sampling frequency).
- a normalization coefficient is calculated for each set of atoms.
- the signal f (that is, its pulse response) supplied from each sensor on the substrate is input to each set of the atom dictionary, and a value is calculated as follows. sup
- C (f, g ⁇ , h ⁇ ) is a distance function.
- the following distance function is selected.
- C (f, g ⁇ , h ⁇ ) ⁇ 2, ⁇ ⁇ ( ⁇ f, g ⁇ > 2 + ⁇ f, h ⁇ > 2 -2 ⁇ 1, ⁇ ⁇ f, g ⁇ > ⁇ ⁇ f, h ⁇ > )
- This distance function makes it possible to accurately determine the position of components (resonance caused by body weight and vehicle vibration) that occur simultaneously in the human body and system.
- a sensor is considered qualified if the signal transmitted by the sensor simultaneously includes system parameters and human biological parameters. It is not eligible if it contains only system parameters.
- the method is capable of predicting human body motion in order to properly retain the relevant set of signals (occupant biological parameters). Therefore, the following approach is used in this example.
- the occurrence of body movement is detected by a sensor on the board where the supplied signal starts to fluctuate.
- the sensors on the substrate at the backrest are the sensors identified by their criteria (i, j), (i + 1, j + 1), and (i, j + 2).
- Means 14 identifies the movement by these sensors, and this same means can predict the direction of this movement, indicated by the arrow 24 in FIG. 3, by interpolation. Thus, while such a movement takes place, the means 14 predicts the future movement, and during the movement, relative to the selected sensor substrate, is in the course of movement and is advantageous during the future movement.
- a sensor capable of supplying a signal is added. In FIG. 3, two sensors (i + 2, j + 2) and (i + 1, j + 2) in the same row as the sensor (i, j + 2) correspond to this. Therefore, the means 14 can take into account the expected signals supplied from these sensors as soon as possible. Thereafter, when it is confirmed that movement is recognized at the position of these sensors, these sensors are held on the substrate. On the other hand, if the prediction is incorrect and no movement is detected for at least one of the sensors, the sensor is removed from the substrate.
- a plurality of accelerometers 16 is used to serve as a reference for vibration noise in the vehicle.
- a three-axis or 3D accelerometer is preferably used unless on-vehicle vibrations are exerted in a single direction, eg, only in a vertical direction. Furthermore, it is preferable to use at least two accelerometers.
- Identifying the location of these accelerometers may prove important to obtain a reliable model. For example, by placing one accelerometer 16 under the seat 10 structure in the seat, the accelerometer senses the occupant's vibrations under the seat. In this embodiment, the second accelerometer is placed on top of the backrest 12. This is because it has been confirmed that this part of the seat can vibrate with some independence from the seat.
- the transfer function modeling performed thereafter can be linear or non-linear.
- these modeling parameters are in this case estimated by a recursive procedure.
- the above parameters are often reestimated during the implementation of the method so that the model adapts precisely to various situations, in particular operating conditions.
- FIG. 5 shows the principle of transfer function modeling. This is responsible for the identification of the function 11 and its parameters, with the x, y, z signals supplied by the two accelerometers at the input and the piezoelectric sensor considered at the output. Is supplied with the signal s. In this way, a modeling of the inherent effects of vibration in the signal supplied by the piezoelectric sensor is obtained.
- the means 20 first determines a model of the most suitable type for obtaining a transfer function from a list of a plurality of types of models according to the situation.
- This list is in this case as follows: Modeling by showing the state, ARMA, ARX, NLARX.
- the noise value is dynamically determined according to the instantaneous signal supplied by the accelerometer using the model identified in this way.
- the input is six signal values of the accelerometer.
- the output is an estimated value due to a single noise on the signal side of the piezoelectric sensor.
- the estimated value is subtracted from the signal supplied from the piezoelectric sensor 6 at the position of the subtractor 13. Thereby, after this subtraction, a signal from which most of the influence of vibration noise is removed is obtained.
- the means 20 uses an ARX type external autoregressive model by default. This means that this type of model is used when better results are not obtained from any of the other types of models in the list. Otherwise, the model that provides the best results is used.
- N C N A + N i ⁇ N B
- the polynomial coefficients are estimated by minimizing the error prediction covariance matrix trace. As mentioned earlier, these estimates of parameters are updated from time to time as operating conditions change. Once the model parameters are estimated at each sampling step, the predicted noise (ie, estimated noise) in the piezoelectric sensor can be calculated. In that case, such estimated noise is removed by the output of the piezoelectric sensor as shown in FIG.
- the signal 15 of the piezoelectric sensor 6 before subtraction is indicated by a thin line
- the signal 17 after subtraction is indicated by a thick line.
- the amplitude (unit: volt) of the ordinate signal is remarkably reduced after subtraction, and a peak corresponding to the heartbeat clearly appears.
- m is the number of sinusoidal waveform components
- a i (t) indicates the amplitude of the sinusoidal waveform component
- ⁇ i (t) represents the phase difference between the fundamental pulse and the harmonic.
- phase difference ⁇ i (t) between the fundamental pulse and the harmonic component and the instantaneous change in phase ⁇ k, i (t) over time is obtained by the following equation.
- ⁇ k + 1, i i ⁇ ⁇ k + ⁇ k, i + v k, i ⁇
- the estimation of the variance of the noise v k component will affect the rate of change of the estimated parameters (pulse, amplitude component and phase component) and the convergence rate of the algorithm.
- the equation showing the observation is non-linear, which supports the use of the extended Kalman filter.
- the observation noise variance nk is related to the noise variance observed in the piezoelectric signal.
- the extended Kalman filter algorithm is executed as follows.
- the prediction formula of the extended Kalman filter is as follows.
- the updated formula is as follows: here Further, where the Q w and R eta, a covariance matrix of each v k and n k, I is an identity matrix.
- this algorithm is adapted as follows. noise and Is assumed to be equal to zero.
- the size of the state vector and the observation vector can also be increased.
- a state vector, a state transition matrix, and a linear matrix indicating observation are as follows.
- one or more piezoelectric sensors can be processed to find the heart beat (and / or respiratory signal).
- This model of state is given as an example and is not limiting.
- Fig. 8 shows the result of the above process as an example.
- Curve 30 shows the actual heart beat signal
- curve 32 shows the heart beat signal estimated by the method of the present invention
- curves 34 and 36 show a tolerance threshold of 5% above and below the actual signal amplitude. showed that.
- These curves indicate, on the ordinate, the number of beats per minute corresponding to the time (unit: seconds) indicated on the abscissa. It can be seen that the deviation between the actual signal and the signal estimated by the method of the present invention is frequently included in a tolerance range of 5% above and below the actual signal. This relates to the results of tests performed under actual driving conditions (urban driving, highway driving, etc.).
- Means 14, 20, and 22 include calculation means such as a microprocessor, and include one or more computers having one or more memories, for example.
- the method according to the present invention can be automatically carried out by a computer program recorded on a data recording medium such as a hard disk, flash memory, CD or DVD disk.
- the computer program includes code instructions that can control the performance of the method according to the invention when executed on a computer.
- Such a program can be configured to be utilized and provided over a telecommunications network for its download, eg, download of an updated program version.
- a unique atom dictionary can be associated with each of the frequency ranges, so in this case two atom dictionaries can be used.
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Heart & Thoracic Surgery (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Combustion & Propulsion (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Dentistry (AREA)
- Chemical & Material Sciences (AREA)
- Mechanical Engineering (AREA)
- Physiology (AREA)
- Cardiology (AREA)
- Transportation (AREA)
- Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
Description
ここで、
φ1(t)=ω(t)・t
φi(t)=i・ω(t)・t+θi(t)、i=2...mであり、
y(t)は、信号を示し、
ω(t)は、信号の瞬間的な基本パルスを示し、
mは、正弦関数の数であり、
ai(t)は、正弦関数の振幅を示し、
φi(t)は、高調波の瞬間的な位相を示し、
θi(t)は、基本パルスと高調波との間の位相差を示す。
ここで、
は、観察者による信号y(k)の推定値であり、
Hは、状態xkを測定値ykに関連づける行列であり、
および
は、それぞれ、観察者によるai,kおよびφi,kの推定値であり、
nkは、観察者により観察されたノイズである。
一群のセンサの中の各センサの信号を受信し、
この信号をアトム辞書(仏語でdictionnaire d’atomes)に投入し、
投入により幾つかのセンサを選択し、
選択されたセンサだけを用いて監視を実施する。
入力における加速度計または少なくとも1つの加速度計のうちの1つの加速度計の少なくとも1つの信号と、出力におけるセンサまたは複数のセンサのうちの1つのセンサの信号との間の伝達関数のモデルを決定し、
モデルによりノイズ値を推定し、
推定されたノイズ値をセンサの信号から除去する。
座席またはベッドの部材の占有者の心臓の拍動および/または呼吸信号の少なくとも1つの生物学的パラメータの監視装置を構成し、本装置は、
部材に結合されて、接触による圧力変動を検知可能な少なくとも1つのセンサと、
非線形フィルタリングによりセンサまたは各センサからの受信信号または各受信信号の処理手段とを含む。
図1から3を参照すると、座席8は、ここでは、座部10と背もたれ部12とにそれぞれ配置された2個のセンサアレイ6を含む。これらのアレイの各々は、この場合には、複数の行と複数の列を含む。例えば、図3に示したように、アレイは、5個の行と5個の列を含んで碁盤状のセンサを形成する。各アレイのセンサは、ほぼ同一平面にある。センサは、装置の他の部分に適切に電気接続され、これらのセンサにより読み取られる信号を手段14および20に伝達する。
ここでは、時間-周波数の同時分析を実施する。そのため、高い周波数精度を提供するアトム分解を用いて重量を推定し、追跡すべき成分の位置を突き止めることができる。これは、一種のウェーブレットパケットによる拡張である。
x=Fα、ここでF=[f1,f2,...,fM]
ここで
ここで、mはアトムの長さを示す。実際には、この長さは、周波数解像度に対して重要な値である。こうした加重がない場合、アトムは、辞書において異なる持続時間を有する可能性がある。
sup|C(f,gγ,hγ)|
ここで、C(f,gγ,hγ)は距離関数である。この例では、次の距離関数を選択する。
C(f,gγ,hγ)=φ2,γ・(<f,gγ>2+<f,hγ>2-2φ1,γ<f,gγ>・<f,hγ>)
本方法は、関連する信号の集合(占有者の生物学的パラメータ)を適切に保持するために、人間の身体の動きを予測することができる。そのため、この例では以下のアプローチを用いる。
車両における振動ノイズの基準の役割を果たすために、複数の加速度計16が使用される。車上振動が単一の方向に及ぼされるのでない限り、例えば単に垂直方向に及ぼされるだけでない限り、好適には、3軸または3D加速度計を用いる。さらに、少なくとも2個の加速度計を使用することが好ましい。
状態を示すことによるモデル化、
ARMA、
ARX、
NLARX。
ここで、
A(q)は、NA個の係数を有する多項式である。
y(t)は、圧電センサの出力信号である。
Bi(q)は、NB個の係数を有する多項式である。
ui(t)(i=1...Ni)は、加速度計により供給される入力信号である。
nkiは、入力における単位遅延数である。
e(t)は、このモデルのエラー信号である。
NC=NA+Ni・NB
こうしたノイズ除去ステップの後、手段20により、心律動と呼吸信号を抽出することが必要である。これは、周波数を直接観察できないパラメータを推定することに帰する。従って、ベイズ推定に位置づけることが特に有効である。さらに、検討されたシステムは非線形であるので、拡張カルマンフィルタを用いることができる。ガウスノイズではないノイズ変動をより密接に認識するために個別フィルタを用いてもよい。
ここで、φ1(t)=ω(t)・t
φi(t)=i・ω(t)・t+θi(t)、i=2...mであり
ω(t)は、心臓の拍動に関連する信号の基本パルスを示す。
mは、正弦波形成分の数であり、
ai(t)は、正弦波形成分の振幅を示し、
φi(t)、i=2...mは、高調波の瞬間的な位相を示し、
θi(t)は、基本パルスと高調波との間の位相差を示す。
ak+1,i=ak,i+vk,i a
ωk+1=ωk+vk ω
φk+1,i=i・ωk+φk,i+vk,i φ
更新される式は次のようになる。
ここで
また、ここで、QwとRηは、それぞれvkとnkの共分散行列であり、Iは、単位行列である。
Claims (10)
- 座席またはベッドの部材の占有者の心臓の拍動および/または呼吸信号の少なくとも1つの生物学的パラメータを監視する方法であって、部材に結合されて接触による圧力変動を検知可能な1つまたは複数のセンサからの受信信号または各受信信号を、非線形フィルタリングにより処理することを特徴とする、生物学的パラメータを監視する方法。
- 前記フィルタリングが、拡張カルマンフィルタまたは個別フィルタ等のベイズ型の再帰的推定量を含む、請求項1に記載の方法。
- 一群のセンサの中の各センサの信号を受信し、
この信号をアトム辞書に投入し、
投入により幾つかのセンサを選択し、
選択されたセンサだけを用いて監視を実施する、請求項1~5のいずれか一項に記載の方法。 - 少なくとも1つの加速度計が部材に結合されており、
入力における加速度計または少なくとも1つの加速度計のうちの1つの加速度計の少なくとも1つの信号と、出力におけるセンサまたは複数のセンサのうちの1つのセンサの信号との間の伝達関数のモデルを決定し、
モデルによりノイズ値を推定し、
推定されたノイズ値をセンサの信号から除去する、請求項1~6のいずれか一項に記載の方法。 - 車両に搭載して利用される、請求項1~7のいずれか一項に記載の方法。
- コンピュータまたは計算機での実行時に請求項1~8のいずれか一項に記載の方法のステップの実行を制御可能なコード命令を含むことを特徴とするコンピュータプログラム。
- 座席またはベッドの部材の占有者の心臓の拍動および/または呼吸信号の少なくとも1つの生物学的パラメータの監視装置であって、
部材に結合されて、接触による圧力変動を検知可能な少なくとも1つのセンサと、
非線形フィルタリングによりセンサまたは各センサからの受信信号または各受信信号の処理手段とを含むことを特徴とする、生物学的パラメータの監視装置。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201080012705.5A CN102355850B (zh) | 2009-03-18 | 2010-03-18 | 生物学参数的监视方法及监视装置 |
EP10753582A EP2409638A4 (en) | 2009-03-18 | 2010-03-18 | BIOLOGICAL PARAMETER MONITORING METHOD, COMPUTER PROGRAM, AND BIOLOGICAL PARAMETER MONITORING DEVICE |
JP2011504883A JP5382110B2 (ja) | 2009-03-18 | 2010-03-18 | 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 |
US13/257,180 US20120197138A1 (en) | 2009-03-18 | 2010-03-18 | Biological parameter monitoring method, computer-readable storage medium and biological parameter monitoring device |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR0951714A FR2943233B1 (fr) | 2009-03-18 | 2009-03-18 | Procede de surveillance d'un parametre biologique d'une personne au moyen d'un filtrage non-lineaire bayesien |
FR0951714 | 2009-03-18 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010107091A1 true WO2010107091A1 (ja) | 2010-09-23 |
Family
ID=41210883
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2010/054701 WO2010107091A1 (ja) | 2009-03-18 | 2010-03-18 | 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 |
Country Status (6)
Country | Link |
---|---|
US (1) | US20120197138A1 (ja) |
EP (1) | EP2409638A4 (ja) |
JP (1) | JP5382110B2 (ja) |
CN (1) | CN102355850B (ja) |
FR (1) | FR2943233B1 (ja) |
WO (1) | WO2010107091A1 (ja) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011200509A (ja) * | 2010-03-26 | 2011-10-13 | Aisin Seiki Co Ltd | 生体情報検出装置 |
CN104554071A (zh) * | 2013-10-22 | 2015-04-29 | 大陆汽车电子(长春)有限公司 | 防止驾驶员突发疾病而引发事故的方法和装置 |
JP2015093038A (ja) * | 2013-11-12 | 2015-05-18 | 沖電気工業株式会社 | 情報処理装置、情報処理方法及びプログラム |
JP2016136989A (ja) * | 2015-01-26 | 2016-08-04 | アイシン精機株式会社 | 生体情報検出装置 |
JP2016187555A (ja) * | 2015-03-27 | 2016-11-04 | イムラ・ヨーロッパ・エスアエスImra Europe S.A.S. | 生体パラメータ推定装置または方法 |
JP2017012650A (ja) * | 2015-07-06 | 2017-01-19 | アイシン精機株式会社 | 生体情報検出装置 |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
LU92090B1 (en) * | 2012-10-29 | 2014-04-30 | Iee Sarl | Piezoelectric and/or electret sensing device |
US9642575B2 (en) * | 2013-03-15 | 2017-05-09 | Zoll Medical Corporation | ECG noise reduction system for removal of vehicle motion artifact |
EP3218674B1 (en) * | 2014-11-11 | 2021-06-23 | Intel Corporation | Extended kalman filter based autonomous magnetometer calibration |
CN109416288B (zh) * | 2016-04-22 | 2022-11-11 | 株式会社村田制作所 | 监测*** |
CN108599737B (zh) * | 2018-04-10 | 2021-11-23 | 西北工业大学 | 一种变分贝叶斯的非线性卡尔曼滤波器的设计方法 |
CN109606316A (zh) * | 2018-10-24 | 2019-04-12 | 惠州市德赛西威汽车电子股份有限公司 | 一种驾驶员身份识别方法 |
CN109497970A (zh) * | 2018-11-29 | 2019-03-22 | 江苏医药职业学院 | 一种远程健康监测装置 |
JP7156210B2 (ja) * | 2019-08-09 | 2022-10-19 | トヨタ自動車株式会社 | 車両用情報提供システム |
GB2589337A (en) * | 2019-11-27 | 2021-06-02 | Continental Automotive Gmbh | Method of determining fused sensor measurement and vehicle safety system using the fused sensor measurement |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR951714A (fr) | 1946-09-27 | 1949-11-02 | Houdaille Hershey Corp | Perfectionnements aux serrures de portes |
JP2005161039A (ja) * | 2003-11-14 | 2005-06-23 | Siemens Medical Solution Health Services Corp | 患者医療パラメータの処理システム、処理方法、及びデータの伝送と保存の所要量の低減方法 |
WO2007143535A2 (en) * | 2006-06-01 | 2007-12-13 | Biancamed Ltd. | Apparatus, system, and method for monitoring physiological signs |
JP2008079931A (ja) * | 2006-09-28 | 2008-04-10 | Aisin Seiki Co Ltd | 心拍検出装置 |
US20080103702A1 (en) | 2006-10-30 | 2008-05-01 | Aisin Seiki Kabushiki Kaisha | Biosignal intensity distribution measuring apparatus and biosignal intensity distribution measuring method |
JP2008113797A (ja) * | 2006-11-02 | 2008-05-22 | Aisin Seiki Co Ltd | 生体信号検出装置 |
Family Cites Families (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH045950A (ja) * | 1990-04-24 | 1992-01-09 | Toyota Central Res & Dev Lab Inc | 心拍計 |
US6024705A (en) * | 1998-11-06 | 2000-02-15 | Bioacoustics, Inc. | Automated seismic detection of myocardial ischemia and related measurement of cardiac output parameters |
US6468234B1 (en) * | 2000-07-14 | 2002-10-22 | The Board Of Trustees Of The Leland Stanford Junior University | SleepSmart |
US6822573B2 (en) * | 2002-01-18 | 2004-11-23 | Intelligent Mechatronic Systems Inc. | Drowsiness detection system |
US8790272B2 (en) * | 2002-03-26 | 2014-07-29 | Adidas Ag | Method and system for extracting cardiac parameters from plethysmographic signals |
US6827695B2 (en) * | 2002-10-25 | 2004-12-07 | Revivant Corporation | Method of determining depth of compressions during cardio-pulmonary resuscitation |
JP3960298B2 (ja) * | 2003-11-19 | 2007-08-15 | 株式会社デンソー | 寝姿及び***検出装置 |
US7565194B2 (en) * | 2004-05-12 | 2009-07-21 | Zoll Medical Corporation | ECG rhythm advisory method |
CN1956680A (zh) * | 2004-05-24 | 2007-05-02 | 日本先锋公司 | 生物信息检测装置 |
JP2006346093A (ja) * | 2005-06-15 | 2006-12-28 | Denso Corp | 車内生体情報検出装置 |
-
2009
- 2009-03-18 FR FR0951714A patent/FR2943233B1/fr not_active Expired - Fee Related
-
2010
- 2010-03-18 JP JP2011504883A patent/JP5382110B2/ja not_active Expired - Fee Related
- 2010-03-18 WO PCT/JP2010/054701 patent/WO2010107091A1/ja active Application Filing
- 2010-03-18 US US13/257,180 patent/US20120197138A1/en not_active Abandoned
- 2010-03-18 CN CN201080012705.5A patent/CN102355850B/zh not_active Expired - Fee Related
- 2010-03-18 EP EP10753582A patent/EP2409638A4/en not_active Withdrawn
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR951714A (fr) | 1946-09-27 | 1949-11-02 | Houdaille Hershey Corp | Perfectionnements aux serrures de portes |
JP2005161039A (ja) * | 2003-11-14 | 2005-06-23 | Siemens Medical Solution Health Services Corp | 患者医療パラメータの処理システム、処理方法、及びデータの伝送と保存の所要量の低減方法 |
WO2007143535A2 (en) * | 2006-06-01 | 2007-12-13 | Biancamed Ltd. | Apparatus, system, and method for monitoring physiological signs |
JP2008079931A (ja) * | 2006-09-28 | 2008-04-10 | Aisin Seiki Co Ltd | 心拍検出装置 |
US20080103702A1 (en) | 2006-10-30 | 2008-05-01 | Aisin Seiki Kabushiki Kaisha | Biosignal intensity distribution measuring apparatus and biosignal intensity distribution measuring method |
JP2008113797A (ja) * | 2006-11-02 | 2008-05-22 | Aisin Seiki Co Ltd | 生体信号検出装置 |
Non-Patent Citations (1)
Title |
---|
See also references of EP2409638A4 |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011200509A (ja) * | 2010-03-26 | 2011-10-13 | Aisin Seiki Co Ltd | 生体情報検出装置 |
CN104554071A (zh) * | 2013-10-22 | 2015-04-29 | 大陆汽车电子(长春)有限公司 | 防止驾驶员突发疾病而引发事故的方法和装置 |
JP2015093038A (ja) * | 2013-11-12 | 2015-05-18 | 沖電気工業株式会社 | 情報処理装置、情報処理方法及びプログラム |
JP2016136989A (ja) * | 2015-01-26 | 2016-08-04 | アイシン精機株式会社 | 生体情報検出装置 |
JP2016187555A (ja) * | 2015-03-27 | 2016-11-04 | イムラ・ヨーロッパ・エスアエスImra Europe S.A.S. | 生体パラメータ推定装置または方法 |
JP2017012650A (ja) * | 2015-07-06 | 2017-01-19 | アイシン精機株式会社 | 生体情報検出装置 |
Also Published As
Publication number | Publication date |
---|---|
EP2409638A1 (en) | 2012-01-25 |
EP2409638A4 (en) | 2012-12-26 |
CN102355850A (zh) | 2012-02-15 |
CN102355850B (zh) | 2014-03-12 |
FR2943233A1 (fr) | 2010-09-24 |
FR2943233B1 (fr) | 2013-10-11 |
JP5382110B2 (ja) | 2014-01-08 |
US20120197138A1 (en) | 2012-08-02 |
JPWO2010107091A1 (ja) | 2012-09-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5382110B2 (ja) | 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 | |
JP5024480B2 (ja) | 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 | |
WO2010107092A1 (ja) | 生物学的パラメータを監視する方法、コンピュータプログラム、および生物学的パラメータの監視装置 | |
Luinge et al. | Inclination measurement of human movement using a 3-D accelerometer with autocalibration | |
JP5919722B2 (ja) | 生体信号推定装置及びプログラム | |
JP4505619B2 (ja) | 心身状態判定システム | |
JP2006346093A (ja) | 車内生体情報検出装置 | |
US11738773B2 (en) | System for controlling autonomous vehicle for reducing motion sickness | |
JP4952688B2 (ja) | 睡眠判定装置及び睡眠判定方法 | |
JPWO2018180331A1 (ja) | 運転者状態検知装置 | |
US9330306B2 (en) | 3D gesture stabilization for robust input control in mobile environments | |
WO2018180330A1 (ja) | 感情推定装置 | |
CN109153332B (zh) | 车辆用户接口的手语输入 | |
WO2019049667A1 (ja) | 心拍検出装置、心拍検出方法及びプログラム | |
US10011176B2 (en) | Method and device for recognising the condition of vehicle occupants | |
JP6738177B2 (ja) | 生体パラメータ推定装置または方法 | |
Murphey et al. | Personalized driver workload estimation in real-world driving | |
KR20210158525A (ko) | 자율주행차 주행 시 멀미 저감 시스템 | |
CN115996666A (zh) | 用于确定认知需求的***和方法 | |
JP6515709B2 (ja) | 生体情報検出装置 | |
JP5852331B2 (ja) | データ処理装置、及びプログラム | |
JP7492257B2 (ja) | 演算装置の作動方法、及び、演算装置 | |
JP6709543B2 (ja) | 脳情報測定装置 | |
WO2022167361A1 (en) | Group biofeedback method and associated system | |
JP2019072456A (ja) | 影響推定装置及びプログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201080012705.5 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 10753582 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011504883 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2010753582 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13257180 Country of ref document: US |