JP5612627B2 - Physical ability determination device and data processing method - Google Patents

Description

The present invention relates to a physical ability determination device and a data processing method for evaluating a physical ability of a subject.

  Locomotive syndrome (motor organ syndrome), along with metabolic syndrome (metabolic syndrome) and dementia, is attracting attention as one of the three major factors that induce a shortened healthy life expectancy, bedridden state or need for care. It has been urgently needed.

  In the decline of physical ability, there is a factor of brain function deterioration as well as muscle strength reduction and skeletal deterioration. While knowledge of orthopedics is required for strength and skeletal strengthening, knowledge of brain science is required for deterioration of brain function. Cross-sectional evaluation (diagnosis) and treatment from both orthopedics and brain science perspectives It is necessary to give guidance.

  As a technique for measuring the state of the body, there is a technique for measuring fluctuations of the body center of gravity and walking motion. For example, Non-Patent Document 1 discusses how changes in mandibular position affect fluctuations in body center of gravity and muscle activity during a human upright posture.

"Fluctuation analysis of body sway due to changes in mandibular position" Gifu Dental Society Vol.35 No.3, 167-181, February 2009

  However, in the conventional technology, from the viewpoints of both orthopedics and brain science, a system for efficiently judging human physical ability, for example, it is possible to measure brain function deterioration and shaping deterioration at the same time, and at low cost and easily. Some systems have not been established.

In order to solve the above problems, an object of the present invention is to provide a physical ability determination device and a data processing method for efficiently determining physical ability.

In order to achieve the above object, a physical ability determination device of the present invention is a physical ability determination device that determines a state of a subject's body,
Based on sensing information supplied from a body state measurement sensor that measures the body motion of the subject, a motion feature amount observation unit that calculates a motion feature amount indicating a feature of the body motion of the subject;
A reference information storage unit that stores reference information for specifying the correspondence between the exercise feature and the evaluation value of physical ability;
Based on the reference information stored in the reference information storage unit, physical ability determination that identifies an evaluation value indicating the physical ability of the subject from the exercise feature amount of the subject calculated by the exercise feature amount observation unit Part
Have
The body condition measurement sensor includes four load sensors, and measures the center-of-gravity movement of the front, rear, left and right during the stepping motion of the subject ,
Correspondence between the Lyapunov index specified from the movement feature amount and the evaluation value of physical ability is described in the reference information,
The physical ability determination unit calculates a Lyapunov index indicating divergence from the motion feature of the subject using a calculation technology of a chaotic dynamical system, and based on the reference information, the body of the subject is calculated from the Lyapunov index of the subject. Identify an evaluation value that indicates your ability,
The motion feature is an attractor obtained by time-series chaos analysis, and from the reconstructed attractor, the Lyapunov exponent is estimated using the Sano-Sawada method, and when the Lyapunov exponent is estimated, a hypersphere The size, the number of neighboring points, and the development time coefficient are used as parameters .

In order to achieve the above object, the data processing method of the present invention is a data processing method in an apparatus that outputs an evaluation value indicating the physical ability of a subject,
The apparatus includes a motion feature amount observation unit, a reference information storage unit, and an evaluation value specifying unit,
The motion feature amount observing unit calculating a motion feature amount indicating a feature of the subject's body movement based on sensing information supplied from a body state measurement sensor that measures the subject's body motion;
The evaluation value specifying unit is calculated by the exercise feature amount observing unit based on reference information for specifying the correspondence between the exercise feature amount and the evaluation value of physical ability stored in the reference information storage unit. A step of identifying an evaluation value indicating the physical ability of the subject from the movement feature amount of the subject,
Have
The body condition measurement sensor includes four load sensors, and measures the center-of-gravity movement of the front, rear, left and right during the stepping motion of the subject ,
Correspondence between the Lyapunov index specified from the movement feature amount and the evaluation value of physical ability is described in the reference information,
The evaluation value specifying unit calculates a Lyapunov index indicating divergence from the motion feature of the subject using a calculation technique of a chaotic dynamical system, and based on the reference information, from the Lyapunov index of the subject, the body of the subject Identify an evaluation value that indicates your ability,
The motion feature is an attractor obtained by time-series chaos analysis, and from the reconstructed attractor, the Lyapunov exponent is estimated using the Sano-Sawada method, and when the Lyapunov exponent is estimated, a hypersphere The size, the number of neighboring points, and the development time coefficient are used as parameters .

  The present invention has the above-described configuration or processing, enables simultaneous measurement (recording) of brain function deterioration and shaping deterioration in order to efficiently determine the physical ability of a subject, and is a low-cost and simple system It has the effect that can be built.

It is a block diagram which shows an example of a structure of the physical capability determination apparatus in embodiment of this invention. It is a block diagram which shows an example of a structure of the physical capability determination part of the physical capability determination apparatus in embodiment of this invention. It is a flowchart which shows an example of the measurement process of the movement feature-value in embodiment of this invention. It is a flowchart which shows an example of the determination process of the athletic ability in embodiment of this invention. It is a flowchart which shows an example of the fluctuation | variation analysis process in embodiment of this invention. It is a flowchart which shows an example of the physical ability determination process in embodiment of this invention. It is a figure which shows a specific example of the fluctuation | variation analysis process in embodiment of this invention. In embodiment of this invention, it is a figure which shows an example of the body state measurement sensor for measuring the state of a test subject's body, and a measurement result. In embodiment of this invention, it is a figure which shows an example of the reference | standard information for evaluating a test subject's physical ability. It is a block diagram which shows an example of a structure of the physical capability determination apparatus in another embodiment of this invention. In another embodiment of this invention, it is a figure which shows an example of the reference | standard information for evaluating a test subject's physical ability. In another embodiment of this invention, it is a figure which shows another example of the reference | standard information for evaluating a test subject's physical ability. In the experiment which concerns on this invention, it is an image figure which shows an example which drawn the attractor by searching the parameter for reconstructing a favorable attractor by making time series data into the space vector P, and plotting the space vector P. In the experiment which concerns on this invention, when an embedding dimension number is 4 and embedding delay time is 30 (s), it is a figure which shows an example of the map (hexagonal shape) which projected the reconstructed attractor on a certain plane. . In the experiment according to the present invention, when the number of embedding dimensions is 4 and the embedding delay time is 30 (s), the reconstructed attractor is projected onto a certain plane based on the centroid position data of 6 subjects. It is a figure which shows an example of the mapping (hexagon shape) which performed. In the experiment according to the present invention, it is an image diagram showing a state of calculating a trajectory discrete for only a motion that draws a trajectory within a hypersphere size range from d1 to d3, with respect to a time-series point trajectory in a mapping space at the time of attractor superposition. . In the experiment according to the present invention, the Lyapunov exponent calculated using the Sano-Sawada method with the hypersphere size of 0.1, the number of neighbors of 20, the development time factor of 30, and the age of the subject It is a figure which shows an example of the graph made into the axis | shaft. In the experiment according to the present invention, based on the gravity center position data of five subjects at each age, when the number of embedding dimensions is 4 and the embedding delay time is 8 (s), the reconstructed attractor is a certain plane. It is a figure which shows an example of the map (butterfly shape) projected on the top. In the experiment according to the present invention, the Lyapunov exponent calculated using the Sano-Sawada method with the hypersphere size of 0.08, the number of neighbors of 20, the development time coefficient of 30, and the age of the subject It is a figure which shows an example of the graph made into the axis | shaft. In the experiment which concerns on this invention, it is a figure which shows an example of the graph which made the vertical axis | shaft the scaling index | exponent calculated by DFA as a parameter | index of stability of the reconfigure | reconstructed attractor, and made the test subject's age the horizontal axis. In this invention, it is a figure which shows an example of the display screen at the time of displaying the evaluation value (brain functional age) obtained from a Lyapunov index, a DFA scaling index, an attractor shape, a Lyapunov index, etc. on one screen in real time.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

The present invention provides a physical ability determination device and a data processing method capable of evaluating a subject's current physical ability by determining the physical ability of the subject. The present invention is configured so as to be realized with a simple configuration, and enables the simultaneous measurement (recording) of brain function deterioration and shaping deterioration of a subject's physical ability by a simple method that does not burden the subject. The physical ability of the subject can be determined.

  First, the configuration of the physical ability determination device in the embodiment of the present invention will be described. FIG. 1A is a block diagram illustrating an example of a configuration of a physical ability determination device according to an embodiment of the present invention, and FIG. 1B illustrates an example of a configuration of a physical ability determination unit of the physical ability determination device according to an embodiment of the present invention. A block showing. The physical ability determination device 100 illustrated in FIG. 1A includes an exercise feature amount observation unit 110, a physical ability determination unit 120, and a reference information storage unit 130. In FIG. 1A and FIG. 1B, each function is illustrated by a block, but each of these functions can be realized by hardware and / or (a program that can be executed by a computer).

  The movement feature quantity observation unit 110 has a function of measuring the movement feature quantity of the subject from information related to the physical state of the subject (body movement of the subject). For example, the motion feature amount observation unit 110 can calculate the motion feature amount of the subject based on sensing information detected from the body state measurement sensor 191 that detects the state of the subject's body. In addition, when the dynamic feature technique is used, the motion feature amount observation unit 110 can also convert the motion feature amount of the subject into an attractor (for example, perform the processing of FIG. 2 described later).

  In addition, as the body state measurement sensor 191, any sensor that measures muscle motion in the body of the subject as a numerical value can be used. For example, it is possible to use a method that detects the center of gravity movement (gravity sway) when the subject is standing upright using a load / acceleration sensor, or a method that detects the body movement or pace when the subject performs a walking motion. It is. In addition, by detecting the trajectory of body feature points from motion capture using a video camera, etc., it detects the center of gravity movement (center of gravity sway) when the subject is standing upright, and the body movement and pace when the subject performs a walking motion It is possible to use a method for detecting. FIG. 6 is a diagram illustrating an example of sensing information detected along with the walking motion of the subject in the embodiment of the present invention. As shown in FIG. 6 (A), the subject wears a wireless biosensor 600 capable of measuring muscle action potential, triaxial acceleration change, and temperature change (in FIG. 6 (A), Wearing the head, waist, and both knees), the subject's movement accompanying the walking motion is measured by having the subject perform the walking motion. By this measurement, for example, as shown in FIG. 6B, it is possible to measure changes in the positions of the head, waist, and both knees of the subject at the time of walking, the center of gravity movement of the subject, the pace of the subject, and the like. . In addition, here, the case where the subject is caused to walk is described as an example. For example, as shown in FIG. It is also possible to measure the vertical center of gravity fluctuation) with the load sensor 700.

  Further, the physical ability determination unit 120 has a function of determining the physical ability of the subject based on the subject's motion feature amount measured by the motion feature amount observation unit 110. The physical ability determination unit 120 compares, for example, a numerical value (for example, a Lyapunov exponent described later) obtained from the subject's movement feature amount with reference information stored in the reference information storage unit 130, thereby exercising the subject's movement. It is possible to specify at which level the feature amount is located. The determination result by the physical ability determination unit 120 can be output as an evaluation value from an output device such as a monitor / speaker 192, for example.

  The physical ability determination unit 120 includes a brain function determination unit 121 and a shaping determination unit 122 as illustrated in FIG. 1B. The brain function determination unit 121 has a function of executing the operation described in the embodiment of the present invention (the operation of evaluating physical ability from the movement feature amount), while the shaping determination unit 122 is a skeleton deterioration・ It has a function to observe muscle strength deterioration from the perspective of orthopedics. For example, regarding skeletal deterioration, it is possible to observe spinal distortion by observing the deterioration of the stepping motion of the right and left feet.

  The reference information storage unit 130 has a function of storing in advance reference information that is referred to when the physical ability determination unit 120 makes a determination. As the reference information, for example, it is possible to use a correspondence table that defines an evaluation value (for example, brain function age described later) corresponding to a numerical value (for example, a later-described Lyapunov exponent) obtained from the subject's movement feature value. is there. Further, for example, each pattern of an attractor group (attractor matrix) may be held and an evaluation value corresponding to each pattern may be defined.

  Next, processing according to the present invention will be described based on the configuration of the physical ability determination device 100 illustrated in FIGS. 1A and 1B.

  FIG. 2 is a flowchart showing an example of the movement feature amount measurement process in the embodiment of the present invention. 2 is a process executed by the motion feature quantity observation unit 110 shown in FIG. 1A in millisecond order (millisecond unit).

  In FIG. 2, the movement feature quantity observation unit 110 acquires a measurement value of the center of gravity movement (step S201), acquires a measurement value of the pace (step S202), and a measurement value of a change in three-axis acceleration (acceleration in a three-dimensional space). Is obtained) (step S203). Each of the sensing information in steps S201 to S203 is supplied from the body state measurement sensor 191 to the motion feature amount observation unit 110 in units of milliseconds, for example. In addition, each process of step S201-S203 is a parallel process. Further, in the embodiment of the present invention, a case where the center of gravity movement, the pace, and the triaxial acceleration change are measured will be described, but any one or a plurality (or all) of these measured values may be used. In addition, other sensing information may be used.

  Here, the movement feature amount observation unit 110 converts the measurement value supplied from the body state measurement sensor 191 into an attractor. For example, the motion feature quantity observation unit 110 converts the measured value of the center of gravity motion into an attractor (step S204), converts the measured value of the gait into an attractor (step S205), and converts the measured value of the triaxial acceleration change into an attractor (step S206). .

  As described above, the motion feature quantity observation unit 110 can measure the motion feature quantity of the subject and make it an attractor in the millisecond order.

  FIG. 3A is a flowchart illustrating an example of physical ability determination processing according to the embodiment of the present invention. Note that the physical ability determination process illustrated in FIG. 3A is performed by the physical ability determination unit 120 illustrated in FIGS. 1A and 1B in an order of several seconds to one minute (a few seconds to one minute unit). It is.

  In FIG. 3A, the physical ability determination unit 120 obtains the motion feature amount of the latest subject from the motion feature amount observation unit 110 (step S310). In addition, the test subject's exercise | movement feature-value acquired in step S310 is a result (what each exercise | movement feature-value was attracted | subtracted) calculated by step S204-S206 illustrated in FIG. Next, the physical ability determination unit 120 performs a fluctuation analysis process on the acquired movement feature amount of the subject (step S330). In the fluctuation analysis process in step S330, the physical ability determination unit 120 performs a process of calculating a Lyapunov exponent from each exercise feature amount, for example. Details of the fluctuation analysis processing in step S330 will be described later with reference to FIG. 3B. Then, the physical ability determination unit 120 accumulates (logs) the transition of the processing result of the fluctuation analysis process (specifically, the calculated Lyapunov exponent) in step S330 as a log (step S350).

  FIG. 3B is a flowchart illustrating an example of a fluctuation analysis process according to the embodiment of the present invention. Note that the fluctuation analysis process illustrated in FIG. 3B explains the details of the fluctuation analysis process in step S330 of FIG. 3A as described above.

  In FIG. 3B, the physical ability determination unit 120 obtains the attractor group (attractor matrix) of the center of gravity movement (step S3311), calculates the Lyapunov index of the center of gravity movement (step S3312), and calculates the Lyapunov index from the past attractor group. The stability tendency is analyzed (step S3313).

  Further, the physical ability determination unit 120 acquires a step attractor group (attractor matrix) (step S3321), calculates a step Lyapunov index (step S3322), and analyzes the stability tendency of the Lyapunov index from the past attractor group. (Step S3323).

  Further, the physical ability determination unit 120 acquires an attractor group (attractor matrix) of the measurement values of the triaxial acceleration change (step S3331), calculates a Lyapunov exponent of the measurement values of the triaxial acceleration change (step S3332), The stability tendency of the Lyapunov exponent is analyzed from the past attractor group (step S3333).

  Note that the specific calculation of the fluctuation analysis process in the embodiment of the present invention can be executed as shown in FIG. 5, for example. FIG. 5 is a diagram illustrating an example of a specific calculation method of the fluctuation analysis process according to the embodiment of the present invention. Note that the calculation method of the fluctuation analysis process is not limited to that shown in FIG. Various studies have been conducted on analysis methods of complex systems, and any analysis method established now and in the future can be applied to the present invention (for example, complex system analysis tools that have come here, http://www.ieice.org/cs/csbn/program/papers/04_1_miao.pdf).

  In the embodiment of the present invention, for example, each measurement value such as the center of gravity movement, the gait, the triaxial acceleration change is acquired (steps S201 to S203 in FIG. 2). Such measured values (sensing data) are acquired as a curve that varies with respect to the time axis, for example, as shown in FIG. 5A, and are handled as motion feature quantities that represent the characteristics of the body condition of the subject. Is possible.

  Here, for example, by acquiring the attractor (trajectory) of the local trend of the motion feature amount (local trend) in the minute window while shifting the range of the predetermined minute window (time width τ), the motion feature amount The attractor group (attractor matrix) is acquired (S3311, S3321, S3331, FIG. 5B in FIG. 3B), and the Lyapunov exponent can be calculated from the thus obtained attractor group (attractor matrix). (S3312, S3322 and S3332 in FIG. 3B, FIG. 5C). At this time, only one Lyapunov exponent is calculated, but by shifting a window (metawindow) including a plurality of predetermined minute windows (time width τ), time series data (temporal) of the Lyapunov exponent is calculated. A plurality of different Lyapunov exponents).

  The physical ability determination unit 120 may output a Lyapunov exponent (or time series data of the Lyapunov exponent) as described above as an exercise feature used for the determination, and the attractor group (attractor matrix) itself. It may be output. For example, when using a correspondence table that defines an evaluation value corresponding to a numerical value (for example, a Lyapunov index) obtained from a subject's motion feature amount as reference information, the Lyapunov index is output as a motion feature amount used for determination, When each pattern of the attractor group (attractor matrix) corresponding to the evaluation value is held, the attractor group (attractor matrix) itself may be output.

  FIG. 4 is a flowchart showing an example of the physical ability determination process in the embodiment of the present invention. In the physical ability determination process illustrated in FIG. 4, the analysis result of the fluctuation analysis process (the processing result of step S330 in FIG. 3A) is acquired by the physical ability determination unit 120 illustrated in FIGS. 1A and 1B. It may be executed every time it is performed, or may be executed after measuring the motion feature quantity of the subject on the order of several tens of seconds to several minutes.

  In FIG. 4, the physical ability determination unit 120 acquires the reference information stored in the reference information storage unit 130 (step S <b> 401), and exercise feature amounts (for example, for use in the determination calculated by the physical ability determination unit 120). (Represented by the Lyapunov exponent based on the motion feature) is acquired (step S402). The reference information acquired in step S401 is a correspondence table of Lyapunov exponents and evaluation values stored in the reference information storage unit 130, each pattern of an attractor group (attractor matrix) corresponding to the evaluation values, and the like. In addition, the motion feature value acquired in step S402 is the analysis result of the fluctuation analysis process in step S330 in FIG. 3A (the process result in step S330 in FIG. 3A), and specifically acquired from each motion feature value. This is the Lyapunov exponent calculated from the attractor group and each motion feature.

  The physical ability determination unit 120 specifies the evaluation value of the physical ability of the subject based on the exercise feature amount and the reference information used for the determination calculated by the physical ability determination unit 120 (step S403). Specifically, the physical ability determination unit 120 determines which level the exercise feature amount used for the determination calculated by the physical ability determination unit 120 corresponds to in the reference information. In this determination, when using a correspondence table that defines an evaluation value (for example, brain function age described later) corresponding to a numerical value (for example, Lyapunov exponent) obtained from the subject's motion feature value, The evaluation value corresponding to the obtained Lyapunov exponent is specified, and when each pattern of the attractor group (attractor matrix) corresponding to the evaluation value is used, the pattern of the attractor group obtained from the motion feature of the subject is analyzed. Thus, it is specified which evaluation value corresponds to the pattern of the attractor group (attractor matrix). When examining the similarity of patterns, the spread width or curved surface shape of a specific local region of the attractor may be compared with the shape of the attractor corresponding to the reference evaluation value.

  FIG. 7 illustrates a correspondence table (an example of reference information) that defines an evaluation value (in this case, described as brain function age) corresponding to the Lyapunov exponent. In general, the larger the Lyapunov exponent value, the more active the brain function that controls the physical ability, and the smaller the Lyapunov exponent value, the less the brain function that controls the physical ability. Although there is a tendency, as described later (explained as another embodiment), it is possible to output a more accurate evaluation value by appropriately selecting a correspondence table to be used according to the age (age) of the subject. Is possible.

  The physical ability determination unit 120 specifies an evaluation value (brain function age) corresponding to the Lyapunov exponent using, for example, the correspondence table illustrated in FIG. 7, and the evaluation value (brain function age) is displayed on the monitor / speaker. 192 or the like, or can be transmitted to a delivery destination (such as a server) via a communication unit (not shown in FIG. 1A). In addition, by registering the subject (user registration), the evaluation value of the subject is sequentially accumulated, the state of the subject's brain function is traced, and it is used to grasp the progress of orthopedic surgery (rehabilitation, etc.) by the subject. It is also possible to do.

  Next, another embodiment of the present invention will be described. According to this embodiment, the physical ability of the subject can be evaluated in more detail by reflecting the age (age) of the subject estimated using DFA (Detrended Fluctuation Analysis) analysis. Become. As described above, the larger the Lyapunov exponent value, the more active the brain function that controls the physical ability, and the smaller the Lyapunov exponent value, the less the brain function that controls the physical ability. However, the evaluation value corresponding to the value of the Lyapunov index differs depending on the age (age) of the subject. Accordingly, it is useful to select reference information according to the age (age) of the subject.

  FIG. 8 is a block diagram showing an example of the configuration of the physical ability determination device according to another embodiment of the present invention. The physical ability determination device 100 illustrated in FIG. 7 includes an exercise feature amount observation unit 110, a physical ability determination unit 120, a reference information storage unit 130, and a DFA analysis unit 140. In FIG. 8, each function is illustrated as a block, but each function can be realized by hardware and / or (a program that can be executed by a computer).

  The physical ability determination device 100 illustrated in FIG. 8 further includes a DFA analysis unit 140 as compared with the configuration illustrated in FIG. 1A. In addition, a function is added to the physical ability determination unit 120, and the reference information storage unit 130 stores reference information that enables more detailed evaluation. Hereinafter, functions and operations characteristic of another embodiment of the present invention will be mainly described.

The DFA analysis unit 140 performs DFA analysis based on information (for example, sensing information from the body state measurement sensor 191) related to the physical state of the subject (body motion of the subject), and the DFA analysis result (scaling index α ) Is output. Note that DFA analysis is a technique for analyzing long-term correlation characteristics of non-stationary time-series data, and since it is a widely known technique, detailed description thereof is omitted here. DFA analysis divides new time-series data created by integrating input data into boxes of length n, calculates an approximate straight line by the method of least squares in each box, obtains a local trend, and calculates the trend from the signal. A value (F (n)) obtained by averaging the dispersions after the removal is calculated. F (n) determined by this DFA analysis can be represented by F (n) ∝n ^α and is characterized by a scaling index α.

  The inventor of the present invention can use the scaling index obtained by the DFA analysis to estimate the age (age) of the subject and determine the physical ability of the subject based on the estimated age (age) of the subject. I found it useful. Specifically, for example, it is possible to divide the age by α ≧ 0.8 and α <0.8 with α = 0.8 as a threshold value. Hereinafter, a case where α ≧ 0.8 is calculated is referred to as a young group, and a case where α <0.8 is calculated as an old group. Here, as an example, the age is divided into two by one threshold (for example, α = 0.8) (classified into two categories of young and old), but the value set as the threshold or The number of threshold values and the number of categories to be classified can be set as appropriate.

  The DFA analysis result by the DFA analysis unit 140 is supplied to the physical ability determination unit 120. The physical ability determination unit 120 has a function of specifying an evaluation value indicating the physical ability of the subject based on the subject's motion feature amount measured by the motion feature amount observation unit 110. For example, the subject's motion feature By comparing the numerical value obtained from the quantity with the reference information stored in the reference information storage unit 130, it is possible to evaluate at which level the motion feature quantity of the subject is located.

  In the present embodiment, physical ability determination unit 120 further has a function of selecting the reference information stored in reference information storage unit 130 based on the DFA analysis result of DFA analysis unit 140. The physical ability determination unit 120 corresponds to the value of the scaling index α from among a plurality of pieces of reference information stored in the reference information storage unit 130 based on the value of the scaling index α supplied from the DFA analysis unit 140, for example. The reference information to be selected is selected, and the evaluation value of the physical ability of the subject is specified by comparing the exercise feature quantity of the subject with the selected reference information.

  For example, the reference information storage unit 130 stores two pieces of reference information for young people and old people. For example, when the Lyapunov index is used for the evaluation of the physical ability of the subject, the correspondence table for young people as shown in FIG. 9 and the correspondence table for old people as shown in FIG. And are stored. For example, when the similarity of the pattern of the attractor group is determined for the evaluation of the physical ability of the subject, the attractor group for the young group and the attractor group for the elderly group are stored.

  As shown in the above example, when the scaling index α supplied from the DFA analysis unit 140 is 0.8 or more, the physical ability determination unit 120 generates a correspondence table for young people (FIG. 9). The evaluation value of the subject's physical ability is specified by selecting as reference information and comparing with the selected reference information. Similarly, when the scaling index α supplied from the DFA analysis unit 140 is less than 0.8, the physical ability determination unit 120 selects and selects the correspondence table for the elderly (FIG. 10) as reference information. By comparing with the reference information, the evaluation value of the physical ability of the subject is specified.

  As described above, according to the present embodiment, the age (age) of the subject is estimated based on the DFA analysis result by the DFA analysis unit 140, and the reference information corresponding to the estimated age (age) of the subject is selected. The physical ability of the subject can be evaluated based on the selected reference information.

  Next, as an example related to the above-described embodiment of the present invention, details of an experiment performed using a prototype system will be described.

  In the experiment according to the present invention, a load sensor (for example, a strain gauge type force sensor) having four points as shown in FIG. 6C is provided as the body condition measurement sensor 191 shown in FIG. The center of gravity of the subject was measured using a plate controller. The center of gravity shake meter generally used in fields such as rehabilitation medicine and sports medicine needs to extract a minute balance shake in order to judge normality / abnormality of center of gravity shake. -It has a highly sensitive load sensor and has a hardware mechanism for removing noise such as vibration propagating through the floor, making it a very expensive device. On the other hand, in the experiment according to the present invention, an extremely simple and inexpensive plate-like board controller (hereinafter referred to as a consumer board controller) such as a widely used game controller is used.

  In the experiment according to the present invention, the subject was stepped on the consumer board controller, and the center-of-gravity movement vector on the floor plane was calculated from the center-of-gravity fluctuation data obtained from the consumer board controller. Next, with respect to the obtained center-of-gravity movement vector, we attempted to reconstruct the attractor using time-series data of the front, rear, left and right components of the consumer board controller. After that, the Lyapunov exponent indicating the degree of instability of the autocorrelation of the reconstructed attractor was estimated, and this value was used as a quantitative indicator of biological fluctuations indicating brain activity. Finally, in order to quantitatively confirm the stability of the motion moment in the stepping motion, the scaling index was calculated using a spectrum analysis method (DFA) that is effective for the analysis of unsteady signals. Note that the scaling index calculated by the DFA can be handled as an index indicating the stability of walking caused by the motion moment, and the higher the stability, the lower the scaling index.

  The stepping position and the load during the stepping operation change in a complicated manner with time, and the data can be measured as non-linear time series data. However, it is not easy to analyze the complex behavior in detail because of the influence of noise on the measurement environment and ecological effects such as the age and physical condition of the subject. The time-series chaos analysis method is an analysis method that derives deterministic properties as a simple nonlinear signal model from such a complex and irregular time-series signal in the real world. This is a technique to visualize the trajectory of dynamic motion with autocorrelation (trend) from the waveform signal (attractor reconstruction) and quantify the tendency characteristics.

  The data obtained from the consumer board controller used in the experiment according to the present invention includes various noises that propagate from the ground plane during measurement, and has low resolution. It is considered that the time series chaos analysis method for analyzing the trend characteristic is also effective for low-resolution center-of-gravity fluctuation data containing a lot of noise, but prior to the experiment according to the present invention, First, a preliminary experiment was conducted to investigate whether or not attractor reconstruction can be performed from the center-of-gravity fluctuation data containing a lot of noise. Specifically, using the Placence's embedding theorem for the stepping motion center of gravity data measured at a sampling rate of 30 (Hz) using a consumer board controller, as shown in FIG. Using time-series data as a space vector P (embedded vector P), parameters for reconstructing a good attractor (number of embedded dimensions n, embedded delay time τ) were searched. In the image diagram of FIG. 11, the trajectory when the three-dimensional delay space is set as the XYZ observation point and the plot is moved in the order of P1, P2, and P3 is the attractor.

  As a result, when the number of embedding dimensions is 4 and the embedding delay time is 30 (s), a map obtained by projecting the reconstructed attractor onto a certain plane has a hexagonal shape as shown in FIG. It was. From this result, it was confirmed that the attractor can be reconfigured for data containing a lot of various noises in a wide band. The found attractor visually represents the tendency characteristic of the motion moment during the stepping motion.

  In the experiment according to the present invention, 13 subjects in their 30s to 70s performed a stepping motion for 3 minutes on the consumer board controller, and the left and right direction of the subject in the stepping position is the X axis, and the front and back direction is the Y axis. The center-of-gravity position data was recorded, and the center-of-gravity movement vector was calculated from the center-of-gravity position data. Then, for each of the X-axis component and Y-axis component of each subject's center-of-gravity movement vector, reconstructing the attractor using the parameters (number of embedding dimensions 4, embedding delay time 30 (s)) obtained in the above preliminary experiment. Went. The attractor can be reconfigured for the X-axis component. FIG. 13 shows a typical shape of the obtained attractor. Note that the attractor shown in the lower right of FIG. 13 is reconstructed from the center-of-gravity movement vector associated with a person with a cerebral infarction half body, and is significantly distorted unlike the shape of the other person. The other attractors had a hexagonal shape that was distorted, but the trend characteristics could be extracted.

  Further, as schematically shown in FIG. 14, the Lyapunov exponent is used as an amount indicating the degree to which the approaching orbit in the dynamic system leaves. The image diagram of FIG. 14 shows how the trajectory discrete is calculated only for the motion that draws the trajectory of the time series point trajectory in the mapping space at the time of attractor superposition within the hypersphere size range from d1 to d3. .

  In the experiment according to the present invention, the degree of orbital instability of the tendency characteristic of time series data expressed as an attractor is indicated by the Lyapunov exponent. Here, the Lyapunov exponent was adjusted using the Sano-Sawada method considering real-time characteristics for the calculation of the Lyapunov exponent, using the hypersphere size, the number of neighboring points, and the development time coefficient as parameters. Specifically, the value of the Lyapunov exponent was calculated using the Sano-Sawada method with the hypersphere size of 0.1, the number of neighboring points of 20, the development time coefficient of 30. FIG. 15 shows a graph in which the Lyapunov exponent value estimated as a result is the vertical axis and the age of the subject is the horizontal axis. From this result, there is no correlation between the Lyapunov exponent and age, and there is a tendency to diverge with aging. Therefore, it is considered that this attractor shape (hexagonal shape) does not show a brain activity tendency of biological fluctuations. However, on the other hand, as described above, the attractor of a person with a cerebral infarction half-body is remarkably distorted. Therefore, this attractor shape is considered to indicate a stable tendency of the motion moment. Therefore, the analysis using the parameters used in the above-mentioned motion feature value analysis is used to visually grasp the orthopedic abnormality related to the stepping motion visually by the attractor shape or quantitatively by the Lyapunov exponent. It is considered effective.

  In addition, based on the result of motion feature analysis, another parameter search for reconstructing an attractor effective as motion feature was performed using the Turkens embedding theorem. As a result, when the number of embedding dimensions was 4 and the embedding delay time was 8 (s), butterfly-shaped attractors were confirmed uniformly from the data of all subjects for the X-axis component of the center-of-gravity movement vector. FIG. 16 is a diagram showing typical attractor shapes for each age group: (a) 30s, (b) 40s, (c) 50s, (d) healthy people over 60s, (e) 60s. An example of the attractor shape obtained from the center-of-gravity movement vector of the non-healthy subject is described above. As shown in FIG. 16, in this butterfly-shaped attractor shape, it was confirmed that the strain breaking tends to increase with aging. Moreover, remarkable distortion could be observed in the attractor ((e) of FIG. 16) extracted from the stepping motion of the elderly with a walking disorder.

  Next, from the reconstructed attractor of the X-axis component, the Lyapunov exponent estimated using the Sano-Sawada method with the hypersphere size of 0.08, the number of neighbors of 20, the development time coefficient of 30, and the vertical axis FIG. 17 shows a graph with the age of Since the test population was small, a strong correlation could not be confirmed from this result, but the tendency of the Lyapunov index to decrease with age was confirmed. FIG. 18 shows a graph in which the scaling index calculated by DFA is used as an index of stability of the reconstructed attractor, and the age of the subject is plotted on the horizontal axis. According to this result, it was confirmed that the scaling index tends to increase with aging.

  Thus, when the parameters used in the above-described motion feature amount analysis (embedding dimension number 4, embedding delay time 8 (s)) are used, the attractor mapping of the X-axis component of the centroid movement vector is Since the subject became a uniform butterfly shape, it is considered that useful tendency characteristics regarding the stepping motion could be extracted. In addition, it was confirmed that the Lyapunov exponent estimated from the reconstructed attractor had a decreasing tendency with aging. Therefore, the Lyapunov exponent extracted from each subject's center-of-gravity movement vector is effective as a quantitative index indicating the brain activity tendency, and, as described above, for example, by mapping as the brain function age indicating the brain activity tendency, It is considered that the feature amount related to the brain function of the subject can be specified as the evaluation value. In addition, since there is a correlation between brain activity and cognitive ability, the Lyapunov index determined by this method may be a quantitative index for diagnosing dementia. Furthermore, the scaling index calculated by DFA showed a tendency to increase with aging, and it was found that the greater the age, the greater the movement of the orthopedic movement moment. In addition, this method has been used as a quantitative index for orthopedic examinations since it was confirmed that the attractors of subjects with gait disturbance were visually distorted compared to other subjects. It is thought that it can be used.

  As described above, even in the case of data including various noises (for example, center-of-gravity fluctuation data) obtained from a very simple sensor (for example, a consumer board controller) by an experiment according to the present invention, chaos dynamics It was confirmed that the instability (brain activity) of biological fluctuation can be grasped from the Lyapunov exponent estimated based on the attractor by performing analysis using the calculation technique of the system.

  Further, it has been found by the experiment according to the present invention that there is a possibility that the stability of the motion moment can be grasped from the reconstructed attractor (the degree of distortion of the attractor). In addition, as a method of time series chaos analysis of the present invention, in particular, search for attractor reconstruction parameters (number of embedding dimensions, embedding delay time) using the Turkens embedding theorem, calculation of Lyapunov exponents using the Sano-Sawada method (Adjusted with the hypersphere size, the number of neighbors, and the development time score as parameters) was confirmed to be applicable. Moreover, it was confirmed from the experiment which concerns on this invention that the scaling index | exponent calculated by DFA is useful for estimation of a test subject's age (age).

  In the experiment of the present invention, data from a three-dimensional sensor (for example, a three-dimensional motion sensor for games) is collected simultaneously with data collection by a consumer board controller to estimate a subject's skeletal motion (body posture). Attempts have also been made to display the results of estimation of the skeletal motion and the information obtained from the sway meter (attractor, DFA, evaluation value (brain function age), etc.) on the tablet computer.

  Further, when the attractor is reconstructed from the center-of-gravity fluctuation data obtained from the consumer board controller or the propulsion fluctuation data recorded once the measurement is completed, as shown in FIG. 19, the Lyapunov exponent is used. , The evaluation value (brain function age) obtained from DFA scaling index, attractor shape and Lyapunov index, etc. is displayed on a single screen in real time, improving the visibility of the user and realizing a highly convenient display screen May be. Further, the skeletal motion of the subject and other biological information (for example, pulse information) based on the data collected by the three-dimensional sensor may be displayed in synchronization.

  The present invention has the effect that the physical ability of the subject can be determined efficiently, the brain function deterioration and the shaping deterioration can be simultaneously measured, and a simple system can be constructed at a low cost. The present invention can be applied to a technique for evaluating the physical ability of a subject and a medical technique for performing functional recovery based on the evaluation result of the physical ability of the subject.

DESCRIPTION OF SYMBOLS 100 Physical ability determination apparatus 110 Motion feature-value observation part 120 Physical ability determination part 121 Brain function determination part 122 Shaping determination part 130 Reference | standard information storage part 140 DFA analysis part 191 Body state measurement sensor 192 Monitor / speaker 700 Load sensor

Claims (8)

A physical ability determination device for determining a physical condition of a subject,
Based on sensing information supplied from a body state measurement sensor that measures the body motion of the subject, a motion feature amount observation unit that calculates a motion feature amount indicating a feature of the body motion of the subject;
A reference information storage unit that stores reference information for specifying the correspondence between the exercise feature and the evaluation value of physical ability;
Based on the reference information stored in the reference information storage unit, physical ability determination that identifies an evaluation value indicating the physical ability of the subject from the exercise feature amount of the subject calculated by the exercise feature amount observation unit Part
Have
The body condition measurement sensor includes four load sensors, and measures the center-of-gravity movement of the front, rear, left and right during the stepping motion of the subject ,
Correspondence between the Lyapunov index specified from the movement feature amount and the evaluation value of physical ability is described in the reference information,
The physical ability determination unit calculates a Lyapunov index indicating divergence from the motion feature of the subject using a calculation technology of a chaotic dynamical system, and based on the reference information, the body of the subject is calculated from the Lyapunov index of the subject. Identify an evaluation value that indicates your ability,
The motion feature is an attractor obtained by time-series chaos analysis, and from the reconstructed attractor, the Lyapunov exponent is estimated using the Sano-Sawada method, and when the Lyapunov exponent is estimated, a hypersphere Physical ability determination device with size, number of nearby points, and development time coefficient as parameters . A plurality of reference information is stored in the reference information storage unit,
Based on the sensing information, a trend removal fluctuation analysis unit that calculates a scaling index by analysis using a trend removal fluctuation analysis method,
The physical ability determination unit selects reference information corresponding to the scaling index calculated by the trend removal fluctuation analysis unit, and specifies an evaluation value indicating the physical ability of the subject based on the selected reference information The physical ability determination device according to claim 1, wherein the physical ability determination device is configured to perform. The physical ability determination device according to claim 1 , wherein a parameter for reconstructing the attractor is searched using a Turkens embedding theorem. The attractor obtained by the time-series chaos analysis used as the motion feature value, the scaling index obtained by the analysis using the trend removal variation analysis method, and the physical ability of the subject specified from the motion feature value of the subject are shown. the evaluation value and the, physical capability determination apparatus according to any one of claims 1 to 3 to be displayed in one display screen in real time. A data processing method in an apparatus for outputting an evaluation value indicating a physical ability of a subject,
The apparatus includes a motion feature amount observation unit, a reference information storage unit, and an evaluation value specifying unit,
The motion feature amount observing unit calculating a motion feature amount indicating a feature of the subject's body movement based on sensing information supplied from a body state measurement sensor that measures the subject's body motion;
The evaluation value specifying unit is calculated by the exercise feature amount observing unit based on reference information for specifying the correspondence between the exercise feature amount and the evaluation value of physical ability stored in the reference information storage unit. A step of identifying an evaluation value indicating the physical ability of the subject from the movement feature amount of the subject,
Have
The body condition measurement sensor includes four load sensors, and measures the center-of-gravity movement of the front, rear, left and right during the stepping motion of the subject ,
Correspondence between the Lyapunov index specified from the movement feature amount and the evaluation value of physical ability is described in the reference information,
The evaluation value specifying unit calculates a Lyapunov index indicating divergence from the motion feature of the subject using a calculation technique of a chaotic dynamical system, and based on the reference information, from the Lyapunov index of the subject, the body of the subject Identify an evaluation value that indicates your ability,
The motion feature is an attractor obtained by time-series chaos analysis, and from the reconstructed attractor, the Lyapunov exponent is estimated using the Sano-Sawada method, and when the Lyapunov exponent is estimated, a hypersphere A data processing method that uses size, number of neighbors, and development time coefficient as parameters . The apparatus has a trend removal fluctuation analysis unit that calculates a scaling index by analysis using a trend removal fluctuation analysis method based on the sensing information,
The evaluation value specifying unit selects reference information corresponding to the scaling index calculated by the trend removal variation analysis unit from a plurality of reference information, and based on the selected reference information, the body of the subject The data processing method according to claim 5, wherein an evaluation value indicating capability is specified. The data processing method according to claim 5 , wherein a parameter for reconstructing the attractor is searched using a Turkens embedding theorem. The attractor obtained by the time-series chaos analysis used as the motion feature value, the scaling index obtained by the analysis using the trend removal variation analysis method, and the physical ability of the subject specified from the motion feature value of the subject are shown. The data processing method according to any one of claims 5 to 7 , wherein the evaluation value is displayed in real time in one display screen.

Images