JP2018038752A - Walking analysis method and walking analyzer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique capable of easily acquiring a moment generated in the knee joints during walking.SOLUTION: A walking analysis method includes: an acquisition step (S11) of acquiring acceleration data of the lumbar part of a subject during walking; and an estimation step (S12) of estimating a moment generated in the knee joints of the subject during the walking using statistical information of walking sample data preliminarily measured from a population and the acceleration data acquired in the acquisition step (S11).SELECTED DRAWING: Figure 1

Description

  The present invention relates to a gait analysis technique, and more particularly to an analysis technique using acceleration during walking.

  Various services are provided to promote walking activities in order to maintain the muscle strength of the elderly and prolong the healthy life expectancy. However, if inappropriate walking activities are continued, there is a possibility that the knees, ankles, hip joints, etc. may be damaged. Therefore, various methods for analyzing the motion state and walking state of the subject have been proposed.

For example, Patent Document 1 below discloses a method for evaluating a subject's motion characteristics by comparing a motion characteristic model with a reference model prepared in advance. In this method, in order to determine the motion characteristic model, the pressure on the sole, the transition of the positional relationship between the sole and the torso, and the height of the center of gravity are used, and the vertical plane perpendicular to the subject's longitudinal direction is used. It is disclosed that the position of the center of gravity is estimated to obtain the transition, and the moment acting on the center of gravity in the vertical plane is obtained.
In Patent Document 2 below, a joint angle of a pedestrian's knee joint is obtained based on acceleration or angular velocity measured by a sensor attached so as to sandwich the pedestrian's knee joint, and the walking state of the pedestrian is calculated based on the joint angle. Is disclosed.
Non-patent documents 1 and 2 will be described later.

JP 2012-161402 A JP 2014-208257 A

"Analysis of gait by joint moment", Chapter 1-Chapter 2, Clinical Gait Analysis Study Group, Bio-Dental Publishing Co., Ltd., July 1997 "AIST walking database 2015", [online], National Institute of Advanced Industrial Science and Technology (AIST), [March 17, 2016 search], Internet <URL: https: //www.dh.aist.go .jp / database / gait2015 / index. html>

One of the older cases with high morbidity is knee osteoarthritis. Osteoarthritis of the knee is considered to be one of the major causes of shortening the healthy life span because it causes pain in the knee and inhibits walking. On the other hand, it is known that the varus or valgus moment applied to the knee joint during walking is involved in the onset of knee osteoarthritis.
However, in order to measure the moment of the knee joint, an advanced device or system using a motion capture technique or a floor reaction force meter is required.

  The present invention has been made in view of such a problem, and provides a technique capable of easily acquiring a moment generated in a knee joint during walking (hereinafter sometimes referred to as a knee joint moment).

  Each aspect of the present invention employs the following configurations in order to solve the above-described problems.

  The first aspect relates to a gait analysis method. The walking analysis method according to the first aspect includes an acquisition step of acquiring acceleration data in the waist of a walking subject, statistical information of walking sample data measured in advance from a population, and acceleration data acquired in the acquisition step And estimating the moment generated at the knee joint of the subject during walking.

  The second aspect relates to a gait analyzer. The walking analysis apparatus according to the second aspect includes an acquisition means for acquiring acceleration data at the waist of a subject who is walking, a storage means for storing statistical information of walking sample data measured in advance from a population, and the statistical information. And calculating means for calculating a moment value generated at the knee joint of the subject during walking using the acquired acceleration data.

  Another aspect of the present invention is a program that causes a computer to execute the method according to the first aspect, or a computer-readable storage medium that records such a program. This recording medium includes a non-transitory tangible medium.

  According to each of the above aspects, it is possible to provide a technique capable of easily acquiring a moment generated in the knee joint during walking.

It is a flowchart which shows the walk analysis method in 1st embodiment. It is a figure which shows a knee joint moment conceptually. It is a figure which shows notionally one general walk cycle. It is a figure which shows notionally the hardware structural example of the gait analyzer in 1st embodiment. It is a figure which shows notionally the process structural example of the gait analyzer in 1st embodiment. It is a flowchart which shows the walk analysis method in 2nd embodiment. It is a figure which shows the mode of a measurement. It is a figure which shows notionally the acceleration component selected as an explanatory variable. It is a graph which shows the precision of the multiple regression type based on the measurement data of the 1st group. It is a graph which shows the precision of the multiple regression type based on the measurement data of the 2nd group.

  Embodiments of the present invention will be described below. In addition, each embodiment given below is an illustration, respectively, and this invention is not limited to the structure of each following embodiment.

[First embodiment]
[Walk analysis method]
First, the gait analysis method in the first embodiment (hereinafter sometimes abbreviated as the first method) will be described. FIG. 1 is a flowchart showing a gait analysis method in the first embodiment.
As shown in FIG. 1, the first method includes a step (S11) of acquiring acceleration data in the waist of a walking subject, and statistical information of walking sample data measured in advance from a population and a step (S11). And (S12) estimating the moment generated in the knee joint of the subject during walking using the acquired acceleration data.

Here, the “lumbar region” means the waist and its periphery. The waist means the part on the back side from the lower part of the spinal column to the periphery of the pelvis. Here, the “waist” includes, in addition to the so-called “waist”, the so-called “waist circumference” from the waist to the abdomen, and further includes the periphery of the “waist circumference”. It is preferable that the periphery of the “waistline” is a range including a portion that moves integrally with the waist.
The acceleration data acquired in the step (S11) is time series data of acceleration, and the acceleration is indicated by one or more direction components. In other words, the acceleration data is a set of acceleration values in one or more directions measured within an arbitrary time. The acceleration data may be an acceleration value in a certain direction at a certain temporary point.

The method for acquiring acceleration data in the step (S11) is not limited.
For example, an acceleration value group measured by an acceleration sensor mounted on the waist of the subject is acquired as acceleration data. In this case, a computer such as a smartphone or a personal computer (hereinafter referred to as a PC) can acquire the acceleration data by communicating with the acceleration sensor. The execution subject of the step (S11) may be a device incorporating the acceleration sensor.
The acceleration data may be calculated using a motion capture technique. Specifically, acceleration data can be calculated by a computer or a sensor measuring trajectory data of the position of the subject's waist using a motion capture technique and performing a differential operation on the trajectory data of the measured position. The position information of the waist may be obtained by detecting a marker attached to the waist of the subject, or may be detected from the subject image without using a marker such as a Kinect (registered trademark) sensor.

In the first method, in step (S12), the knee joint moment during walking of the subject is estimated using the acceleration data acquired in step (S11).
FIG. 2 is a diagram conceptually showing the knee joint moment. As shown in FIG. 2, the knee joint moment is a rotational force generated around an axis (axis extending in the front-rear direction of the human body) Ax perpendicular to the frontal plane in the knee, and includes a counterclockwise inner moment and clockwise rotation. One or both of the valgus moments. The valgus moment tends to dominate at the knee joint of the O-leg, the valgus moment tends to dominate at the knee joint of the X leg, and the inner or outer side of the knee joint is worn away by the varus or valgus moment. It is known that knee osteoarthritis develops.

The moment estimated in the step (S12) may be normalized by physique data such as body weight, or may not be normalized (Nm (Newton meter)).
Further, the moment estimated in the step (S12) may be either the right knee joint or the left knee joint, or both.
The knee joint moment changes sequentially even during one walking cycle.
Here, “one walking cycle” means a walking period from when one of the feet is landed to when it is again landed. However, the start point of one walking cycle may be any time, not necessarily when the foot has landed.
FIG. 3 is a diagram conceptually illustrating a general one walking cycle. One gait cycle is formed from the stance phase and the swing phase when focusing on one leg. The stance phase is a period in which one leg is landing, and the free leg period is a period in which the one leg is getting out of bed. The stance phase is further formed from a heel-contact phase (at the time of landing), a mid-stance phase, and a stepping phase.
For example, the moment of the right knee joint is different in the stance phase and the swing leg phase in the right leg, and in the stance phase, it is different in the heel contact phase, the mid stance phase, and the stepping phase. For this reason, the knee joint moment estimated in the step (S12) may be a predetermined moment or a moment at any time during walking, or may be a maximum or minimum moment during a predetermined period. It may be the average moment of the period. It is only necessary to estimate the knee joint moment according to the purpose.

In step (S12), in addition to the acceleration data acquired in step (S11), the knee joint moment is estimated using statistical information of walking sample data measured in advance from the population.
Here, “walking sample data” is data that can be measured from each walking person forming the population, and is data that can derive the knee joint moment of each walking person. The “walking sample data” preferably includes the waist acceleration of each person during walking or data from which the acceleration can be derived, and includes other data related to the population, such as physique data. May be.
Any known method may be used to obtain the knee joint moment. For example, it is known that the knee joint moment value can be approximately calculated only from the floor reaction force, and can be calculated by further using the joint position, joint acceleration (angular velocity), body parameters, and the like (see Non-Patent Document 1). Thus, for example, the walking sample data is the position data or acceleration data of the hips and knees of each person who is walking the population, and floor reaction force data.

  The “statistical information of walking sample data” used in the step (S12) is information obtained by statistically processing the walking sample data based on the causal relationship between the acceleration of the waist and the knee joint moment. The causal relationship between the acceleration of the lumbar region and the knee joint moment is newly found by the present inventors.For example, a regression equation obtained by regression analysis based on such a causal relationship is an example of the statistical information. As mentioned. In this example, the statistical information corresponds to a single regression equation or a multiple regression equation in which the acceleration value of the waist during walking is an explanatory variable and the knee joint moment value during walking is an objective variable. When the “statistical information of the walking sample data” is such a regression formula, the first method substitutes the acceleration data acquired in the step (S11) into the regression formula in the step (S12), thereby the subject. Estimate the knee joint moment.

  Further, the present inventors have found that among the acceleration data during walking, acceleration data at a plurality of local timings during one walking cycle shows a stronger causal relationship with the knee joint moment during walking. Therefore, in the step (S12), the knee joint moment is estimated using the acceleration data at each of a plurality of predetermined timings in one walking cycle obtained from the acceleration data acquired in the step (S11). Is preferred.

The knee joint moment during one walking cycle is greater in the stance phase of the leg with the knee joint than in the swing phase of the leg. Also in the lumbar region, the movements during walking are different between the right lumbar region and the left lumbar region, the right lumbar region is highly interlocked with the right leg, and the left waist region is highly interlocked with the left leg. That is, the acceleration of the right waist is more strongly associated with the moment of the right knee joint than the left knee joint, and the acceleration of the left waist is more strongly associated with the moment of the left knee joint than the right knee joint.
Therefore, in the step (S12), when acceleration data of the right hip is acquired in the step (S11), the acceleration data at each of a plurality of timings determined in advance from the right stance phase is used during walking. When the moment generated in the right knee joint of the subject is estimated and acceleration data of the left hip is acquired, the acceleration data at each of a plurality of predetermined timings from the left stance phase is used during walking. It is preferable to estimate the moment generated in the subject's left knee joint. As a result, the magnitude of the moment of the knee joint during one walking cycle can be estimated with high accuracy.

  Furthermore, the present inventors have found that acceleration data and knee joint moment show a stronger causal relationship when the above-described plural timings are individually determined for each of two or more directional components of acceleration. Thereby, in the step (S12), it is more preferable to estimate the knee joint moment by using the acceleration value in the direction at each predetermined timing for each of the two or more direction components. Thereby, the estimation accuracy of the knee joint moment can be further improved.

Each of the above-mentioned “local plural timings during one walking cycle” may be a certain temporary point or a period having an arbitrary time width. Each of the “multiple timings” may have different time widths.
When each of the plurality of timings has a predetermined time width individually, in step (S12), using the average value of acceleration values in the same direction within each time width of the plurality of timings, the knee joint moment Can be estimated.

Further, one walking cycle in the acceleration data acquired in the step (S11) can be specified based on the vertical direction component included in the acquired acceleration data. For example, the maximum value of acceleration in the vertical direction indicated by the acceleration data is detected, and the interval between the maximum values arranged in time series is specified as one walking cycle. Since the maximum value of acceleration in the vertical direction occurs at the time of landing on the target leg, that is, when the heel touches down, the maximum value that is aligned in time series from the time of landing of one leg to the next landing of that leg. Can be specified between values.
Therefore, the first method may further include a step of specifying one walking cycle in the acceleration data based on the vertical direction component indicated by the acceleration data acquired in step (S11). In this case, in the step (S12), the acceleration data at each of the plurality of timings can be specified and used from the acceleration data in one walking cycle specified in the step.
The identification of one walking cycle can also be realized by other methods. For example, by using a motion capture technique, a locus of a predetermined position of the right foot (or right leg) and the left foot (or left leg) is detected, and one walking cycle can be specified from the positional relationship between the right foot and the left foot. . Also, one walking cycle can be specified from the positional relationship between the right foot and the right knee. Furthermore, it is possible to specify one walking cycle by causing the subject to perform predetermined operations such as button operation and utterance at the start time and end time of one walking cycle and detecting the predetermined operations.

As described above, in the first embodiment, the statistical information obtained by statistically processing the walking sample data based on the causal relationship between the acceleration of the waist and the knee joint moment, and the acceleration data of the waist of the subject who is walking are used. Thus, the knee joint moment of the subject during walking is estimated. Since the acceleration data of the waist during walking can be acquired by a sensor such as an acceleration sensor or Kinect (registered trademark) sensor, according to the first embodiment, compared to a method using a floor reaction force meter or the like. The knee joint moment during walking can be easily estimated.
Moreover, since the acceleration data of the waist is less noise than the acceleration data of the knee itself, it can be said that the estimation accuracy can be prevented from being lowered according to the first embodiment. This is because the knee moves violently in various directions than the waist during walking, and therefore, the acceleration of the knee may include various components not correlated with the knee joint moment.
Furthermore, due to the ease of wearing the acceleration sensor, obtaining the acceleration data for the lower back reduces the burden on the subject than obtaining the acceleration data for the knee, and achieves high usability in the first method. Can do.

[Walk analysis device]
Next, a gait analyzer (hereinafter, sometimes referred to as a first device) in the first embodiment will be described. The first device can execute the first method described above.

FIG. 4 is a diagram conceptually illustrating a hardware configuration example of the gait analyzer 10 in the first embodiment.
The first device 10 is a so-called computer, and includes, for example, a CPU (Central Processing Unit) 11, a memory 12, an input / output interface (I / F) 13, a communication unit 14, and the like that are connected to each other via a bus. The number of hardware elements forming the first device 10 is not limited, and these hardware elements may be collectively referred to as an information processing circuit. The first device 10 may include hardware elements not shown in FIG. 4, and the hardware configuration of the first device 10 is not limited.

The CPU 11 may be configured by an application specific integrated circuit (ASIC), a DSP (Digital Signal Processor), a GPU (Graphics Processing Unit), or the like in addition to a general CPU.
The memory 12 is a RAM (Random Access Memory), a ROM (Read Only Memory), or an auxiliary storage device (such as a hard disk).
The input / output I / F 13 can be connected to user interface devices such as the output device 15 and the input device 16. The output device 15 is at least one of a device such as an LCD (Liquid Crystal Display) or a CRT (Cathode Ray Tube) display, a device that displays a screen corresponding to drawing data processed by the CPU 11, a printing device, and the like. The input device 16 is a device that receives an input of a user operation such as a keyboard and a mouse. The output device 15 and the input device 16 may be integrated and realized as a touch panel.
The communication unit 14 performs communication via a communication network with other computers, exchange of signals with other devices, and the like. A portable recording medium or the like can be connected to the communication unit 14. The communication unit 14 may be connected to an acceleration sensor 17 or a camera (not shown).
The first device 10 may be a device that includes the acceleration sensor 17.

FIG. 5 is a diagram conceptually illustrating a processing configuration example of the gait analyzer 10 in the first embodiment.
The first device 10 includes an acquisition unit 21, an information storage unit 23, a calculation unit 25, and the like. These processing modules are software elements, and are realized, for example, by executing a program stored in the memory 12 by the CPU 11. This program is installed from, for example, a portable recording medium such as a CD (Compact Disc) or a memory card or another computer on the network via the input / output I / F 13 or the communication unit 14, and stored in the memory 12. May be.

The acquisition unit 21 acquires acceleration data in the waist of the subject who is walking. That is, the acquisition unit 21 performs the above-described step (S11). The “lumbar region” and the “acceleration data” are as described above.
When the acceleration sensor 17 is attached to the waist of the subject, the acquisition unit 21 acquires acceleration data measured by the acceleration sensor 17. At this time, the acquisition unit 21 can acquire acceleration data from the acceleration sensor 17 by communicating with the acceleration sensor 17 connected to the communication unit 14 wirelessly or by wire. Moreover, the acquisition part 21 may acquire the said acceleration data via a cloud service or a portable recording medium.
When the first apparatus 10 includes the acceleration sensor 17, the acquisition unit 21 can acquire acceleration data measured by the acceleration sensor 17 through an internal bus.
Moreover, the acquisition part 21 can also calculate the said acceleration data using a motion capture technique. For example, the acquisition unit 21 receives video data of a subject who is walking through the communication unit 14 and detects the waist of the subject from the video data. The acquisition unit 21 can also calculate the acceleration data by acquiring the locus data of the position on the lower back of the subject based on the detection information and differentiating the locus data of the position.

  The information storage unit 23 stores statistical information of walking sample data measured in advance from the population. “Walking sample data” and “statistical information of walking sample data” are as described above. The “statistical information of the walking sample data” may be generated by the first device 10 or may be generated by another computer.

  The calculation unit 25 uses the statistical information of the walking sample data stored in the information storage unit 23 and the acceleration data acquired by the acquisition unit 21 to calculate a moment value generated at the subject's knee joint during walking. That is, the calculation unit 25 performs the above-described step (S12).

The calculation method of the moment value by the calculation unit 25 is the same as the knee joint moment estimation method described in the step (S12).
For example, the calculation unit 25 calculates the moment value using acceleration data at each of a plurality of predetermined timings in one walking cycle obtained from the acceleration data acquired by the acquisition unit 21.
Further, when the acceleration data acquired by the acquisition unit 21 includes two or more direction components orthogonal to each other, the calculation unit 25 performs the direction at each timing individually predetermined for each of the two or more direction components. It is preferable to calculate the moment value using the acceleration value.
Further, when the acceleration data of the right waist is acquired by the acquisition unit 21, the calculation unit 25 uses the acceleration data at each of a plurality of timings determined in advance from the right stance phase, and the subject's during walking. It is preferable to calculate a moment value generated in the right knee joint. When the left waist acceleration data is acquired by the acquisition unit 21, the calculation unit 25 uses the acceleration data at each of a plurality of timings determined in advance from the left stance phase, and the left knee of the subject during walking. It is preferable to calculate a moment value generated in the joint. These grounds are as described above.
When each of the plurality of timings has a predetermined time width individually, the calculation unit 25 calculates an average value of acceleration values in the same direction within each time width of the plurality of timings, and calculates the calculated average value The moment value may be calculated using
Further, the calculation unit 25 specifies one walking cycle in the acceleration data based on the vertical direction component indicated by the acceleration data acquired by the acquisition unit 21, and from the acceleration data in the specified one walking cycle. The acceleration data at each of the plurality of timings can be specified. The method for specifying one walking cycle using the vertical direction component is also as described above.

The first device 10 can cause the output device 15 to output the moment value calculated by the calculation unit 25. Thereby, the moment value can be output in various modes such as display, sound, and printing.
According to the first device 10, since the first method described above is executed, the same effect as the first method described above can be obtained.

[Second Embodiment]
In the second embodiment, the knee joint moment normalized by the physique data is estimated with high accuracy by further using the physique data of the subject. In the following description, the same contents as those in the first embodiment will be omitted as appropriate, and the contents different from those in the first embodiment will be mainly described.

[Walk analysis method]
First, the gait analysis method in the second embodiment (hereinafter may be abbreviated as the second method) will be described. FIG. 6 is a flowchart showing a gait analysis method in the second embodiment.
As shown in FIG. 6, the second method includes the step of acquiring physique data of the subject (S21), the step of acquiring acceleration data of the waist of the subject during walking (S22), and the knee joint of the subject during walking. The step (S23) which estimates the moment which arises is included.

The “physique data” in the step (S21) is a kind of value or a plurality of kinds of values that can indicate the physique of the subject, for example, height, weight, waist height, knee height, leg length. Is at least one kind of value. That is, the physique data is a physique value or a physique value group.
The acquisition method of the physique data in a process (S21) is not restrict | limited.
For example, when the computer executes the step (S21), the computer can display a physique data input screen and acquire the physique data in accordance with a user input operation on the screen. The computer can also obtain the physique data of the subject from an external database that stores the subject information. The computer may acquire a weight value as a physique data from a connected scale, or extract image information of a subject based on image data obtained from a camera with a predetermined angle of view. Based on the obtained image information and angle of view information, at least one of the subject's height, waist height, knee height, leg length, etc. can be calculated as physique data.

Step (S22) is the same as step (S11) in the first method.
Step (S23) corresponds to step (S12) in the first method, and differs from step (S12) in the following points.
The statistical information of the walking sample data used for estimating the knee joint moment reflects the physique data of the population. In other words, statistical information obtained by statistically processing the walking sample data and the physique data regarding each person in the population is used. In the step (S23), the knee joint moment normalized with the physique data is estimated using the physique data acquired in the step (S21), the acceleration data acquired in the step (S22), and the statistical information.

  In the second method, the statistical information is generated on the basis of a causal relationship between a combination of waist acceleration and physique information and a knee joint moment normalized by the physique information. This causal relationship has also been newly found by the present inventors. For example, the statistical information is a multiple regression equation obtained by multiple regression analysis using at least the walking sample data and physique data of the population, and includes a plurality of acceleration values and a plurality of acceleration values at a plurality of timings in one walking cycle. It is a multiple regression equation in which a physique value is an explanatory variable and a moment value normalized by one or more physique values is an objective variable. In this case, in the step (S23), the moment normalized by the physique data is applied to the multiple regression equation by applying the physique data obtained in the step (S21) and the acceleration data obtained in the step (S22). A value can be calculated.

As described above, in the second embodiment, the physique data of the subject is further acquired, and the knee joint moment normalized with the physique data is estimated using the acceleration data and the physique data in the waist while walking. In addition, since the estimated knee joint moment is normalized by the physique data, a standard index indicating the load applied to the knee joint can be obtained.
In the second embodiment, the knee joint moment normalized by the physique data is estimated using the acceleration data and the physique data of the waist of the subject. However, since the knee joint moment is affected by the physique information such as the weight and the knee length, it is possible to estimate the unnormalized knee joint moment using the acceleration data and the physique data.

[Walk analysis device]
Hereinafter, the gait analyzer in the second embodiment (hereinafter may be abbreviated as the second device) will be described. The second device can execute the second method described above.
The hardware configuration and processing configuration of the second device are the same as those of the first device (see FIGS. 4 and 5).

The acquisition unit 21 of the second device further acquires the physique data of the subject. About the definition of "physique data" and the acquisition method of the physique data by the acquisition part 21, it is the same as that of the above-mentioned process (S21). That is, the acquisition unit 21 of the second apparatus executes the above-described step (S21) and step (S22).
The statistical information stored in the information storage unit 23 of the second device includes the physique data of the population as described above. In other words, the information storage unit 23 stores statistical information obtained by statistically processing the walking sample data and the physique data regarding each person in the population.

The calculation unit 25 of the second device further uses the physique data of the subject acquired by the acquisition unit 21 to calculate the moment value normalized by the physique data. The calculation method of the moment value by the calculation unit 25 is the same as the normalized moment estimation method in the above-described step (S23). That is, the calculation unit 25 performs the above-described step (S23).
For example, the statistical information stored in the information storage unit 23 is a multiple regression equation acquired by multiple regression analysis using at least the walking sample data and physique data of the population, and a plurality of statistical information at a plurality of timings in one walking cycle. The calculation unit 25 is acquired by the acquisition unit 21 when a multiple regression equation having an acceleration value and one or more physique values as explanatory variables and a moment value normalized by the one or more physique values as an objective variable is included. The moment value normalized with the physique data is calculated by applying the acceleration data and the physique data to the multiple regression equation.

According to the second device, since the second method described above is executed, the same effects as those of the second method described above can be obtained.
Examples will be given below to describe the above-described embodiments in more detail. In the following examples, the effects of the above-described embodiments are demonstrated by explaining a part of the experiment conducted by the present inventors and part of the results. In addition, the content of each above-mentioned embodiment is not limited to the following content.

150 males and females (75 males and 75 females) aged 20 to 75 who can walk on their own with no pain when walking are taken as the population of subjects. Classified into groups. As will be described later, the first group is a group for acquiring a regression equation, and the second group is a group for evaluating the acquired regression equation.
First group: 75 (37 men, 38 women)
Second group: 75 (38 men, 37 women)
Furthermore, t tests were performed on the body characteristics (age, height, weight, BMI) of the first group and the second group, and it was confirmed that no significant difference was found in the body characteristics of both groups.
The following measurements were made on the population. In the following measurements, each subject was instructed to refrain from excessive exercise and drinking the previous day.

First, a reflective marker is attached to the subject's right waist (upper anterior iliac spine) and right knee joint, and the position of the reflective marker during walking of the subject is a motion capture system (VICON MX system, VICON NEXUS made by VICON). Measured with. Furthermore, during the walking, the subject was passed over a floor reaction force meter (BP400600-1000P manufactured by AMTI) to measure the floor reaction force. The position of the reflective marker was measured at 200 hertz (Hz), and the floor reaction force was measured at 1000 Hz.
FIG. 7 is a diagram showing a state of measurement. As shown in FIG. 7, the subject wears a reflective marker at a plurality of predetermined sites and walks linearly at a predetermined place. The subject passes over the floor reaction force meter FR installed during the walking. The attachment positions of the reflective markers shown in FIG. 7 are merely examples, and all the reflective markers need not be used.

The position coordinate data (time series data) and floor reaction force data of each subject measured as described above were processed as follows. Between the data of the position coordinate data (between the measurement intervals of 200 Hz) was appropriately interpolated.
As preprocessing, first, noise such as position blur was removed by applying a fourth-order Butterworth low-pass filter to the position coordinate data of each subject in both groups. The cut-off frequency at this time was set to 10 Hz. In the subsequent processing, the position coordinate data from which noise is removed in this way is used, and in the following description, it is simply referred to as position coordinate data.
Subsequently, acceleration data of the waist of each subject was calculated from the position coordinate data of the waist of each subject in both groups. The acceleration data was calculated for the directional components in the front / rear direction, the left / right direction, and the vertical direction of the waist. Each direction component of acceleration was calculated by differentiating the position coordinates twice.
Then, a single walking cycle is cut out from the calculated acceleration data according to the extreme value of the vertical acceleration accompanying the landing of the right foot, and the extracted acceleration data of one walking cycle is divided into 101 equal parts (about 10 milliseconds). In total, 303 acceleration components were extracted for each subject.
Moreover, the weight value of each test subject of both groups was extracted based on the measured value by a floor reaction force meter.

Next, using the position coordinate data and floor reaction force data of each subject, the maximum value of the right knee joint reaction moment during one walking cycle is calculated for each subject in both groups, and the calculated maximum value is Normalized by subject's weight. Hereinafter, this normalized value is used as the maximum value of the right knee joint varus moment of each subject, and the normalized value is also expressed as the right knee joint varus moment maximum value in the following description.
In addition, the calculation method of the above-mentioned acceleration data and right knee joint varus moment is based on the whole body walking measurement pattern in the walking database of National Institute of Advanced Industrial Science and Technology (AIST) (see Non-Patent Document 2). .

Then, a multiple regression equation was obtained by performing multiple regression analysis on the maximum value of the right knee joint varus moment calculated for the first group, acceleration data (303 components) of one walking cycle, and body weight value. Specifically, the following multiple regression equations were obtained with the maximum value of the right knee joint varus moment as the objective variable and the specific component in the weight value (kg) and acceleration data of each subject as the explanatory variables. .
Maximum knee joint varus moment (Nm / kg)
= 0.643 + (weight value) × (−0.007) + (X [17-19]) × (0.008) + (X [24-29]) × (0.018) + (Y [3 −5]) × (−0.004) + (Y [18-20]) × (0.006) + (Y [29-31]) × (0.003) + (Z [25-27]) × (−0.004)
In the above equation, X [] represents the average value of acceleration in the left-right direction at the timing indicated by [] within one walking cycle, and Y [] represents the front-rear direction at the timing indicated by [] within one walking cycle. Z [] indicates the average acceleration in the vertical direction at the timing indicated by [] within one walking cycle. The timing here is shown in a time rate range where the entire walking cycle is 100%. For example, X [24-29] indicates an average value of acceleration in the left-right direction between 24% and 29% of the time rate within one walking cycle.

Multiple regression analysis was performed by the stepwise method, and the optimum explanatory variable that explained the maximum value of the right knee joint varus moment with high accuracy was selected.
In particular, regarding the acceleration data, the degree of correlation with the maximum value of the right knee joint varus moment was considered for the direction component to be used and the timing within one walking cycle (range of time ratio).
FIG. 8 is a diagram conceptually showing the acceleration component selected as the explanatory variable.
As shown in FIG. 8, one or more timings (time rate ranges) were individually selected for each of the three directional components of acceleration. Specifically, for the left-right direction component of acceleration, an average acceleration value at each of two timings (a range of 17% to 19% of time and a range of 24% to 29% of time) is selected. For the direction component, the average acceleration value of each of the three timings (range 3% to 5%, range 18% to 20%, and range 29% to 31%) is selected. For the vertical component of acceleration, an average acceleration value at one timing (range from 25% to 27% of time rate) was selected.
Further, as shown in FIG. 8, in this example, acceleration in the stance phase of the right leg was used, and acceleration in the swing phase of the right leg was not used. This is presumably because the club moment of the right knee joint is the estimation target.

FIG. 9 is a graph showing the accuracy of the multiple regression equation based on the measurement data of the first group. In the graph of FIG. 9, the right knee joint is estimated (calculated) by substituting the acceleration data and the weight value of each subject in the first group into the estimated value, that is, the multiple regression equation, on the vertical axis. The maximum value of the club moment is shown, and the horizontal axis shows the actual value, that is, the maximum value of the right knee joint ball moment calculated from the measurement data (including floor reaction force data) of each subject in the first group. Since the coefficient of determination R ² indicating the degree of correlation between the estimated value and measured value is 0.74, that the estimation accuracy of the knee joint varus moment maximum value according to the multiple regression equation is high have been demonstrated.

Furthermore, the accuracy of the multiple regression equation was verified using the measurement data of the second group.
FIG. 10 is a graph showing the accuracy of the multiple regression equation based on the measurement data of the second group. In the graph of FIG. 10 as well, the estimated value is shown on the vertical axis and the measured value is shown on the horizontal axis, as in FIG.
Since the coefficient of determination R ² between the measured value and the estimate for the second group is 0.41, for the measured data other than the regression equation acquisition group (first group), the knee joint reaction by the multiple regression equation It was demonstrated that the maximum moment value can be estimated with high accuracy.

  The contents given in this embodiment are merely examples. Therefore, the estimated knee joint moment may not be the maximum value during the position walking cycle, and may not be normalized by the weight. Further, the direction component of acceleration used for estimating the knee joint moment and its timing (time rate range) are not limited to the contents of this embodiment. As described in each of the above-described embodiments, acceleration values at the same timing may be used for each direction component, or not all three orthogonal direction components may be used, but one or two direction components may be used. Good.

  In addition, in the some flowchart used by the above-mentioned description, although several process (process) is described in order, the execution order of the process performed by each embodiment is not restrict | limited to the order of the description. For example, the step (S21) and the step (S22) shown in FIG. 6 may be executed in reverse order or may be executed in parallel. In each embodiment, the order of the illustrated steps can be changed within a range that does not hinder the contents. Moreover, each above-mentioned embodiment can be combined in the range in which the content does not conflict.

The above contents can also be specified as follows. However, the above-mentioned content is not limited to the following description.
<1> An acquisition step of acquiring acceleration data in the waist of a subject who is walking;
An estimation step of estimating a moment generated in the knee joint of the subject during walking using statistical information of walking sample data measured in advance from a population and acceleration data acquired in the acquisition step;
Gait analysis method including:

<2> In the estimation step, the moment is estimated using acceleration data at each of a plurality of predetermined timings in one walking cycle obtained from the acceleration data acquired in the acquisition step.
The walking analysis method according to <1>.
<3> The acceleration data acquired in the acquisition step includes two or more direction components orthogonal to each other,
In the estimating step, the moment is estimated by using an acceleration value in the direction at each timing individually determined for each of the two or more direction components.
The walking analysis method according to <2>.
<4> In the acquisition step, acceleration data in the right waist or left waist of the subject during walking is acquired,
In the estimation step, when acceleration data of the right waist is acquired, acceleration data at each of the plurality of timings determined in advance from the right stance phase is used to apply to the right knee joint of the subject during walking. When the left hip acceleration data is acquired by estimating the moment to be generated, the left knee joint of the subject during walking using the acceleration data at each of the predetermined timings from the left stance phase To estimate the moment that occurs in
The gait analysis method according to <2> or <3>.
<5> Each of the plurality of timings has a predetermined time width individually,
In the estimating step, the moment is estimated using an average value of acceleration values in the same direction within each time width of the plurality of timings.
The gait analysis method according to any one of <2> to <4>.
<6> A step of identifying one walking cycle in the acceleration data based on a vertical direction component indicated by the acceleration data acquired in the acquisition step;
Further including
In the estimation step, the acceleration data at each of the plurality of timings is specified from the acceleration data in the specified one walking cycle.
The gait analysis method according to any one of <2> to <5>.
<7> A step of acquiring the physique data of the subject,
Further including
The statistical information reflects physique data of the population,
In the estimating step, further using the acquired physique data of the subject to estimate the moment normalized with the physique data,
The gait analysis method according to any one of <1> to <6>.
<8> In the estimation step, a multiple regression equation obtained by multiple regression analysis using at least the walking sample data and the physique data of the population, and a plurality of acceleration values at a plurality of timings in one walking cycle And using one or more physique values as explanatory variables and a multiple regression equation with the moment values normalized by the physique values as objective variables, the moment is estimated.
The gait analysis method according to <7>.
<9> Acquisition means for acquiring acceleration data at the waist of the subject who is walking;
Storage means for storing statistical information of gait sample data measured in advance from a population;
Using the statistical information and the acquired acceleration data, calculating means for calculating a moment value generated in the knee joint of the subject during walking;
A gait analyzer comprising:
<10> The calculation unit calculates the moment value using acceleration data at each of a plurality of predetermined timings in one walking cycle, which is obtained from the acceleration data acquired by the acquisition unit.
The gait analyzer according to <9>.
<11> The acceleration data acquired by the acquisition unit includes two or more direction components orthogonal to each other,
The calculation means calculates the moment value using an acceleration value in the direction at each timing individually determined for each of the two or more direction components.
The gait analyzer according to <10>.
<12> The acquisition means acquires acceleration data in the right waist or left waist of the subject during walking,
When the acceleration data of the right waist is acquired, the calculating means uses the acceleration data at each of the plurality of timings determined in advance from the right stance phase to the right knee joint of the subject during walking. When the moment value to be generated is calculated and the acceleration data of the left hip is acquired, the acceleration data at each of the plurality of timings determined in advance from the left stance phase is used to determine the left knee of the subject during walking. Calculate the moment value generated in the joint,
The gait analyzer according to <10> or <11>.
<13> Each of the plurality of timings has a predetermined time width individually,
The calculation means calculates an average value of acceleration values in the same direction within each time width of the plurality of timings, and calculates the moment value using the calculated average value.
The gait analyzer according to any one of <10> to <12>.
<14> The calculation unit specifies one walking cycle in the acceleration data based on the vertical direction component indicated by the acceleration data acquired by the acquiring unit, and the acceleration data in the specified one walking cycle To identify acceleration data at each of the plurality of timings,
The gait analyzer according to any one of <10> to <13>.
<15> The acquisition means further acquires the physique data of the subject,
The statistical information includes physique data of the population,
The calculation means further uses the physique data of the subject to calculate the moment value normalized with the physique data.
The gait analyzer according to any one of <9> to <14>.
<16> The statistical information is a multiple regression equation obtained by multiple regression analysis using at least the walking sample data and the physique data of the population, and a plurality of acceleration values at a plurality of timings in one walking cycle. And a multiple regression equation having one or more physique values as explanatory variables, and a moment value normalized by the one or more physique values as an objective variable,
The calculation means calculates the moment value normalized by the physique data by applying the acceleration data and the physique data acquired by the acquisition means to the multiple regression equation.
The gait analyzer according to <15>.

10 Gait analyzer (first device)
11 CPU
12 Memory 13 Input / output I / F
DESCRIPTION OF SYMBOLS 14 Communication unit 15 Output device 16 Input device 17 Acceleration sensor 21 Acquisition part 23 Information storage part 25 Calculation part

Claims (9)

An acquisition step of acquiring acceleration data in the waist of a walking subject;
An estimation step of estimating a moment generated in the knee joint of the subject during walking using statistical information of walking sample data measured in advance from a population and acceleration data acquired in the acquisition step;
Gait analysis method including: In the estimation step, the moment is estimated using acceleration data obtained from the acceleration data acquired in the acquisition step at each of a plurality of predetermined timings during one walking cycle.
The gait analysis method according to claim 1. The acceleration data acquired in the acquisition step includes two or more direction components orthogonal to each other,
In the estimating step, the moment is estimated by using an acceleration value in the direction at each timing individually determined for each of the two or more direction components.
The gait analysis method according to claim 2. In the acquisition step, the acceleration data in the right waist or left waist of the subject during walking is acquired,
In the estimation step, when acceleration data of the right waist is acquired, acceleration data at each of the plurality of timings determined in advance from the right stance phase is used to apply to the right knee joint of the subject during walking. When the left hip acceleration data is acquired by estimating the moment to be generated, the left knee joint of the subject during walking using the acceleration data at each of the predetermined timings from the left stance phase To estimate the moment that occurs in
The gait analysis method according to claim 2 or 3. Each of the plurality of timings has a predetermined time width individually,
In the estimating step, the moment is estimated using an average value of acceleration values in the same direction within each time width of the plurality of timings.
The gait analysis method according to any one of claims 2 to 4. A step of identifying one walking cycle in the acceleration data based on a vertical direction component indicated by the acceleration data acquired in the acquisition step;
Further including
In the estimation step, the acceleration data at each of the plurality of timings is specified from the acceleration data in the specified one walking cycle.
The walking analysis method according to any one of claims 2 to 5. Obtaining the physique data of the subject,
Further including
The statistical information reflects physique data of the population,
In the estimating step, further using the acquired physique data of the subject to estimate the moment normalized with the physique data,
The gait analysis method according to any one of claims 1 to 6. In the estimation step, a multiple regression equation obtained by multiple regression analysis using at least the walking sample data and the physique data of the population, wherein a plurality of acceleration values and one or more at a plurality of timings in one walking cycle Estimating the moment using a multiple regression equation with the physique value of as an explanatory variable and the moment value normalized with the physique value as the objective variable
The gait analysis method according to claim 7. An acquisition means for acquiring acceleration data in the waist of the walking subject;
Storage means for storing statistical information of gait sample data measured in advance from a population;
Using the statistical information and the acquired acceleration data, calculating means for calculating a moment value generated in the knee joint of the subject during walking;
A gait analyzer comprising:

Images