JP2005342254A - Walking period detecting apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a walking period detecting apparatus capable of providing a highly precise walking period from a frequency peak of a plurality of walking accelerations generated when a person shuffles or the like. <P>SOLUTION: This walking period detecting apparatus is provided with an acceleration detecting part 1 detecting accelerations of a subject, a peak frequency detecting part 2 performing a frequency analysis to the accelerations detected by the acceleration detecting part 1 and detecting a frequency in the peak of a power spectrum satisfying a predetermined condition, a period calculation part 3 calculating a walking period, or an interval of one step of walking, from the frequency detected by the peak frequency detecting part 2, a storage part 4 storing the results of the period calculation part 3, and a display part 5 displaying the detection results of the walking period. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a walking cycle detection device for detecting a walking cycle of a person, and more particularly to a walking cycle detection device for accurately detecting a walking cycle of a subject under rehabilitation conditions such as a footrest.

  Under the background of the arrival of an aging society in recent years, it is urgently necessary to establish a device and a treatment method for assisting or recovering the walking function so that the elderly can walk safely and do not become bedridden. Further, for example, in artificial joint patients and osteoarthritis patients, it is one of important issues to quantitatively grasp the degree of recovery or deterioration of lower limb function.

  Moreover, not only an elderly person and an artificial joint patient but by extracting a gait, it becomes possible to measure and analyze human behavior and posture. If the behavior pattern can be grasped, it becomes possible to perform more comfortable control of lighting and air conditioning, to operate the equipment safely, and to improve the living environment.

  In this way, when assisting or restoring the walking function or quantitatively grasping it, it is necessary to quantitatively evaluate how the subject is walking. Therefore, it is very important to accurately detect the walking cycle, which is the interval of one step of the subject's walking.

  Conventionally, several devices for detecting a walking cycle have been proposed. As a conventional walking cycle detection device, there is one that measures the acceleration of a subject who is walking, performs frequency analysis on acceleration data, and detects the walking cycle from the reciprocal of the frequency at which the power spectrum is maximized ( For example, see Patent Document 1).

  FIG. 21 is a reference diagram showing a conventional walking cycle detection method described in Patent Document 1. Reference numerals 501 to 504 in FIG. 21 indicate frequency analysis of acceleration data when the subject walks with various walking cycles, and the horizontal axis indicates the frequency of walking and the vertical axis indicates the common logarithm of the power spectrum. is there. In FIG. 21, when the walking cycle is detected, the reciprocal of the frequency at the largest value of the power spectrum is used as the walking cycle. For example, in 503, since the frequency at the peak time of the power spectrum is 1 Hz, by calculating the reciprocal of this frequency, the walking cycle is 1 second, that is, the time required for one step is 1 second.

As described above, in the configuration of the conventional walking cycle detection apparatus, the reciprocal of the frequency at which the power spectrum is maximized is set as the walking cycle as a result of performing frequency analysis on the acceleration data during walking.
Japanese Patent Application No. 2003-132392

  However, when the patient is a leg or the like in a hospital or the like, the movement of the body during walking is more complicated than normal, and the movement of the body in the stepping of each of the left and right feet is different, which is calculated from the acceleration of the patient A plurality of power spectrum peaks may occur, and the power spectrum may be maximized at a position other than the frequency of one step.

  In this case, if the reciprocal of the frequency in the maximum power spectrum is determined to be a one-step walking cycle, an incorrect walking cycle is calculated, degrading the accuracy of walking evaluation, resulting in a decrease in accuracy in quantifying walking functions. This is a challenge.

  The present invention solves the above-described conventional problem, and can detect a walking cycle more accurately even when the subject is in various walking states such as a normal walking file or the like. The purpose is to provide.

In order to solve the conventional problem, a walking cycle detection device according to the present invention is [Claim 1].
With this configuration, in the walking cycle detection device according to the present invention, the threshold setting unit determines the threshold of the power spectrum used for calculating the walking cycle, and the peak detection unit detects the peak of the power spectrum that is equal to or higher than the threshold, In the walking cycle detection means, in order to detect the walking cycle using these peak frequencies, the walking cycle can be detected more accurately regardless of how the subject normally walks or walks. Can do.

The walking cycle detection device according to the present invention is [Claim 5].
With this configuration, the walking cycle detection unit can detect the walking cycle according to the number of power spectrum peaks equal to or higher than the threshold detected by the peak detection unit. Therefore, the walking cycle detection device according to the present invention. Can detect the walking cycle more accurately regardless of how the subject normally walks or walks.

Furthermore, the walking cycle detection device according to the present invention is [Claim 12].
For this reason, in the displacement detection means, the displacement of the subject is detected, and the acceleration calculation means can calculate the acceleration from this data. Calculation can be performed.

  In the present invention, not only can it be realized as a walking cycle detection device, but also a walking cycle detection method using steps provided in the walking cycle detection device as a step, and a program for realizing the walking cycle detection method with a computer or the like. Needless to say, the program can be realized or distributed via a recording medium such as a DVD or CD-ROM or a transmission medium such as a communication network.

  According to the walking cycle detection device of the present invention, it is possible to more accurately detect the walking cycle of the left and right feet during walking, regardless of how the subject walks, such as a normal walking file or a walking foot. It becomes.

Embodiments of a walking cycle detection device according to the present invention will be described below with reference to the drawings.
(Embodiment 1)
FIG. 1 is an overall view showing a specific configuration example of the walking cycle detection device according to the first embodiment. The acceleration detection unit is provided in the acceleration detection device 10 fixed to the subject's waist. The acceleration detection device 10 will be described in detail with reference to FIG.

  The computer 7 includes each processing unit shown in FIG. 4 described later, and carries a peak frequency detection unit, a cycle calculation unit, a storage unit, a display unit, and a program for executing the walking cycle detection method of the present invention. A walking cycle detection method can be implemented based on acceleration data transmitted from the acceleration detection device 10 including a medium. In addition, the person 8 is a test subject whose walking cycle is detected, and is a healthy person, a lower limb dysfunction patient (subject), or the like.

  In FIG. 1, the communication method of the acceleration detection device 10 is wireless communication, but other communication methods such as wired communication may be used. Further, the acceleration detection device 10 may be equipped with a peak frequency detection unit, a period calculation unit, a storage unit, and a display unit included in the computer 7 and attached to the body.

  FIG. 2 is an external view of a specific device of the acceleration detection device 10 including the acceleration detection unit according to the first embodiment. This acceleration detection device 10 is for detecting acceleration when a subject wearing this device is walking using a built-in acceleration sensor. This apparatus has a slightly thin rectangular parallelepiped shape (for example, 5 cm (vertical) × 8 cm (horizontal) × 1.5 cm (thickness)) of a business card size, and includes buttons 101, a liquid crystal panel 102, and a speaker 103. And a microphone 104. The acceleration detecting device 10 is attached to the human body by a fixing band 105 for fixing through the belt, for example.

  The buttons 101 are buttons pressed by the subject or the user in order to determine the position where the acceleration detection device 10 is mounted. For example, when the acceleration detection device 10 is worn on the left waist of the subject, the “left” button is pressed. Similarly, the right button when worn on the subject's right waist, the front button when worn on the subject's abdomen, and the back button when worn on the lower back of the subject, Each is pressed.

  The liquid crystal panel 102 displays the operation mode of the acceleration detection unit 1 and an error message. The speaker 103 outputs sound for giving a predetermined instruction to the user when operating the acceleration detection device 10. The microphone 104 receives an instruction from the user by voice when operating the acceleration detection device 10.

FIG. 3 is an external view showing a wearing example of the acceleration detection device 10 according to the first embodiment in a subject.
This figure is a view of the subject as viewed from the left. The arrows 31 and 32 indicate that the forward direction of the subject is the positive direction of the acceleration information in the front-rear direction and the upward direction of the subject is the positive direction of the acceleration information in the vertical direction, respectively.

  FIG. 4 is a functional block diagram of walking cycle detection apparatus 100 according to Embodiment 1 of the present invention. In the first embodiment, the acceleration detection device 10 includes the acceleration detection unit 1 and the computer 7 includes other processing units.

  The walking cycle detection device 100 is a device that calculates the walking cycle of the subject's left and right legs, performs frequency analysis on the acceleration detected by the acceleration detection unit 1 and the acceleration detection unit 1, and has a power spectrum that satisfies a predetermined condition. Peak frequency detection unit 2 that detects a frequency at a peak, cycle calculation unit 3 that calculates a walking cycle that is a cycle of one step from the frequency detected by the peak frequency detection unit 2, and an accumulation that stores a result of the cycle calculation unit 3 The display part 5 which displays the calculation result of the walk period accumulate | stored in the part 4 and the accumulation | storage part 4 is provided.

In the following, each processing unit will be described in detail.
The acceleration detection unit 1 includes a biaxial acceleration sensor, an amplifier circuit, and a Bluetooth unit as a wireless communication module. In addition, it is fixed to the human waist by means of a belt-like wearing tool, and acceleration generated in the front-rear direction and motion generated in the vertical direction as the human body moves is detected as a voltage change, and the voltage signal is wirelessly converted after A / D conversion. It outputs to the peak frequency detection part 2 by communication.

  The peak frequency detection unit 2 includes an input / output interface including a wireless communication module corresponding to the wireless communication of the acceleration detection unit 1, a RAM that temporarily stores data, a processing unit that performs numerical calculation on acceleration information, A frequency analysis is performed on the acceleration information in the vertical direction or the front-rear direction, and a program for detecting a frequency at a peak satisfying a predetermined condition in the power spectrum of the acceleration is provided, and the result is output to the period calculation unit 3.

  The period calculation unit 3 performs a numerical operation on the input / output interface corresponding to the output from the peak frequency detection unit 2, the RAM that temporarily stores data, and the frequency at the peak of the power spectrum that satisfies a predetermined condition. A processing unit and a program for obtaining a walking cycle from the output of the peak frequency detection unit 2 are provided, and the result is output to the storage unit 4.

The accumulation unit 4 is a processing unit that includes an input / output interface corresponding to the output from the cycle calculation unit 3 and a disk device, and accumulates the output from the cycle calculation unit 3 in the disk device.
The display unit 5 is specifically a monitor, and displays the results stored in the storage unit 4.

Next, with reference to FIG. 5, it is a flowchart explaining the operation | movement procedure at the time of the walk period detection of the walk period detection apparatus 100 which concerns on this invention.
First, the acceleration detector 1 detects accelerations in the front-rear direction, the up-down direction, and the like when the subject walks (S501).

  Next, the peak frequency detection unit 2 analyzes the frequency of the acceleration data and detects the power spectrum for each frequency. The power spectrum is an index indicating which frequency component is contained in a large amount.

  And the peak frequency detection part 2 sets the threshold value of a power spectrum. A method for determining the threshold will be described with reference to FIG. FIG. 6 is a diagram illustrating a table 600 illustrating a method for determining a power spectrum threshold. As shown in this figure, the threshold value determining method in the walking cycle detection apparatus 100 is, for example, a method of determining as an empirical rule from an experiment according to the maximum value of the power spectrum calculated by frequency analysis (this embodiment) 1), a method using half of the maximum value of the power spectrum as a threshold (corresponding to the second embodiment described below), and a method using a value divided by the maximum value of the power spectrum as a threshold (described in the third embodiment below). Response) etc. can be considered.

  Next, the period calculation unit 2 determines the number of peak frequencies corresponding to the peak of the power spectrum that is equal to or greater than the set threshold, and calculates the walking period according to the number of peak frequencies. FIG. 7 is a diagram illustrating a table 700 for explaining a case where the walking cycle calculation method is changed according to the number of peak frequencies in the walking cycle detection apparatus 100 according to the present invention. For example, when the number of peak frequencies is one, the reciprocal value of one peak frequency is the walking cycle, and when the number of peak frequencies is two, as shown in A or B, the smaller one If the reciprocal of the value twice the peak frequency or the reciprocal of the larger peak frequency is the walking cycle, and the number of peak frequencies is 3 or more, as shown in A or B, the minimum peak frequency The reciprocal of the double value and the reciprocal of the average value of the remaining peak frequencies excluding the minimum frequency can be used as the walking cycle.

  FIG. 8 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result using the walking cycle detection method according to the first embodiment. FIG. 8 shows data when the subject walks a healthy person. In the first embodiment, as shown in FIG. 3, the upward direction corresponds to the positive direction for the acceleration information in the vertical direction.

  FIG. 8 (a) displays the acceleration in the vertical direction when the subject walks in time series, and FIG. 8 (b) displays the result of frequency analysis in relation to the frequency and power spectrum. . The graph in FIG. 8A corresponds to the vertical acceleration, and the graph in FIG. 8B corresponds to the result of frequency analysis performed on the vertical acceleration shown in FIG.

  In FIG. 8 (a), since the subject is a healthy person and the left and right legs are in a constant rhythm during walking, the time required for the right footpad grounding-left footpad grounding in the vertical acceleration waveform (0.53 seconds) and the left footpad are shown. The time required for grounding-right footpad grounding is approximately the same (0.52 seconds), and the amplitudes of acceleration of the left and right feet are also substantially the same. For this reason, the acceleration waveform is close to a sine wave, and the frequency analysis result of FIG. 8B also shows a high peak at a frequency (1.83 Hz) corresponding to one step of walking (hereinafter referred to as a walking cycle). Yes.

  In the peak frequency detection method, the walking cycle detection device 100 determines the frequency at the peak of the power spectrum when the power spectrum value exceeds a predetermined value Pth as a result of frequency analysis of acceleration in the vertical direction or the front-back direction. It is characterized by the peak frequency. In FIG. 8A, the predetermined value Pth is determined by an empirical value based on experimental data, for example, and is 80000, which is about half of the maximum value of the power spectrum.

  Also in FIG. 8B, the predetermined value Pth is set to 80000, and the peak frequency at which the power spectrum value is equal to or greater than Pth is 1.83 Hz, which is only one corresponding to A. This indicates that the left and right legs are walking at a constant rhythm when the subject walks.

  FIG. 9 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result according to the first embodiment. In FIG. 9, the test subject is a patient with lower limb dysfunction, specifically, a patient with a left foot function failure. In FIG. 9, the forward direction corresponds to the positive direction for the acceleration information in the front-rear direction.

  FIG. 9A shows the acceleration in the front-rear direction when the subject walks on the pedestrian in time series, and FIG. 9B shows the frequency analysis result in a relationship between the frequency and the power spectrum. It is. The graph of FIG. 9A corresponds to the result of frequency analysis performed on the longitudinal acceleration, and FIG. 9B corresponds to the result of frequency analysis performed on the longitudinal acceleration.

  In FIG. 9 (a), since the subject is a leg dysfunction patient and is walking on the foot, the time required for kicking the right foot toe-left foot toe kick in the longitudinal acceleration waveform (0.67 seconds) and the left foot toe kick However, the time required for kicking out the right foot toe (0.67 seconds) is almost equal. However, when the leg on the healthy side is stepped on, the force of the foot on the affected side is weak, so it is not possible to step on the foot greatly, the back and forth motion of the body is reduced, and the acceleration amplitude when the leg on the healthy side is stepped on Becomes smaller. For this reason, the acceleration waveform is close to a synthetic wave having a walking cycle (0.67 seconds) and a walking two-step cycle (1.30 seconds), and the frequency analysis result in FIG. 9B also corresponds to the walking cycle. High peaks appear at the frequency (1.48 Hz) and the frequency of walking two steps (0.76 Hz). In the figure, B indicates the peak frequency on the healthy side, and A indicates the peak frequency of movement of the entire body.

  In the peak frequency detection method in FIG. 9B, when the predetermined value Pth is set to 35000, there are two peak frequencies at which the power spectrum value is equal to or greater than Pth, which are 0.76 Hz and 1.48 Hz.

  This is because, in the acceleration in the front-rear direction in the walking of the subject's foot, the amplitude of the acceleration when the foot on the healthy side is stepped down becomes smaller, so the acceleration waveform has a walking cycle (0.67 seconds) and a cycle of two steps. It is close to the synthesized wave having (1.30 seconds), and high peaks appear at the frequency corresponding to the walking cycle (1.48 Hz) and the frequency of walking two steps (0.76 Hz). This confirms the fact that can be read from FIG.

  FIG. 10 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result according to the first embodiment. In FIG. 10, the test subject is a patient with lower limb dysfunction and a patient with a left foot function failure.

  Fig. 10 (a) shows the acceleration in the vertical direction when the subject walks on a pedestrian in time series. Fig. 10 (b) shows the result of the frequency analysis in terms of the relationship between frequency and power spectrum. It is displayed. Further, the graph of FIG. 10A corresponds to the result of frequency analysis performed on the vertical acceleration, and FIG. 10B corresponds to the result of frequency analysis performed on the vertical acceleration.

  In FIG. 10 (a), since the subject is a leg dysfunctional patient and is walking on the foot, the time required for grounding the right footpad-grounding the left footpad (0.68 seconds) and grounding the left footpad-grounding the right footpad in the vertical acceleration waveform. The time required for (0.45 seconds) is different. Furthermore, when the leg on the healthy side touches down, a large impact is generated by covering the affected leg, and on the contrary, a small impact occurs when the leg on the affected side is stepped on, so that the affected heel touches the ground. The amplitude at the time becomes smaller. For this reason, the acceleration waveform is close to a composite wave having a walking cycle of the right foot (0.68 seconds), a walking cycle of the left foot (0.45 seconds), and a cycle of two steps of walking (1.19 seconds). The frequency analysis result of 10 (b) also shows the frequency corresponding to the walking cycle of the right foot (B: 1.46 Hz in the figure), the frequency corresponding to the left foot (C: 2.21 Hz in the figure), and the frequency in two steps of walking. A high peak appears at (A: 0.84 Hz in the figure).

  In the peak frequency detection method in FIG. 10B, when the predetermined value Pth is set to 20000, there are three peak frequencies at which the power spectrum value is equal to or greater than Pth, 0.84 Hz, 1.46 Hz, and 2.21 Hz. It is.

  This means that the acceleration in the vertical direction of the subject's gait walking is the walking cycle of the right foot (0.68 seconds), the walking cycle of the left foot (0.45 seconds), and the cycle of walking 2 steps (1.19 seconds). A high peak appears in the frequency (1.46 Hz) corresponding to the walking cycle of the right foot, the frequency (2.21 Hz) corresponding to the left foot, and the frequency (0.84 Hz) in two steps of walking, which is close to the synthesized wave. This supports the fact that it can be read from FIG.

  Note that the predetermined value Pth set in the first embodiment is not limited to the above value, and may take other values. Further, the threshold value of the power spectrum can be a default value or an input value.

Hereinafter, a method of calculating the walking cycle according to the number of peak frequencies equal to or higher than the detected threshold in the cycle calculation unit 3 according to the present invention will be described.
FIG. 8B shows the result of performing frequency analysis on the acceleration information of the subject in the first embodiment and obtaining the peak frequency by the peak frequency detection method. The graph in the figure relates to the first embodiment and displays the relationship between the frequency and the power spectrum by performing frequency analysis on the vertical acceleration when the subject walks.

  In FIG. 8B, there is only one peak frequency detected by the peak frequency detection method. This peak frequency just corresponds to the walking cycle. In this case, the cycle calculation unit 3 calculates the reciprocal of the detected peak frequency as a walking cycle.

That is, the walking cycle is 1 / 1.83 = 0.55 (seconds) in the case of FIG.
FIG. 9B shows the relationship between the frequency and the power spectrum by performing frequency analysis on the acceleration in the front-rear direction when the subject walks with a foot.

  In FIG. 9B, there are two peak frequencies detected by the peak frequency detection method. These correspond to the frequency (B in the figure) resulting from the walking cycle and the frequency (A in the figure) resulting from the cycle of two steps. In this case, the reciprocal of the value obtained by doubling the smaller peak frequency (A in the figure) due to two steps of walking is calculated as the walking cycle.

That is, in the case of FIG. 9B, the walking cycle is 1 / (0.76 × 2) = 0.66 (seconds).
In addition, in FIG. 9B, as the period calculation method, the detected two peak frequencies are frequencies resulting from the walking period (B in the figure), and frequencies resulting from the period of two walking steps (A in the figure). Therefore, the reciprocal of the larger peak frequency (B in the figure) due to the walking cycle may be used as the walking cycle.

In this case, in the case of FIG. 9B, the walking cycle is 1 / 1.48 = 0.68 (seconds).
FIG. 10 (b) shows the result of performing frequency analysis on the acceleration information of the subject and obtaining the peak frequency by the peak frequency detection method.

  FIG. 10B shows a case where there are three peak frequencies detected by the peak frequency detection method. These correspond to the frequency resulting from the walking cycle of the right foot (C in the figure), the frequency resulting from the left foot (B in the figure), and the frequency resulting from two steps of walking (A in the figure). In this case, the reciprocal of the value obtained by doubling the minimum peak frequency (A in the figure) due to two steps of walking is defined as the walking cycle.

That is, in the case of FIG. 10B, the walking cycle is 1 / (0.84 × 2) = 0.59 (seconds).
In addition, in FIG. 10B, the walking cycle calculation method includes a frequency in which the detected peak frequency is derived from the walking cycle of the right foot (C in the diagram), a frequency derived from the left foot (B in the diagram), and walking. Since it corresponds to the frequency (A in the figure) due to two steps, the frequency (C in the figure) due to the walking cycle of the right foot and the walking cycle of the left foot, excluding the minimum peak frequency (A in the figure) The reciprocal of the value obtained by averaging the frequencies (B in the figure) may be used as the walking cycle.

In this case, in the case of FIG. 10B, the walking cycle is 2 / (1.46 + 2.21) = 0.55 (seconds).
FIG. 11 is a diagram illustrating a display example of the walking cycle of the subject displayed on the monitor that is the display unit 5 of the walking cycle detection apparatus 100 according to the first embodiment.

  The monitor includes, for example, a walking cycle, a walking cycle for one step of the right foot (described as “right foot walking cycle” in the diagram), a walking cycle for one step of the left foot (described as “left foot walking cycle” in the diagram), The acceleration data in the front-rear direction or the vertical direction, and the result of frequency analysis performed on the acceleration data are displayed. Moreover, the walking cycle displayed largely on the monitor is an average value of the right foot and the left foot.

Then, the subject himself / herself refers to the monitor, thereby confirming his / her walking cycle and recognizing the degree of recovery from rehabilitation.
As described above, the walking cycle detection apparatus 100 according to the first embodiment performs frequency analysis on acceleration data at the time of walking of the subject in the peak frequency detection unit 2, and sets a threshold value for the detected power spectrum. Then, a peak frequency corresponding to a peak equal to or higher than the threshold value is acquired. Moreover, the period calculation part 3 changes the calculation method of a walk period based on the number of this acquired frequency.

  Therefore, no matter how the subject walks, the cycle detection unit 3 can more accurately detect the walking cycle, which is the cycle of one step of the subject during walking, and the walking of the subject can be performed using the detected walking cycle. The function can be quantitatively evaluated.

  Further, in the acceleration data detected by the acceleration detector 1, the difference between the walking cycles of both legs in the body movement is detected from the region corresponding to each leg, and the degree of functional recovery and deterioration is quantified over a predetermined period. Can be analyzed. For example, if the difference between the walking periods of the left and right legs of the subject is reduced, the movement function of the subject is recovering smoothly, and if the difference between the walking periods of the left and right legs is not shortened, the movement of the subject It can be quantitatively evaluated that the function is not recovering.

In the first embodiment, the acceleration in the front-rear direction and the vertical direction of the subject is detected. However, the acceleration in other directions such as the left-right direction may be detected.
Furthermore, the acceleration direction is positive in the forward direction and positive in the upward direction, but other directions may be positive directions.

  Furthermore, the display unit 5 is a monitor. However, the display unit 5 is not limited to display using such visual stimuli, and may be capable of presenting the walking cycle to the subject, such as voice stimulation, light stimulation, and vibration stimulation. That's fine.

  Needless to say, the walking cycle detection apparatus 100 of the present invention can be applied to simply counting the number of steps.

(Embodiment 2)
Next, a second embodiment of the walking cycle detection apparatus 100 according to the present invention will be described with reference to the drawings. In the walking cycle detection device according to the second embodiment, the peak frequency detection method is used to frequency-analyze the vertical acceleration or the longitudinal acceleration, to calculate the maximum value Pmax of the peak of the power spectrum, and to determine the value of the power spectrum. Is a difference between Pmax and a predetermined value Pth or less, the frequency at the peak is detected.

  FIG. 12 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result according to the second embodiment, similar to FIG. 8 described above. Similarly to FIG. 8, the graph of FIG. 12A corresponds to the result of frequency analysis performed on the vertical acceleration, and FIG. 12B corresponds to the result of frequency analysis performed on the vertical acceleration.

In FIG. 12A, an analysis result similar to that in FIG. 8A described above is obtained.
In the peak frequency detection method in FIG. 12B, the predetermined value Pth is Pmax / 2, and in that case, there is only one peak frequency at which the power spectrum value is the difference between Pmax and the predetermined value Pth or less. 1.83 Hz, which indicates that the left and right legs are walking at a constant rhythm when the subject walks.

  FIG. 13 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result according to the second embodiment, as in FIG. In FIG. 13, the subject is a patient with lower limb dysfunction and the function of the left foot is in the same manner as in FIG.

The two graphs in the figure correspond to the result of frequency analysis performed on the longitudinal acceleration in the graph of FIG. 13A and the acceleration in the longitudinal direction of FIG. 13B.
In FIG. 13A, the analysis result is the same as in FIG. 9A.

  In FIG. 13B, the predetermined value Pth is set to Pmax / 2, and in this case, there are two peak frequencies at which the power spectrum value is less than the predetermined value Pth, 0.76 Hz and 1 48 Hz. Accordingly, the analysis result is the same as that in FIG.

  FIG. 14 is a diagram showing the correspondence between the acceleration information of the subject and the frequency analysis result according to the second embodiment, as in FIG. In FIG. 14, the subject is a patient with lower limb dysfunction and the function of the left foot is in the same manner as in FIG. Further, the graph of FIG. 14A corresponds to the result of frequency analysis performed on the vertical acceleration, and FIG. 14B corresponds to the result of frequency analysis performed on the vertical acceleration.

In FIG. 14A, the analysis result is the same as that in FIG.
In the peak frequency detection method in FIG. 14B, the predetermined value Pth is set to Pmax / 2. In this case, there are three peak frequencies where the power spectrum value is the difference between Pmax and the predetermined value Pth or less. .84 Hz, 1.46 Hz and 2.21 Hz. This indicates that the analysis result is the same as that in FIG.

FIG. 12 (b) shows the result of obtaining the peak frequency by the peak frequency detection method in the second embodiment, as in FIG. 8 (b).
In FIG. 12B, since there is only one peak frequency detected by the peak frequency detection method, this peak frequency just corresponds to the walking cycle as in FIG. 8B. That is, in the case of FIG. 12B, the walking cycle is 1 / 1.83 = 0.55 (seconds).

  FIG. 13 (b) relates to the second embodiment, and similarly to FIG. 9 (b), frequency analysis is performed on the acceleration in the front-rear direction when the subject walks on the pedestrian, and the relationship between the frequency and the power spectrum is displayed. It is a thing.

  In FIG. 13B, there are two peak frequencies detected by the peak frequency detection method. In this case, for example, the reciprocal of a value obtained by doubling the smaller peak frequency (A in the figure) due to two steps of walking can be calculated as the walking cycle, that is, the walking cycle is 1 / (0.76 × 2 ) = 0.66 (seconds).

  Moreover, as a period calculation method in FIG.13 (b), it is good also considering the reciprocal number of the larger peak frequency (B in the figure) as a walk period similarly to FIG.9 (b). In this case, the walking cycle is 1 / 1.48 = 0.67 (seconds).

  FIG. 14B shows the result of performing frequency analysis on the acceleration information of the subject in the second embodiment and obtaining the peak frequency by the peak frequency detection method, similarly to FIG. 10B.

  In FIG. 14B, there are three peak frequencies detected by the peak frequency detection method. In this case, the reciprocal of a value obtained by doubling the minimum peak frequency (A in the figure) due to two steps of walking. Can be a walking cycle. That is, the walking cycle is 1 / (0.84 × 2) = 0.59 (seconds).

  Further, as the period calculation method in FIG. 14B, as in FIG. 10B, the frequency (C in the figure) resulting from the walking period of the right foot, excluding the minimum peak frequency (A in the figure), The reciprocal of the value obtained by averaging the frequencies (B in the figure) resulting from the walking cycle of the left foot may be used as the walking cycle. In this case, the walking cycle is 2 / (1.46 + 2.21) = 0.55 (seconds).

  As described above, in the walking cycle detection device according to the second embodiment, the peak frequency detected in the peak frequency detection method is set with a half of the maximum value of the power spectrum as a threshold, and the peak frequency detection is performed. The unit 2 detects a frequency corresponding to the peak of the power spectrum equal to or higher than the threshold as the peak frequency. In addition, it is the same as that of Embodiment 1 mentioned above that the period calculation part 3 changes the calculation method of a walk period according to the number of peak frequencies.

  Therefore, in the walking cycle detection device according to the second embodiment, the walking cycle can be detected more accurately regardless of the walking state of the subject.

(Embodiment 3)
Next, Embodiment 3 of the walking cycle detection apparatus 100 according to the present invention will be described with reference to the drawings. In the peak frequency detection method according to the third embodiment, the vertical acceleration or the longitudinal acceleration is subjected to frequency analysis, the maximum value Pmax of the peak of the power spectrum is calculated, and the value of the power spectrum at each frequency is the maximum value. When the value divided by Pmax is equal to or greater than a predetermined ratio Rth, the peak frequency at the peak is detected.

  FIG. 15 is the same as FIG. 8 and FIG. 12 and shows the correspondence between the acceleration information of the subject (healthy person) and the frequency analysis result according to the third embodiment. The graph of FIG. 15A corresponds to the result of frequency analysis performed on the vertical acceleration, and FIG. 15B corresponds to the result of frequency analysis performed on the vertical acceleration.

  In the peak frequency detection method in FIG. 15B, the predetermined ratio Rth is set to 0.5, and in this case, there is only one peak frequency at which the ratio obtained by dividing the power spectrum by Pmax is equal to or greater than Rth. 83 Hz. This confirms the fact that the left and right legs are walking at a constant rhythm in the walking of the subject, which can be read from FIG.

  FIG. 16 is a diagram showing the correspondence between the acceleration information and the frequency analysis result of the subject (lower limb dysfunction patient, left foot dysfunction) according to the third embodiment, as in FIGS. 9 and 13. The graph of FIG. 16A corresponds to the result of frequency analysis of the longitudinal acceleration, and FIG. 16B corresponds to the result of the frequency analysis of the longitudinal acceleration.

  In the peak frequency detection method in FIG. 16B, the predetermined ratio Rth is set to 0.5. In this case, there are two peak frequencies at which the value obtained by dividing the power spectrum value by Pmax is equal to or greater than the predetermined ratio Rth. 0.76 Hz and 1.48 Hz. This is because, in the acceleration in the front-rear direction in the walking of the subject's foot, the amplitude of the acceleration when the foot on the healthy side is stepped down becomes smaller, so the acceleration waveform has a walking cycle (0.67 seconds) and a cycle of two steps. It is close to the synthesized wave with (1.30 seconds), and a high peak appears at the frequency corresponding to the walking cycle (1.48 Hz) and the frequency of walking two steps (0.76 Hz). This confirms the fact that it can be read from 16 (a).

  FIG. 17 is the same as FIG. 10 and FIG. 14 described above, and shows the correspondence between the acceleration information and the frequency analysis result of the subject (lower limb dysfunction patient, left foot dysfunction) according to the third embodiment. is there. The graph of FIG. 17A corresponds to the result of frequency analysis performed on the vertical acceleration, and FIG. 17B corresponds to the result of frequency analysis performed on the vertical acceleration.

  In the peak frequency detection method in FIG. 17B, the predetermined ratio Rth is set to 0.5. In this case, there are three peak frequencies at which the value obtained by dividing the power spectrum value by Pmax is equal to or greater than the predetermined ratio Rth. 0.84 Hz, 1.46 Hz and 2.21 Hz. This means that the acceleration in the vertical direction of the subject's gait walking is the walking cycle of the right foot (0.68 seconds), the walking cycle of the left foot (0.45 seconds), and the cycle of walking 2 steps (1.19 seconds). A high peak appears in the frequency (1.46 Hz) corresponding to the walking cycle of the right foot, the frequency (2.21 Hz) corresponding to the left foot, and the frequency (0.84 Hz) in two steps of walking, which is close to the synthesized wave. The fact that it can be read from FIG.

  In FIG. 15B, similarly to the above-described FIGS. 8 and 12, there is one peak frequency detected by the peak frequency detection method, and this peak frequency just corresponds to the walking cycle. That is, the walking cycle is 1 / 1.83 = 0.55 (seconds).

  In FIG. 16B, similarly to the above-described FIG. 9 and FIG. 13, there are two peak frequencies detected by the peak frequency detection method. In this case, the smaller peak frequency due to two steps of walking. The reciprocal of the value obtained by doubling (A in the figure) is calculated as the walking cycle. That is, the walking cycle is 1 / (0.76 × 2) = 0.66 (seconds).

  In addition, as a period calculation method in FIG. 16B, the reciprocal of the larger peak frequency (B in the figure) may be used as the walking period. In this case, the walking period is 1 / 1.48 = 0.67 ( Seconds).

  FIG. 17B is the same as FIG. 10 and FIG. 14 described above, and there are three peak frequencies detected by the peak frequency detection method. In this case, the reciprocal of the value obtained by doubling the minimum peak frequency (A in the figure) due to two steps of walking is the walking cycle, that is, the walking cycle is 1 / (0.84 × 2) = 0.59 ( Seconds).

  Moreover, as a period calculation method in FIG.17 (b), except the minimum peak frequency (A in a figure), the frequency (C in a figure) resulting from the walking period of a right foot, and the frequency (C in a figure) ( The reciprocal of the value obtained by averaging B) in the figure may be used as the walking cycle. In this case, the walking cycle is 2 / (1.46 + 2.21) = 0.55 (seconds).

  As described above, in the walking cycle detection device according to the third embodiment, the peak frequency detected by the peak frequency detection method is set as a threshold value obtained by dividing the peak frequency by the maximum value of the power spectrum. Since the frequency detection unit 2 detects the frequency corresponding to the peak of the power spectrum equal to or higher than the threshold as the peak frequency, the walking cycle can be detected more accurately regardless of the walking state of the subject. Become.

(Embodiment 4)
Next, the walking cycle detection apparatus 100 according to Embodiment 4 will be described with reference to the drawings. Note that the walking cycle detection apparatus 100 according to the fourth embodiment detects the subject's displacement in the displacement detection unit using a mark or the like attached to the subject, and performs differentiation twice on the detected displacement. Acceleration is calculated, and frequency analysis is performed based on the acceleration.

FIG. 18 is a block diagram of walking cycle detection apparatus 100 in the fourth embodiment of the present invention.
The walking cycle detection apparatus 100 according to the fourth embodiment replaces the configuration of the acceleration detection unit 1 described in the first embodiment, and a displacement detection unit 31 that detects the displacement of the subject, and the displacement detection unit 31. An acceleration calculation unit 32 that performs time differentiation on the displacement detected by the above and calculates an acceleration is provided. Other processing units are the same as those in the first embodiment, and a detailed description thereof is omitted in the fourth embodiment.

The displacement detection unit 31 detects the displacement of the subject using a camera or the like, detects a displacement that occurs in the vertical direction or the front-rear direction with the movement of the human body, and outputs the detected displacement to the acceleration calculation unit 32.
The acceleration calculation unit 32 includes an input / output interface corresponding to the output of the displacement detection unit 31, a RAM that temporarily stores data, a processing unit that performs numerical operations on the displacement information, and a displacement in the vertical direction or the front-rear direction. A program for performing time differentiation on information and calculating an acceleration in the vertical direction or the front-rear direction is provided, and the result is output to the peak frequency detection unit 2. In the fourth embodiment, the peak frequency detection unit 2 includes an input / output interface corresponding to the output from the acceleration calculation unit 32.

FIG. 19 is an overall view showing a specific configuration example of the walking cycle detection device 100 according to the fourth embodiment.
The displacement detector 31 includes a marker 51 and a camera 52. The marker 51 is attached to the subject 8, and the displacement of the subject is detected by tracking the position of the marker 51 fixed to the waist of the subject 8 with the camera 52. And output to the computer 7.

  Then, the computer 7 includes an input / output interface corresponding to an output from the displacement detection unit 31, an acceleration calculation unit 32, a peak frequency detection unit 2, a cycle calculation unit 3, a storage unit 4, a display unit 5, and the walking of the present invention. A medium carrying a program for executing the cycle detection method is provided, and the walk cycle detection method based on displacement can be implemented.

  FIG. 20 is a reference showing a graph when the acceleration calculation unit 32 according to the fourth embodiment calculates acceleration from displacement by performing time differentiation twice on the displacement information output from the displacement detection unit 31. FIG.

The acceleration information calculated by the acceleration calculation unit 32 is output to the peak frequency detection unit 2. The processing after the peak frequency detection unit 2 is the same as that in the first embodiment.
As described above, in the walking cycle detection device 100 according to the fourth embodiment, the acceleration of the subject is calculated by detecting the displacement of the subject using a camera or the like and adding a time derivative thereto, It is possible to calculate the walking cycle of the subject using the same walking cycle detection method as in the above-described embodiments.

  Therefore, it is not necessary for the subject to wear an acceleration detection device including an acceleration sensor on the body, or to set the acceleration detection device, and it is possible to detect an accurate walking cycle more easily.

  In the displacement detection method 31 in the fourth embodiment, the displacement of the subject is detected by the marker and the camera. However, the position detection using the GPS, the method of marking the motion data captured from the video image, the laser is applied. Any method can be used as long as it can detect the position of the subject, such as a method of detecting the position by walking in the surrounding space.

  In the fourth embodiment, the acceleration is calculated by performing the time differentiation twice with respect to the displacement of the subject. However, when the walking speed of the subject is clear, the time differentiation is performed once for the speed information. By applying, acceleration may be calculated.

  The walking cycle detection device according to the present invention can be used as one of the motion evaluation devices for evaluating the recovery of the motor ability in the walking motion of the patient after rehabilitation, particularly for the patient who has been treated for the hip joint or the like. It can be usefully used in hospitals and clinics that manage the recovery of athletic ability.

Overall view showing a specific configuration example of the walking cycle detection device according to the first embodiment External view of a specific device of the acceleration detection device including the acceleration detection unit according to the first embodiment FIG. 3 is an external view showing an example of wearing of the acceleration detection device according to Embodiment 1 by a subject. Functional block diagram of the walking cycle detection device in the first embodiment The flowchart explaining the operation | movement procedure at the time of the walk period detection of the walk period detection apparatus which concerns on this invention. The figure which shows the table which illustrated the determination method of the threshold value of a power spectrum The figure which shows the table at the time of changing the calculation method of a walk cycle according to the number of peak frequencies in the walk cycle detection apparatus which concerns on this invention. The figure which showed the response | compatibility of a test subject's acceleration information and a frequency analysis result using the walking cycle detection method which concerns on Embodiment 1. FIG. The figure which showed the response | compatibility of the test subject's acceleration information and frequency analysis result concerning Embodiment 1 The figure which showed the response | compatibility of the test subject's acceleration information and frequency analysis result concerning Embodiment 1 The figure which shows the example of a display of the test subject's walk cycle displayed on the monitor of the walk cycle detection apparatus which concerns on Embodiment 1. FIG. The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 2 The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 2 The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 2 The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 3 The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 3 The figure which showed the response | compatibility of the acceleration information of a test subject and frequency analysis result concerning Embodiment 3 Block diagram of walking cycle detection device in embodiment 4 Overall view showing a specific configuration example of a walking cycle detection device according to Embodiment 4 Reference diagram showing a graph when the acceleration calculation unit according to the fourth embodiment calculates acceleration from displacement using the displacement information output from the displacement detection unit Reference diagram showing the conventional walking cycle detection method described in Cited Document 1

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Acceleration detection part 2 Peak frequency detection part 3 Period calculation part 4 Accumulation part 5 Display part 7 Computer 8 Test subject 10 Acceleration detection apparatus 31 Displacement detection part 32 Acceleration calculation part 51 Marker 52 Camera 100 Walking period detection apparatus 101 Buttons 102 Liquid crystal panel 103 Speaker 104 Microphone 105 Fixing band

Claims (18)

A walking cycle detection device for detecting a walking cycle during walking of a subject,
Acceleration detecting means for detecting acceleration at least in the vertical direction or the front-rear direction during walking of the subject;
Calculating means for frequency analysis of the acceleration detected by the acceleration detecting means to calculate a power spectrum;
Threshold setting means for setting a threshold for detection of the walking cycle in the calculated power spectrum;
Peak detection means for detecting a peak of the power spectrum equal to or higher than the threshold;
A walking cycle detection device comprising: a walking cycle detection unit that detects the walking cycle from a peak frequency corresponding to a peak detected by the peak detection unit. The threshold value setting means sets a half value of the maximum value of the power spectrum calculated by the calculation means as the threshold value,
The walking cycle detection device according to claim 1, wherein the peak detection unit detects a peak of a power spectrum that is equal to or greater than the threshold value. The threshold value setting means sets a difference value from the maximum value of the power spectrum calculated by the calculation means as the threshold value,
The peak detection unit detects a peak of the power spectrum when a difference value between the maximum value of the power spectrum calculated by the calculation unit and the power spectrum in each frequency band is equal to or greater than the threshold value. The walking cycle detection device according to claim 1. The threshold value setting means sets the division value divided by the maximum value of the power spectrum calculated by the calculation means as the threshold value,
The peak detection means detects a peak of the power spectrum when a division value obtained by dividing the power spectrum in each frequency band calculated by the calculation means by the maximum value of the power spectrum is equal to or greater than the threshold value. The walking period detection device according to claim 1, wherein the detection unit detects a peak of a power spectrum that is equal to or greater than the threshold value. The walking cycle detection unit according to claim 1, wherein the walking cycle detection unit changes the detection method of the walking cycle according to the number of the peak frequencies corresponding to the peaks detected by the peak detection unit. apparatus. 6. When the number of peak frequencies corresponding to the peak detected by the peak detecting means is one, the walking cycle detecting means detects a reciprocal value of the peak frequency as a walking cycle. The walking cycle detection device described. When the number of peak frequencies corresponding to the peak detected by the peak detection means is two, the walking cycle detection means calculates the reciprocal of a value twice the smaller peak frequency of the two peak frequencies. It detects as a walking cycle. The walking cycle detection apparatus of Claim 5 characterized by the above-mentioned. When the number of peak frequencies corresponding to the peak detected by the peak detection means is two, the walking cycle detection means uses the reciprocal of the larger peak frequency value of the two peak frequencies as a walking cycle. The walking cycle detection device according to claim 5, wherein the walking cycle detection device is detected. When the number of peak frequencies corresponding to the peak detected by the peak detection means is three or more, the walking cycle detection means is the reciprocal of a value twice the minimum peak frequency among the detected peak frequencies. The walking cycle detection device according to claim 5, wherein the walking cycle is detected as a walking cycle. When the number of peak frequencies corresponding to the peak detected by the peak detection means is three or more, the walking cycle detection means is an average value of peak frequencies excluding a minimum peak frequency among the detected peak frequencies. The walking cycle detection device according to claim 5, wherein the reciprocal of is detected as a walking cycle. The walking cycle detection device according to claim 1, wherein the acceleration detection unit is an acceleration sensor. The walking cycle detection device further includes:
Displacement detecting means for detecting displacement from a certain reference value including at least one axis in the vertical direction or the front-rear direction with respect to the traveling direction;
The walking cycle detection device according to claim 1, further comprising: an acceleration calculation unit that calculates the acceleration by performing time differentiation twice with respect to the displacement obtained from the displacement detection unit. The walking cycle detection device further includes:
A speed detecting means for detecting a speed including at least one axis in the vertical direction or the front-rear direction with respect to the traveling direction;
The walking cycle detection device according to claim 12, wherein the acceleration calculation means calculates the acceleration by performing time differentiation once on the speed obtained from the speed detection means. The walking cycle detection device according to claim 12, wherein the displacement detection means is a camera or a GPS. A walking cycle detection method for use in a walking cycle detection device for detecting a walking cycle during walking of a subject,
Acceleration detection step for detecting at least the acceleration in the vertical direction or the front-rear direction during walking of the subject;
A frequency analysis of the acceleration detected in the acceleration detection step to calculate a power spectrum;
A threshold setting step for setting a threshold for detection of the walking cycle in the calculated power spectrum;
A peak detection step for detecting a peak of the power spectrum equal to or greater than the threshold;
A walking cycle detection step of detecting the walking cycle from a peak frequency corresponding to the peak detected in the peak detection step. The walking cycle detection step according to claim 15, wherein, in the walking cycle detection step, the detection method of the walking cycle is changed according to the number of the peak frequencies corresponding to the peaks detected in the peak detection step. Detection method. A program used for a walking cycle detection method for detecting a walking cycle during walking of a subject,
Acceleration detection step for detecting at least the acceleration in the vertical direction or the front-rear direction during walking of the subject;
A frequency analysis of the acceleration detected in the acceleration detection step to calculate a power spectrum;
A threshold setting step for setting a threshold for detection of the walking cycle in the calculated power spectrum;
A peak detection step for detecting a peak of the power spectrum equal to or greater than the threshold;
A walking cycle detection step of detecting the walking cycle from a peak frequency corresponding to the peak detected in the peak detection step. The program according to claim 17, wherein in the walking cycle detection step, the walking cycle detection method is changed according to the number of the peak frequencies corresponding to the peaks detected in the peak detection step.

Images