Cognitive function evaluation method, cognitive function evaluation device and storage medium
By detecting the angle parameters of the ankle and knee joints during walking and combining them with a predictive model, the problem of insufficient accuracy in assessing cognitive function in existing technologies has been solved, and high-precision cognitive function assessment has been achieved.
Patent Information
- Application Number
- CN202010869111.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-02-14
- Filing Date
- 2020-08-26
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2040-08-26
AI Technical Summary
Existing technologies struggle to assess cognitive function simply and with high accuracy, particularly in assessing cognitive function in older adults' walking behavior.
By acquiring walking data from test subjects, the angles of the ankle and knee joints are detected, and these angle parameters are used to determine the degree of cognitive function, which is then evaluated in conjunction with a predictive model.
It achieves high-precision assessment of cognitive function, can reliably identify mild cognitive impairment, and does not require large devices; the assessment process is simple.
Smart Images

Figure CN112438723B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a technology for evaluating cognitive function based on the walking movements of a test subject. Background Art
[0002] In recent years, in order to understand the health status of the elderly, technology that can easily estimate body functions has been developed.
[0003] In particular, a technique for evaluating cognitive function or motor function based on parameters measured from walking performed daily is proposed.
[0004] For example, Japanese Patent Application Laid-Open No. 2013-255786 discloses a method for evaluating susceptibility to geriatric disorders (geriatric disorder risk) based on walking parameters measured during walking behavior.
[0005] Moreover, for example, International Publication No. 2018 / 066422 discloses a cognitive function assessment device, which comprises: a storage unit for storing reference data representing the relationship between the periodicity of a person's body movement when walking and the degree of the person's cognitive function; an acquisition unit for acquiring body motion data representing the detected body motion from a body motion sensor that detects the body motion of a subject when walking; and a calculation unit for calculating the periodicity of body motion when walking based on the acquired body motion data, and determining the degree of cognitive function corresponding to the calculated periodicity by comparing the reference data stored in the storage unit with the calculated periodicity.
[0006] However, it is difficult to evaluate cognitive function simply and accurately with the above-mentioned conventional technologies, and further improvement is needed. Summary of the invention
[0007] The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a technology that can evaluate cognitive function simply and with high accuracy.
[0008] One aspect of the present invention relates to a cognitive function assessment method, which is a cognitive function assessment method of a cognitive function assessment device that assesses cognitive function based on the walking movements of a test subject, acquiring walking data related to the walking of the test subject; detecting at least one of the angle of an ankle joint of one leg of the test subject and the angle of a knee joint of the one leg from the walking data; and judging the degree of cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 It is a block diagram showing the configuration of a cognitive function evaluation system according to an embodiment of the present invention.
[0010] Figure 2 This is a schematic diagram for explaining the process of extracting skeleton data from two-dimensional image data in the present embodiment.
[0011] Figure 3 It is a schematic diagram for explaining the walking cycle of this embodiment.
[0012] Figure 4 This is a flowchart for explaining the cognitive function evaluation process using the walking motion of the test subject in the present embodiment.
[0013] Figure 5 It is used to illustrate Figure 4 This is a flowchart of the cognitive function determination process of step S4.
[0014] Figure 6 It is used to illustrate Figure 4 This is a flowchart of another example of the cognitive function determination process of step S4.
[0015] Figure 7 Schematic diagram showing changes in the angle of the ankle joint of one leg during one gait cycle in the present embodiment.
[0016] Figure 8 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of this embodiment.
[0017] Fig. 9 1 is a schematic diagram showing changes in the angle of the ankle joint of one leg during one gait cycle in the first modified example of the present embodiment.
[0018] Fig.10 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the first modification example of the present embodiment.
[0019] Fig.11 This is a schematic diagram showing the average of the average values of time series data of the angle of the ankle joint of one leg of healthy people during the period from 85% to 88% of a walking cycle and the average of the average values of time series data of the angle of the ankle joint of one leg of patients with mild cognitive impairment (MCI) during the period from 85% to 88% of a walking cycle in the first variant of the present embodiment.
[0020] Fig.12 Schematic diagram showing a second modified example of the present embodiment, showing changes in the angle of a knee joint of one leg during one gait cycle.
[0021] Fig.13Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the second modification example of the present embodiment.
[0022] Fig.14 This is a schematic diagram showing changes in the angle of the knee joint of one leg during one gait cycle in a third modified example of the present embodiment.
[0023] Fig.15 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the third modification example of the present embodiment.
[0024] Fig.16 This is a schematic diagram showing the average of the average values of time series data of the angle of the knee joint of one leg of healthy persons during the period from 45% to 49% of a gait cycle and the average of the average values of time series data of the angle of the knee joint of one leg of patients with mild cognitive impairment (MCI) during the period from 45% to 49% of a gait cycle in the third variant of the present embodiment.
[0025] Fig.17 FIG. 4 is a schematic diagram showing the vertical displacement of the waist during one walking cycle in a fourth modified example of the present embodiment.
[0026] Fig.18 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fourth modification example of the present embodiment.
[0027] Fig.19 This is a schematic diagram showing changes in the angle of the ankle joint of one leg during one gait cycle in the fifth modified example of the present embodiment.
[0028] Fig. 20 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fifth modification example of the present embodiment.
[0029] Fig.21 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the sixth modification example of the present embodiment.
[0030] Fig. 22 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the seventh modification example of the present embodiment.
[0031] Fig.23Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the eighth modification example of the present embodiment.
[0032] Fig.24 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the ninth modification example of the present embodiment.
[0033] Fig.25 1 is a schematic diagram showing changes in the angle of the ankle joint of one leg of a male test subject during one gait cycle in the tenth variation of the present embodiment.
[0034] Fig.26 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the tenth modification example of the present embodiment.
[0035] Fig. 27 1 is a schematic diagram showing the change in the angle of the ankle joint of one leg of a male test subject during one gait cycle in the eleventh variation of the present embodiment.
[0036] Fig.28 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the eleventh variation of the present embodiment.
[0037] Fig.29 1 is a schematic diagram showing changes in the angle of the knee joint of one leg of a male test subject during one gait cycle in the twelfth variation of the present embodiment.
[0038] Fig.30 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the twelfth variation of the present embodiment.
[0039] Fig.31 FIG. 1 is a schematic diagram showing the vertical displacement of the waist of a male test subject in one gait cycle in the thirteenth variation of the present embodiment.
[0040] Fig.32 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the thirteenth variation of the present embodiment.
[0041] Fig.33 14 is a schematic diagram showing changes in the angle of the ankle joint of one leg of a female test subject during one gait cycle, in the fourteenth variation of the present embodiment.
[0042] Fig.34Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fourteenth variation of the present embodiment.
[0043] Fig.35 15 is a schematic diagram showing changes in the angle of the knee joint of one leg of a female test subject during one gait cycle, in the fifteenth variation of the present embodiment.
[0044] Fig.36 It is a schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fifteenth variation of the present embodiment.
[0045] Fig.37 FIG. 1 is a schematic diagram showing the vertical displacement of the waist of a female test subject in one gait cycle in the sixteenth variation of the present embodiment.
[0046] Fig.38 It is a schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the sixteenth variation of the present embodiment.
[0047] Fig.39 FIG. 1 is a schematic diagram showing the vertical displacement of the waist of a female test subject in one gait cycle in the seventeenth variation of the present embodiment.
[0048] Fig.40 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the seventeenth variation of the present embodiment.
[0049] Fig.41 It is a schematic diagram showing an example of the evaluation result screen displayed in this embodiment. DETAILED DESCRIPTION
[0050] (Basic knowledge of the present invention)
[0051] The measurement of walking parameters disclosed in Japanese Patent Publication No. 2013-255786 utilizes a sheet-type pressure sensor or a three-dimensional motion analysis system. The sheet-type pressure sensor measures the pressure distribution during walking and measures the walking parameters based on the pressure distribution. The three-dimensional motion analysis system obtains image information of tags installed on the legs from multiple cameras, and measures the walking parameters by analyzing the movements based on the image information. However, it is troublesome to set up such a sheet-type pressure sensor or a three-dimensional motion analysis system. Therefore, the technology disclosed in Japanese Patent Publication No. 2013-255786 is difficult to simply evaluate the risk of geriatric disorders.
[0052] Moreover, as walking parameters disclosed in Japanese Patent Publication No. 2013-255786, two or more can be selected from kicking, stepping, walking ratio, stride, step interval, walking angle, toe angle, left-right difference of stepping, left-right difference of step interval, left-right difference of walking angle, and left-right difference of double-foot support period. The walking angle is the angle between the straight line connecting one left heel and the other heel and the direction of travel. The toe angle is the angle between the straight line connecting the heel and the toe and the direction of travel. Moreover, the technology of Japanese Patent Publication No. 2013-255786 evaluates the risk of elderly disorders selected from at least knee pain, low back pain, urinary incontinence, dementia, and muscle wasting. However, Japanese Patent Publication No. 2013-255786 does not disclose the content of evaluating the risk of elderly disorders using walking parameters other than the above, and the evaluation accuracy of evaluating the risk of elderly disorders can be further improved by using other walking parameters.
[0053] The calculation unit in International Publication No. 2018 / 066422 determines whether the person is a normal person (NC), a mild cognitive impairment (MCI), or Alzheimer's disease (AD) based on the integrated value of the spectrum, the stride length, or the time of one step. However, International Publication No. 2018 / 066422 does not disclose the use of parameters other than the above to evaluate the degree of cognitive function. By using other walking parameters, the accuracy of evaluating cognitive function can be further improved.
[0054] In order to solve the above problems, one aspect of the present invention relates to a cognitive function assessment method, which is a cognitive function assessment method of a cognitive function assessment device that assesses cognitive function based on the walking movements of a test subject, acquiring walking data related to the walking of the test subject; detecting at least one of the angle of an ankle joint of one leg of the test subject and the angle of a knee joint of the one leg from the walking data; and judging the degree of cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint.
[0055] According to this configuration, at least one of the angle of the ankle joint of one leg and the angle of the knee joint of one leg of the walking test subject is used as a parameter related to the degree of cognitive function of the test subject. The walking motion of the test subject with reduced cognitive function has a different tendency from the walking motion of the test subject with normal cognitive function. Therefore, since the degree of cognitive function of the test subject is judged by using the parameter related to the degree of cognitive function of the test subject during walking, the cognitive function of the test subject can be evaluated with high accuracy.
[0056] Furthermore, since at least one of the angle of the ankle joint of one leg of the walking test subject and the angle of the knee joint of one leg can be easily detected from image data obtained by, for example, photographing the walking test subject, no large-scale device is required. Therefore, the present configuration can easily evaluate the cognitive function of the test subject.
[0057] Moreover, in the cognitive function assessment method, it may be that, in the detection, time series data of the angle of the ankle joint during a specified period of time when the foot of one leg is suspended is detected, and in the judgment, an average value of the time series data of the angle of the ankle joint is used to judge the degree of cognitive function of the test subject.
[0058] The angle of the ankle joint of the test subject during the prescribed period of the foot of one leg being suspended in the air is significantly different between the test subject with impaired cognitive function and the test subject without impaired cognitive function. Therefore, according to this configuration, the cognitive function of the test subject can be reliably evaluated by using the average value of the time series data of the angle of the ankle joint of the test subject during the prescribed period of the foot of one leg being suspended in the air while walking.
[0059] Moreover, in the cognitive function assessment method, the period from when one foot of the test subject touches the ground to when the foot touches the ground again may be represented as a walking cycle, and when the one walking cycle is represented by 1% to 100%, the specified period may be a period from 81% to 100% of the one walking cycle.
[0060] According to this configuration, the period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and one walking cycle is represented by 1% to 100%. In this case, by using the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of one walking cycle, the cognitive function of the test subject can be reliably evaluated.
[0061] Furthermore, in the cognitive function evaluation method, the predetermined period may be a period from 85% to 88% of the one walking cycle.
[0062] According to this configuration, by using the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle, the cognitive function of the test subject can be evaluated more reliably.
[0063] Moreover, in the cognitive function assessment method, it may be that, in the detection, time series data of a first angle of the ankle joint in a first period of the stance period of the one leg and time series data of a second angle of the ankle joint in a second period of the foot suspending period of the one leg are detected, and in the judgment, the average value of the time series data of the first angle of the ankle joint and the average value of the time series data of the second angle of the ankle joint are used to judge the degree of cognitive function of the test subject.
[0064] According to this configuration, by combining and utilizing the average value of time series data of a first angle of the ankle joint during a first period of a stance period of one leg and the average value of time series data of a second angle of the ankle joint during a second period of a foot hanging in the air on one leg, cognitive function can be evaluated with high accuracy, compared with utilizing separately the average value of time series data of the first angle of the ankle joint during a first period of a stance period of one leg and the average value of time series data of the second angle of the ankle joint during a second period of a foot hanging in the air on one leg.
[0065] Moreover, in the cognitive function assessment method, the period from when one foot of the test subject touches the ground to when the foot touches the ground again may be represented as a walking cycle, and when the one walking cycle is represented by 1% to 100%, the first period is a period from 45% to 50% of the one walking cycle, and the second period is a period from 85% to 88% of the one walking cycle.
[0066] According to this configuration, the period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one gait cycle, and one gait cycle is represented by 1% to 100%. In this case, by using the average value of the time series data of the angle of the ankle joint of one leg during 45% to 50% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle, the cognitive function of the test subject can be reliably evaluated.
[0067] Moreover, in the cognitive function assessment method, it may be that, in the detection, time series data of the angle of the knee joint during a specified period of the stance period of the one leg is detected, and in the judgment, the average value of the time series data of the angle of the knee joint is used to judge the degree of cognitive function of the test subject.
[0068] The angle of the knee joint during the predetermined period of the stance phase of one leg is significantly different between the test subjects with impaired cognitive function and the test subjects without impaired cognitive function. Therefore, according to this configuration, the cognitive function of the test subject can be reliably evaluated by using the average value of the time series data of the angle of the knee joint of the test subject during the predetermined period of the stance phase of one leg while walking.
[0069] Moreover, in the cognitive function assessment method, the period from when one foot of the test subject touches the ground to when the foot touches the ground again may be represented as a walking cycle, and when the one walking cycle is represented by 1% to 100%, the specified period may be a period from 41% to 50% of the one walking cycle.
[0070] According to this configuration, the period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and one walking cycle is represented by 1% to 100%. In this case, by using the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one walking cycle, the cognitive function of the test subject can be reliably evaluated.
[0071] Moreover, in the cognitive function assessment method, it may be that, in the detection, the time series data of the angle of the knee joint in the first period of the stance period of the one leg and the time series data of the angle of the ankle joint in the second period of the foot suspending period of the one leg are detected, and in the judgment, the average value of the time series data of the angle of the knee joint and the average value of the time series data of the angle of the ankle joint are used to judge the degree of cognitive function of the test subject.
[0072] According to this configuration, time series data of the angle of the knee joint during the first period of the stance phase of one leg and time series data of the angle of the ankle joint during the second period of the foot-hanging phase of one leg are detected. The average value of the time series data of the angle of the knee joint and the average value of the time series data of the angle of the ankle joint are used to determine the degree of cognitive function of the test subject.
[0073] Therefore, by combining the average value of the time series data of the angle of the knee joint during the first period of the stance period of one leg and the average value of the time series data of the angle of the ankle joint during the second period of the foot hanging period of one leg, cognitive function can be evaluated with higher accuracy than by separately using the average value of the time series data of the angle of the knee joint during the first period of the stance period of one leg and the average value of the time series data of the angle of the ankle joint during the second period of the foot hanging period of one leg.
[0074] Moreover, in the cognitive function assessment method, the period from when one foot of the test subject touches the ground to when the foot touches the ground again may be represented as a walking cycle, and when the one walking cycle is represented by 1% to 100%, the first period is a period from 41% to 50% of the one walking cycle, and the second period is a period from 81% to 100% of the one walking cycle.
[0075] According to this configuration, the period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one gait cycle, and one gait cycle is represented by 1% to 100%. In this case, by using the average value of the time series data of the first angle of the knee joint of one leg in the first period of 41% to 50% of one gait cycle and the average value of the time series data of the second angle of the knee joint of one leg in the second period of 81% to 100% of one gait cycle, the cognitive function of the test subject can be reliably evaluated.
[0076] Moreover, in the cognitive function assessment method, it may be that the displacement of the waist of the test subject in the vertical direction is also detected from the walking data, and in the judgment, the degree of cognitive function of the test subject is judged using at least one of the angle of the ankle joint and the angle of the knee joint and the displacement of the waist in the vertical direction.
[0077] According to this configuration, by utilizing at least one of the angle of the ankle joint and the angle of the knee joint in combination with the displacement of the waist in the vertical direction, cognitive function can be evaluated with higher accuracy than by utilizing the displacement of the waist in the vertical direction alone.
[0078] Moreover, in the cognitive function assessment method, the gender of the test subject may also be identified. When the test subject is identified as a male, in the detection, the time series data of the angle of the ankle joint at the initial stage of the stance period of one leg is detected, and in the judgment, the average value of the time series data of the angle of the ankle joint is used to judge the degree of cognitive function of the test subject.
[0079] When the test subject is a male, the angle of the ankle joint of the test subject at the initial stage of the stance phase of one leg is significantly different between the test subject with impaired cognitive function and the test subject without impaired cognitive function. Therefore, according to this configuration, by using the average value of the time series data of the angle of the ankle joint of the test subject at the initial stage of the stance phase of one leg while walking, the cognitive function of the male test subject can be reliably evaluated.
[0080] Moreover, in the cognitive function assessment method, the gender of the test subject may also be identified. When the test subject is identified as a male, in the detection, the time series data of the angle of the knee joint during the period when the foot of one leg is suspended is detected, and in the judgment, the average value of the time series data of the angle of the knee joint is used to judge the degree of cognitive function of the test subject.
[0081] When the test subject is a male, the angle of the knee joint of the test subject when the foot of one leg is suspended in the air is significantly different between the test subject with impaired cognitive function and the test subject without impaired cognitive function. Therefore, according to this configuration, by using the average value of the time series data of the angle of the knee joint of the test subject when the foot of one leg is suspended in the air, the cognitive function of the male test subject can be reliably evaluated.
[0082] Moreover, in the cognitive function assessment method, in the judgment, when the angle of the ankle joint is less than a threshold value or when the angle of the knee joint is less than a threshold value, the test subject is judged to be suffering from mild cognitive impairment.
[0083] According to this configuration, when the angle of the ankle joint is less than the threshold value or the angle of the knee joint is less than the threshold value, the test subject is judged to have mild cognitive impairment. Therefore, by comparing the angle of the ankle joint or the angle of the knee joint with the threshold value, it is possible to simply judge whether the test subject has mild cognitive impairment.
[0084] Moreover, in the cognitive function assessment method, it may be that, in the judgment, whether the test subject has mild cognitive impairment is judged by inputting at least one of the detected angles of the ankle joint and the knee joint into a prediction model, and the prediction model takes at least one of the angles of the ankle joint and the knee joint as input values and generates whether the test subject has mild cognitive impairment as an output value.
[0085] According to this configuration, the prediction model uses at least one of the angle of the ankle joint and the angle of the knee joint as an input value, and generates whether the test subject has mild cognitive impairment as an output value. Moreover, by inputting at least one of the detected angle of the ankle joint and the angle of the knee joint into the prediction model, it is determined whether the test subject has mild cognitive impairment. Therefore, by pre-storing the prediction model, it is possible to easily determine whether the test subject has mild cognitive impairment.
[0086] Another aspect of the present invention relates to a cognitive function assessment device, which assesses cognitive function based on the walking movements of a test subject, and includes: an acquisition unit, for acquiring walking data related to the walking of the test subject; a detection unit, for detecting at least one of the angle of an ankle joint of one leg of the test subject and the angle of a knee joint of the one leg from the walking data; and a judgment unit, for judging the degree of the cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint.
[0087] According to this configuration, at least one of the angle of the ankle joint of one leg and the angle of the knee joint of one leg of the walking test subject is used as a parameter related to the degree of cognitive function of the test subject. The walking motion of the test subject with reduced cognitive function has a different tendency from the walking motion of the test subject with normal cognitive function. Therefore, since the degree of cognitive function of the test subject is judged by using the parameter related to the degree of cognitive function of the test subject during walking, the cognitive function of the test subject can be evaluated with high accuracy.
[0088] Furthermore, since at least one of the angle of the ankle joint of one leg of the walking test subject and the angle of the knee joint of one leg can be easily detected from image data obtained by, for example, photographing the walking test subject, no large-scale device is required. Therefore, the present configuration can easily evaluate the cognitive function of the test subject.
[0089] Another aspect of the present invention relates to a storage medium, which is a non-temporary computer-readable storage medium that stores a cognitive function assessment program for assessing cognitive function based on the walking movements of a test subject, so that the computer has the following functions: obtaining walking data related to the walking of the test subject; detecting at least one of the angle of the ankle joint of one leg of the test subject and the angle of the knee joint of one leg from the walking data; and judging the degree of cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint.
[0090] According to this configuration, at least one of the angle of the ankle joint of one leg and the angle of the knee joint of one leg of the walking test subject is used as a parameter related to the degree of cognitive function of the test subject. The walking motion of the test subject with reduced cognitive function has a different tendency from the walking motion of the test subject with normal cognitive function. Therefore, since the degree of cognitive function of the test subject is judged by using the parameter related to the degree of cognitive function of the test subject during walking, the cognitive function of the test subject can be evaluated with high accuracy.
[0091] Furthermore, since at least one of the angle of the ankle joint of one leg of the walking test subject and the angle of the knee joint of one leg can be easily detected from image data obtained by, for example, photographing the walking test subject, no large-scale device is required. Therefore, the present configuration can easily evaluate the cognitive function of the test subject.
[0092] The embodiments of the present invention are described below with reference to the accompanying drawings. In addition, the embodiments described below are all examples of the present invention, and are not intended to limit the technical protection scope of the present invention.
[0093] (Implementation Method)
[0094] The following is based on Figure 1 The cognitive function evaluation system according to the present embodiment will be described.
[0095] Figure 1 It is a block diagram showing the configuration of a cognitive function evaluation system according to an embodiment of the present invention.
[0096] Figure 1 The illustrated cognitive function evaluation system includes a cognitive function evaluation device 1 , a camera 2 , and a display unit 3 .
[0097] The camera 2 captures the walking test subject. The camera 2 outputs dynamic image data showing the walking test subject to the cognitive function evaluation device 1. The camera 2 is connected to the cognitive function evaluation device 1 by wire or wirelessly.
[0098] The cognitive function assessment device 1 includes a processor 11 and a memory 12 .
[0099] The processor 11 is, for example, a CPU (central processing unit), and includes a data acquisition unit 111 , a walking parameter detection unit 112 , a cognitive function determination unit 113 , and an evaluation result presentation unit 114 .
[0100] The memory 12 is a storage device that can store various information, such as RAM (Random Access Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), or flash memory.
[0101] The data acquisition unit 111 acquires walking data related to the walking of the test subject. The walking data is, for example, dynamic image data obtained by photographing the walking test subject. The data acquisition unit 111 acquires dynamic image data output by the camera 2.
[0102] The walking parameter detection unit 112 extracts skeletal data representing the skeleton of the test subject from the dynamic image data acquired by the data acquisition unit 111. The skeletal data is represented by the coordinates of a plurality of feature points representing the joints of the test subject and the straight lines connecting the feature points. The walking parameter detection unit 112 may also use software (such as Open Pose or 3D pose-baseline) that detects the coordinates of the feature points of a person from two-dimensional image data.
[0103] Here, the process of extracting skeleton data from two-dimensional image data will be described.
[0104] Figure 2 This is a schematic diagram for explaining the process of extracting skeleton data from two-dimensional image data in the present embodiment.
[0105] The walking parameter detection unit 112 extracts skeleton data 21 from the two-dimensional image data 20 including the image of the walking test subject 200. The skeleton data 21 includes a feature point 201 indicating the head, a feature point 202 indicating the center of both shoulders, a feature point 203 indicating the right shoulder, a feature point 204 indicating the right elbow, a feature point 205 indicating the right hand, a feature point 206 indicating the left shoulder, a feature point 207 indicating the left elbow, a feature point 208 indicating the left hand, a feature point 209 indicating the waist, a feature point 210 indicating the right hip joint, a feature point 211 indicating the right knee joint, a feature point 212 indicating the right ankle joint, a feature point 213 indicating the toe of the right foot, a feature point 214 indicating the left hip joint, a feature point 215 indicating the left knee joint, a feature point 216 indicating the left ankle joint, and a feature point 217 indicating the toe of the left foot.
[0106] The dynamic image data is composed of a plurality of two-dimensional image data. The walking parameter detection unit 112 extracts time-series skeleton data from each of the plurality of two-dimensional image data constituting the dynamic image data. Moreover, the walking parameter detection unit 112 may extract skeleton data from the two-dimensional image data of all frames, or may extract skeleton data from the two-dimensional image data of each specified frame. Moreover, in the present embodiment, cognitive function is evaluated mainly based on the movement of the lower limbs of the test subject during walking. For this purpose, the walking parameter detection unit 112 may also extract only the skeleton data of the lower limbs of the test subject.
[0107] Then, the walking parameter detection unit 112 extracts the skeleton data corresponding to one walking cycle of the test subject from the time-series skeleton data extracted from the moving image data. The walking motion of a person is a periodic motion.
[0108] Here, the walking cycle of the test subject is described.
[0109] Figure 3 It is a schematic diagram for explaining the walking cycle of this embodiment.
[0110] like Figure 3 As shown, the period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle. Figure 3 The one walking cycle shown is the period from when the right foot of the test subject touches the ground to when the right foot touches the ground again. Moreover, one walking cycle is standardized to 1% to 100%. The period from 1% to 60% of one walking cycle is called the stance period when one foot (e.g., the right foot) is touching the ground, and the period from 61% to 100% of one walking cycle is called the foot suspension period when one foot (e.g., the right foot) is leaving the ground. One walking cycle includes the stance period and the foot suspension period. In addition, one walking cycle can also be the period from when the left foot of the test subject touches the ground to when the left foot touches the ground again.
[0111] The walking parameter detection unit 112 detects at least one of the angle of the ankle joint of one leg and the angle of the knee joint of one leg of the test subject from the walking data.
[0112] In this embodiment, the walking parameter detection unit 112 detects the angle of the ankle joint of one leg of the test subject from the walking data. The walking parameter detection unit 112 detects the angle of the ankle joint of one leg of the test subject from the time series of skeletal data corresponding to one walking cycle. Figure 2 As shown, the angle θ of the ankle joint on the arrow-shaped surface is the angle formed by the straight line connecting the feature point 212 representing the right ankle joint and the feature point 211 representing the right knee joint and the straight line connecting the feature point 212 representing the right ankle joint and the feature point 213 representing the toe of the right foot.
[0113] In particular, the walking parameter detection unit 112 detects the time series data of the angle of the ankle joint during the prescribed period of the foot suspension period of one leg. Specifically, the prescribed period is a period from 81% to 100% of one walking cycle. In addition, the prescribed period may be a period from 85% to 88% of one walking cycle. The walking parameter detection unit 112 calculates the average value of the time series data of the angle of the ankle joint during the prescribed period of the foot suspension period of one leg as the walking parameter.
[0114] In the present embodiment, since one gait cycle is the period from when the subject's right foot lands on the ground until the right foot lands on the ground again, the gait parameter detection unit 112 detects the angle θ of the ankle joint of the right leg. In the case where one gait cycle is the period from when the subject's left foot lands on the ground until the left foot lands on the ground again, the gait parameter detection unit 112 may detect the angle θ of the ankle joint of the left leg.
[0115] Furthermore, in a modification of the present embodiment, detection of the angle of the knee joint of one leg of the test subject will be described.
[0116] The cognitive function determination unit 113 determines the degree of the cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint.
[0117] In this embodiment, the cognitive function determination unit 113 uses the angle of the ankle joint to determine the degree of cognitive function of the test subject. The cognitive function determination unit 113 determines whether the test subject has mild cognitive impairment by inputting the angle of the ankle joint detected by the walking parameter detection unit 112 into a prediction model, wherein the prediction model uses the angle of the ankle joint as an input value and generates whether the test subject has mild cognitive impairment as an output value.
[0118] In addition, a modification of the present embodiment will be described in which the degree of cognitive function of a test subject is determined using the angle of a knee joint.
[0119] The memory 12 pre-stores a prediction model generated by using the angle of the ankle joint as an input value and whether the test subject has mild cognitive impairment as an output value. The prediction model is a regression model that uses whether the test subject has mild cognitive impairment as a target variable and the time series data of the angle of the ankle joint in a walking cycle as an explanatory variable. The prediction model outputs one of a value indicating that the test subject has mild cognitive impairment (e.g., 1) or a value indicating that the test subject does not have mild cognitive impairment (e.g., 0).
[0120] In particular, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint during the prescribed period of the foot suspension period of one leg. Specifically, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint of one leg during the period of 81% to 100% of one walking cycle. Furthermore, the cognitive function determination unit 113 may also determine the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint of one leg during the period of 85% to 88% of one walking cycle.
[0121] In addition, the prediction model can also be generated by machine learning. As machine learning, for example, there is teacher learning that uses teacher data that gives labels (output information) to input information to learn the relationship between input and output, untaught learning that constructs data only based on input without labels, semi-taught learning that processes both labeled and unlabeled data, and reinforcement learning that maximizes the action of reward through trial and error learning. Moreover, as specific methods of machine learning, there are neural networks (including deep learning using multi-layer neural networks), genetic algorithms, decision trees, Bayesian networks, or support vector machines (SVM), etc. In the machine learning of the present invention, any of the specific examples listed above can be used.
[0122] Furthermore, the prediction model may also output a value indicating the degree of cognitive function. The value indicating the degree of cognitive function is, for example, represented by 0.0 to 1.0. In this case, for example, the cognitive function determination unit 113 may determine that the patient does not have mild cognitive impairment when the value indicating the degree of cognitive function is less than 0.5, and may determine that the patient has a high possibility of having mild cognitive impairment when the value indicating the degree of cognitive function is greater than 0.5.
[0123] The evaluation result presenting unit 114 presents the evaluation result of the degree of cognitive function determined by the cognitive function determining unit 113. The evaluation result presenting unit 114 outputs the evaluation result determined by the cognitive function determining unit 113 to the display unit 3. The evaluation result is at least one of information indicating whether the test subject determined by the cognitive function determining unit 113 has mild cognitive impairment and an evaluation message.
[0124] The display unit 3 displays the evaluation result output from the evaluation result presenting unit 114. The display unit 3 is, for example, a liquid crystal display panel or a light emitting element.
[0125] In addition, the display unit 3 may display the transition of the value indicating the degree of cognitive function in a graph in order to compare the value indicating the degree of cognitive function determined this time with the value indicating the degree of cognitive function in the past. In addition, the value indicating the degree of cognitive function in the past may be stored in the memory 12 and read out from the memory 12.
[0126] In addition, the cognitive function evaluation device 1 may include a camera 2 and a display unit 3. Furthermore, the cognitive function evaluation device 1 may include a display unit 3. The cognitive function evaluation device 1 may be a personal computer or a server.
[0127] Secondly, use Figure 4 The cognitive function evaluation process according to the present embodiment will be described.
[0128] Figure 4 This is a flowchart for explaining the cognitive function evaluation process using the walking motion of the test subject in the present embodiment. Figure 4 The flowchart shown shows a method of evaluating the degree of cognitive function using the cognitive function evaluating device 1 .
[0129] The test subject walks in front of the camera 2. The camera 2 captures the test subject walking. The camera 2 transmits dynamic image data of the test subject walking to the cognitive function evaluation device 1.
[0130] First, in step S1 , the data acquisition unit 111 acquires the moving image data transmitted from the camera 2 .
[0131] Next, in step S2 , the walking parameter detection unit 112 extracts time-series skeleton data from the moving image data.
[0132] Next, in step S3, the walking parameter detection unit 112 detects walking parameters for determining the degree of cognitive function based on the time series skeletal data. Here, the walking parameter of this embodiment is the average value of the time series data of the angle of the ankle joint of one leg of the test subject during a specified period of a walking cycle. The specified period is, for example, a period from 81% to 100% of a walking cycle. In addition, the method for determining the walking parameter will be described later.
[0133] Next, in step S4, the cognitive function determination unit 113 executes a cognitive function determination process for determining the degree of the cognitive function of the test subject using the walking parameter. The cognitive function determination process will be described later.
[0134] Next, in step S5, the evaluation result presentation unit 114 outputs the evaluation result of the degree of cognitive function determined by the cognitive function determination unit 113 to the display unit 3. The evaluation result of the degree of cognitive function indicates whether the test subject has mild cognitive impairment. In addition, the evaluation result presentation unit 114 may output not only the degree of cognitive function but also an evaluation message corresponding to the degree of cognitive function to the display unit 3. The display unit 3 displays the evaluation result of the degree of cognitive function outputted from the evaluation result presentation unit 114.
[0135] Here, Figure 4 The cognitive function determination process of step S4 will be described.
[0136] Figure 5 It is used to illustrate Figure 4 This is a flowchart of the cognitive function determination process of step S4.
[0137] First, in step S11 , the cognitive function determination unit 113 reads the prediction model from the memory 12 .
[0138] Next, in step S12, the cognitive function determination unit 113 inputs the walking parameter detected by the walking parameter detection unit 112 into the prediction model. The walking parameter of this embodiment is the average value of the time series data of the angle of the ankle joint of one leg of the test subject during 81% to 100% of one walking cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the ankle joint of one leg of the test subject during 81% to 100% of one walking cycle into the prediction model.
[0139] Next, in step S13, the cognitive function determination unit 113 obtains the determination result of the degree of cognitive function from the prediction model. The cognitive function determination unit 113 obtains whether the test subject has mild cognitive impairment as the determination result from the prediction model.
[0140] In addition, the cognitive function determination process of this embodiment determines the degree of cognitive function by inputting walking parameters into a pre-generated prediction model, but the present invention is not limited to this. In other examples of the cognitive function determination process of this embodiment, the degree of cognitive function can also be determined by comparing a pre-stored threshold value and walking parameters.
[0141] In this case, the memory 12 stores in advance a threshold value for determining whether the test subject has mild cognitive impairment.
[0142] Moreover, the cognitive function judging unit 113 judges that the test subject has mild cognitive impairment when the angle of the ankle joint is less than a threshold value. In the present embodiment, the cognitive function judging unit 113 judges whether the average value of the time series data of the angle of the ankle joint of one leg of the test subject during the period from 81% to 100% of a walking cycle is less than a threshold value. The cognitive function judging unit 113 judges that the test subject has mild cognitive impairment when the average value of the time series data of the angle of the ankle joint of one leg of the test subject during the period from 81% to 100% of a walking cycle is less than a threshold value. On the other hand, the cognitive function judging unit 113 judges that the test subject does not have mild cognitive impairment, that is, the test subject is a healthy person, when the average value of the time series data of the angle of the ankle joint of one leg of the test subject during the period from 81% to 100% of a walking cycle is greater than a threshold value.
[0143] Figure 6 It is used to illustrate Figure 4 This is a flowchart of another example of the cognitive function determination process of step S4.
[0144] First, in step S21 , the cognitive function determination unit 113 reads the threshold value from the memory 12 .
[0145] Next, in step S22, the cognitive function determination unit 113 determines whether the walking parameter detected by the walking parameter detection unit 112 is less than a threshold value. The walking parameter of this embodiment is the average value of the time series data of the angle of the ankle joint of one leg of the test subject during the period from 81% to 100% of one walking cycle. The cognitive function determination unit 113 determines whether the average value of the time series data of the angle of the ankle joint of one leg of the test subject during the period from 81% to 100% of one walking cycle is less than a threshold value.
[0146] Here, when it is determined that the walking parameter is smaller than the threshold value (YES in step S22 ), the cognitive function determination unit 113 determines in step S23 that the test subject has mild cognitive impairment.
[0147] On the other hand, when it is determined that the walking parameter is equal to or greater than the threshold value (No in step S22), in step S24, the cognitive function determination unit 113 determines that the test subject does not have mild cognitive impairment, that is, the test subject is a healthy person.
[0148] Thus, in this embodiment, the angle of the ankle joint of one leg of the walking test subject is a parameter related to the degree of cognitive function of the test subject. The walking motion of the test subject with low cognitive function has a different tendency from the walking motion of the test subject with normal cognitive function. Therefore, since the degree of cognitive function of the test subject is judged by using the parameter related to the degree of cognitive function of the test subject during walking, the cognitive function of the test subject can be evaluated with high accuracy.
[0149] Furthermore, since the angle of the ankle joint of one leg of the walking test subject can be easily detected from image data obtained by, for example, photographing the walking test subject, no large-scale device is required. Therefore, the configuration of this embodiment can easily evaluate the cognitive function of the test subject.
[0150] In addition, the cognitive function determination unit 113 may also determine whether the difference between the maximum and minimum values of the angle of the ankle joint of one leg of the test subject during the foot-in-the-air period is less than a threshold value. The cognitive function determination unit 113 may determine that the test subject has mild cognitive impairment when the difference between the maximum and minimum values of the angle of the ankle joint of one leg of the test subject during the foot-in-the-air period is less than a threshold value. On the other hand, the cognitive function determination unit 113 may determine that the test subject does not have mild cognitive impairment, that is, the test subject is a healthy person, when the difference between the maximum and minimum values of the angle of the ankle joint of one leg of the test subject during the foot-in-the-air period is greater than a threshold value.
[0151] The walking parameters and prediction model of this embodiment can be determined by experiments. Hereinafter, a method for determining the walking parameters and prediction model of this embodiment will be described.
[0152] The total number of subjects who participated in the experiment was 92. There were 27 male subjects and 65 female subjects. The subjects were screened for dementia, namely the Mini Mental State Examination (MMSE). Subjects with MMSE scores of 27 to 30 were judged to be healthy, and subjects with MMSE scores of 22 to 26 were judged to be suspected of mild cognitive impairment. In this experiment, patients with suspected mild cognitive impairment were regarded as patients with mild cognitive impairment. According to the results of MMSE, there were 10 patients with mild cognitive impairment and 82 healthy subjects among the subjects. Among the patients with mild cognitive impairment, there were 4 male patients and 6 female patients. During the experiment, the subjects walked in front of the camera. The walking subjects were filmed with a camera, and the skeletal data of each subject was extracted from the dynamic image data. In addition, the time series data of the angle of the ankle joint of one leg of each subject was detected from the extracted skeletal data.
[0153] Figure 7 Schematic diagram showing the change in angle of the ankle joint of one leg during one walking cycle in this embodiment. Figure 7 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Figure 7 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of healthy subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of patients with mild cognitive impairment.
[0154] In the experiment, a standardized walking cycle was divided into ten intervals, and the average angle of the ankle joint of one leg in one interval or two or more consecutive intervals was calculated for each test subject. In addition, multiple prediction models were created with whether the test subject had mild cognitive impairment as the target variable and the average angle of the ankle joint of one leg in one interval or two or more consecutive intervals as the explanatory variable. Multiple prediction models were evaluated by cross-validation. Leave-one-out cross-validation was used as cross-validation. In addition, the ROC (Receiver Operating Characteristic) curves of each of the multiple prediction models were calculated. Furthermore, the AUC (Area Under Curve) value of the ROC curve of each of the multiple prediction models was calculated, and the prediction model with the highest AUC value was selected.
[0155] In the present embodiment, the prediction model created by using the average value of the angle of the ankle joint of one leg during the period from 81% to 100% of one gait cycle as an explanatory variable has the highest AUC value.
[0156] Figure 8 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of this embodiment.
[0157] The prediction model of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as the explanatory variable. Figure 8 In the figure, the vertical axis represents the true positive rate (True Positive Rate) and the horizontal axis represents the false positive rate (False Positive Rate). The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0158] Figure 8 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created by using the average value of the angle of the ankle joint during 81% to 100% of a walking cycle as an explanatory variable. Figure 8 The AUC value of the ROC curve shown is 0.660. The AUC value is the area of the lower half of the ROC curve. It can be said that the larger the AUC value (the closer to 1), the higher the performance of the prediction model. In this case, the average value of the angle of the ankle joint during 81% to 100% of a walking cycle is determined as the walking parameter. Moreover, the prediction model created using the average value of the angle of the ankle joint during 81% to 100% of a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function judgment unit 113.
[0159] The memory 12 stores in advance a prediction model, which uses the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of a gait cycle as an input value, and generates whether the test subject has mild cognitive impairment as an output value. The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of a gait cycle. The cognitive function judgment unit 113 inputs the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of a gait cycle into the prediction model, and obtains a judgment result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0160] Moreover, in Figure 7As shown in FIG. 1 , the time series data of the angle of the ankle joint of one leg of the patients with mild cognitive impairment during 81% to 100% of one gait cycle is smaller than the time series data of the angle of the ankle joint of one leg of the healthy persons. For this purpose, the difference between the average of the time series data of the angle of the ankle joint of one leg of the patients with mild cognitive impairment during 81% to 100% of one gait cycle and the average of the time series data of the angle of the ankle joint of one leg of the healthy persons during 81% to 100% of one gait cycle obtained by the experiment may be stored in the memory 12 as a threshold. The cognitive function determination unit 113 may also determine the degree of cognitive function by comparing the average of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle with a pre-stored threshold.
[0161] In this embodiment, the walking parameter is the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one walking cycle, but the present invention is not limited thereto. Various examples of the walking parameter of this embodiment are described below.
[0162] First, walking parameters according to a first modified example of the present embodiment will be described.
[0163] The walking parameter of the first modified example of the present embodiment may be an average value of time series data of the angle of the ankle joint of one leg during a period from 85% to 88% of one walking cycle.
[0164] Fig. 9 Schematic diagram showing the change in angle of the ankle joint of one leg in one walking cycle in the first variant of the present embodiment. Fig. 9 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Fig. 9 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of healthy subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of patients with mild cognitive impairment.
[0165] The first variant of the present embodiment is the same as the above-mentioned experiment, and the time series data of the angle of the ankle joint of one leg of each of the multiple test subjects is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the ankle joint of one leg during 85% to 88% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0166] Fig.10 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the first modification example of the present embodiment.
[0167] The prediction model of the first variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the ankle joint of one leg during 85% to 88% of a walking cycle as the explanatory variable. Fig.10 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0168] Fig.10 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the angle of the ankle joint during 85% to 88% of one walking cycle as an explanatory variable. Fig.10 The AUC value of the ROC curve shown is 0.6866. In this case, the average value of the angle of the ankle joint during 85% to 88% of a gait cycle is determined as the walking parameter. Furthermore, the prediction model created using the average value of the angle of the ankle joint during 85% to 88% of a gait cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0169] The memory 12 stores in advance a prediction model generated by using the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0170] The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint of one leg during 85% to 88% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one walking cycle.
[0171] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle into the prediction model, and obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0172] Fig.11 This is a schematic diagram showing the average of the time series data of the angle of the ankle joint of one leg of healthy persons during the period from 85% to 88% of a gait cycle and the average of the time series data of the angle of the ankle joint of one leg of patients with mild cognitive impairment (MCI) during the period from 85% to 88% of a gait cycle in the first variant of the present embodiment.
[0173] like Fig.11 As shown, the average value of the time series data of the angle of the ankle joint of one leg of healthy people during 85% to 88% of a walking cycle is 13.2 degrees, and the average value of the time series data of the angle of the ankle joint of one leg of patients with mild cognitive impairment (MCI) during 85% to 88% of a walking cycle is 9.5 degrees.
[0174] Thus, the average of the time series data of the angle of the ankle joint of one leg of the patients with mild cognitive impairment during 85% to 88% of one gait cycle is smaller than the average of the time series data of the angle of the ankle joint of one leg of the healthy subjects. For this reason, a value between the average of the time series data of the angle of the ankle joint of one leg of the patients with mild cognitive impairment during 85% to 88% of one gait cycle and the average of the time series data of the angle of the ankle joint of one leg of the healthy subjects during 85% to 88% of one gait cycle obtained by the experiment may be stored in the memory 12 as a threshold. The cognitive function determination unit 113 may also determine the degree of cognitive function by comparing the average of the time series data of the angle of the ankle joint of one leg of the test subject during 85% to 88% of one gait cycle with a pre-stored threshold.
[0175] Next, walking parameters according to a second modified example of the present embodiment will be described.
[0176] The walking parameter of the second modified example of the present embodiment may be an average value of time series data of the angle of the knee joint of one leg during a period from 41% to 50% of one walking cycle.
[0177] Fig.12 Schematic diagram showing the change in the angle of the knee joint of one leg in one walking cycle in the second variant of the present embodiment. Fig.12 In the figure, the vertical axis represents the angle of the knee joint, and the horizontal axis represents a standardized walking cycle. Fig.12 In FIG. 1 , the dotted line represents the average waveform of the angle of the knee joint of one leg of healthy subjects, and the solid line represents the average waveform of the angle of the knee joint of one leg of patients with mild cognitive impairment.
[0178] The second variant of this embodiment is the same as the above-mentioned experiment, and the time series data of the knee joint angle of each leg of multiple test subjects is detected from the skeletal data of multiple test subjects including healthy people and patients with mild cognitive impairment. Figure 2 As shown, the angle γ of the knee joint is the angle formed between the straight line connecting the feature point 211 representing the right knee joint and the feature point 210 representing the right hip joint on the arrow-shaped plane and the straight line connecting the feature point 211 representing the right knee joint and the feature point 212 representing the right ankle joint.
[0179] In the experiment, a standardized walking cycle was divided into ten intervals, and the average angle of the knee joint of one leg in one interval or two or more consecutive intervals was calculated for each test subject. Then, multiple prediction models were created with whether the test subject had mild cognitive impairment as the target variable and the average angle of the knee joint of one leg in one interval or two or more consecutive intervals as the explanatory variable. Multiple prediction models were evaluated by cross-validation. Leave-one-out cross-validation was used as cross-validation. Then, the ROC curves of each of the multiple prediction models were calculated. In addition, the AUC values of the ROC curves of the multiple prediction models were calculated respectively, and the prediction model with the highest AUC value was selected.
[0180] In the second modified example of the present embodiment, the prediction model created using the average value of the angle of the knee joint of one leg during the period from 41% to 50% of one gait cycle as an explanatory variable has the highest AUC value.
[0181] Fig.13 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the second modification example of the present embodiment.
[0182] The prediction model of the second variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as the explanatory variable. Fig.13 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0183] Fig.13 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the angle of the knee joint during 41% to 50% of one walking cycle as an explanatory variable. Fig.13The AUC value of the ROC curve shown is 0.682. In this case, the average value of the angle of the knee joint during 41% to 50% of one gait cycle is determined as the walking parameter. Furthermore, the prediction model created using the average value of the angle of the knee joint during 41% to 50% of one gait cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0184] The walking parameter detection unit 112 detects the angle of the knee joint of one leg of the test subject from the walking data. The walking parameter detection unit 112 detects the angle of the knee joint of one leg of the test subject from the time series of skeletal data corresponding to a captured walking cycle. In particular, the walking parameter detection unit 112 detects the time series data of the angle of the knee joint during a specified period of the stance phase of one leg. Specifically, the specified period is a period from 41% to 50% of a walking cycle. The walking parameter detection unit 112 detects the time series data of the angle of the knee joint of one leg during a period from 41% to 50% of a walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the knee joint of one leg during a period from 41% to 50% of a walking cycle.
[0185] In the second modified example of the present embodiment, since one gait cycle is the period from when the subject's right foot lands on the ground to when the right foot lands again, the gait parameter detection unit 112 detects the angle γ of the knee joint of the right leg. In the case where one gait cycle is the period from when the subject's left foot lands on the ground to when the left foot lands again, the gait parameter detection unit 112 may detect the angle γ of the knee joint of the left leg.
[0186] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the angle of the knee joint. The cognitive function determination unit 113 determines whether the test subject has mild cognitive impairment by inputting the angle of the knee joint detected by the walking parameter detection unit 112 into a prediction model, which uses the angle of the knee joint as an input value and generates whether the test subject has mild cognitive impairment as an output value.
[0187] The memory 12 pre-stores a prediction model generated by using the angle of the knee joint as an input value and whether the test subject has mild cognitive impairment as an output value. The prediction model is a regression model that uses whether the test subject has mild cognitive impairment as a target variable and the time series data of the angle of the knee joint in a walking cycle as an explanatory variable. In particular, the memory 12 pre-stores a prediction model generated by using the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0188] Furthermore, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint during the prescribed period of the stance phase of one leg. Specifically, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one gait cycle into the prediction model, and obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0189] Moreover, in Fig.12 In the period from 41% to 50% of a walking cycle shown in FIG. 1 , the average waveform of the angle of the knee joint of one leg of the patients with mild cognitive impairment is smaller than the average waveform of the angle of the knee joint of one leg of the healthy subjects. For this reason, a value between the average of the time series data of the angle of the knee joint of one leg of the patients with mild cognitive impairment obtained through the experiment and the average of the time series data of the angle of the knee joint of one leg of the healthy subjects during the period from 41% to 50% of a walking cycle can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the angle of the knee joint of one leg of the test subject during the period from 41% to 50% of a walking cycle with a pre-stored threshold.
[0190] Next, walking parameters according to a third modified example of the present embodiment will be described.
[0191] The walking parameter of the third modified example of the present embodiment may be an average value of time series data of the angle of the knee joint of one leg during a period from 45% to 49% of one walking cycle.
[0192] Fig.14 Schematic diagram showing the change in angle of the knee joint of one leg during one walking cycle in the third variant of the present embodiment. Fig.14 In the figure, the vertical axis represents the angle of the knee joint, and the horizontal axis represents a standardized walking cycle. Fig.14 In FIG. 1 , the dotted line represents the average waveform of the angle of the knee joint of one leg of healthy subjects, and the solid line represents the average waveform of the angle of the knee joint of one leg of patients with mild cognitive impairment.
[0193] The third variant of the present embodiment is the same as the above-mentioned experiment, and the time series data of the angle of the knee joint of one leg of each of the multiple test subjects is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the knee joint of one leg during 45% to 49% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of each prediction model is calculated.
[0194] Fig.15 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the third modification example of the present embodiment.
[0195] The prediction model of the third variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the knee joint of one leg during 45% to 49% of a walking cycle as the explanatory variable. Fig.15 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0196] Fig.15 The ROC curve shown plots the true positive rate and false positive rate of a prediction model created using the average value of the angle of the knee joint during 45% to 49% of a walking cycle as an explanatory variable. Fig.15 The AUC value of the ROC curve shown is 0.5829. In this case, the average value of the angle of the knee joint during 45% to 49% of one gait cycle is determined as the walking parameter. Furthermore, the prediction model created using the average value of the angle of the knee joint during 45% to 49% of one gait cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0197] The memory 12 stores in advance a prediction model generated by using the average value of the time series data of the angle of the knee joint of one leg during 45% to 49% of one gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0198] The walking parameter detection unit 112 detects the time series data of the angle of the knee joint of one leg during 45% to 49% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the knee joint of one leg during 45% to 49% of one walking cycle.
[0199] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint of one leg during 45% to 49% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the knee joint of one leg during 45% to 49% of one gait cycle into the prediction model, and obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0200] Fig.16 This is a schematic diagram showing the average of the average values of time series data of the angle of the knee joint of one leg of healthy persons during the period from 45% to 49% of a gait cycle and the average of the average values of time series data of the angle of the knee joint of one leg of patients with mild cognitive impairment (MCI) during the period from 45% to 49% of a gait cycle in the third variant of the present embodiment.
[0201] like Fig.16 As shown, the average value of the time series data of the angle of the knee joint of one leg of healthy people during 45% to 49% of a gait cycle was 27.9 degrees, and the average value of the time series data of the angle of the knee joint of one leg of patients with mild cognitive impairment (MCI) during 45% to 49% of a gait cycle was 23.1 degrees.
[0202] As described above, during the period of 45% to 49% of a walking cycle, the average value of the time series data of the angle of the knee joint of one leg of the patients with mild cognitive impairment is smaller than the average value of the time series data of the angle of the knee joint of one leg of the healthy subjects. For this reason, a value between the average value of the time series data of the angle of the knee joint of one leg of the patients with mild cognitive impairment obtained by the experiment during the period of 45% to 49% of a walking cycle and the average value of the time series data of the angle of the knee joint of one leg of the healthy subjects during the period of 45% to 49% of a walking cycle can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average value of the time series data of the angle of the knee joint of one leg of the test subject during the period of 45% to 49% of a walking cycle with a pre-stored threshold.
[0203] Next, walking parameters according to a fourth modified example of the present embodiment will be described.
[0204] The walking parameter of the fourth modified example of the present embodiment may be an average value of time series data of the vertical displacement of the waist during a period from 21% to 30% of one walking cycle.
[0205] Fig.17 FIG. 4 is a schematic diagram showing the vertical displacement of the waist in one walking cycle in the fourth variation of the present embodiment. Fig.17 In the figure, the vertical axis represents the vertical displacement of the waist, and the horizontal axis represents a standardized walking cycle. Fig.17 In FIG. 1 , the dotted line represents the average waveform of the vertical displacement of the waist of healthy subjects, and the solid line represents the average waveform of the vertical displacement of the waist of patients with mild cognitive impairment.
[0206] The fourth variant of this embodiment is the same as the above-mentioned experiment, and detects the time series data of the vertical displacement of the waist of each of the multiple test subjects from the skeletal data of multiple test subjects including healthy people and patients with mild cognitive impairment. Figure 2 As shown, the displacement α of the waist in the vertical direction represents the displacement of the feature point 209 of the waist in the vertical direction.
[0207] In the experiment, a standardized walking cycle was divided into ten intervals, and the average displacement of the waist in the vertical direction in one interval or two or more consecutive intervals was calculated for each test subject. Then, multiple prediction models were created with whether the test subject had mild cognitive impairment as the target variable and the average displacement of the waist in the vertical direction in one interval or two or more consecutive intervals as the explanatory variable. Multiple prediction models were evaluated by cross-validation. Leave-one-out cross-validation was used as cross-validation. Then, the ROC curves of each of the multiple prediction models were calculated. In addition, the AUC values of the ROC curves of the multiple prediction models were calculated respectively, and the prediction model with the highest AUC value was selected.
[0208] In the fourth modified example of the present embodiment, the prediction model created using the average value of the vertical displacement of the waist during the period from 21% to 30% of one gait cycle as an explanatory variable has the highest AUC value.
[0209] Fig.18 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fourth modification example of the present embodiment.
[0210] The prediction model of the fourth variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle as the explanatory variable. Fig.18 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0211] Fig.18 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the displacement of the waist in the vertical direction during 21% to 30% of one walking cycle as an explanatory variable. Fig.18 The AUC value of the ROC curve shown is 0.587. In this case, the average value of the vertical displacement of the waist during 21% to 30% of one gait cycle is determined as the gait parameter. Furthermore, the prediction model created using the average value of the vertical displacement of the waist during 21% to 30% of one gait cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0212] The walking parameter detection unit 112 detects the displacement of the waist of the test subject in the vertical direction from the walking data. The walking parameter detection unit 112 detects the displacement of the waist of the test subject in the vertical direction from the skeleton data of the time series corresponding to a captured walking cycle. In particular, the walking parameter detection unit 112 detects the time series data of the displacement of the waist in the vertical direction during a specified period of the stance period of one leg. Specifically, the specified period is a period from 21% to 30% of a walking cycle. The walking parameter detection unit 112 detects the time series data of the displacement of the waist in the vertical direction during a period from 21% to 30% of a walking cycle. Moreover, the walking parameter detection unit 112 calculates the average value of the time series data of the displacement of the waist in the vertical direction during a period from 21% to 30% of a walking cycle.
[0213] In the fourth modified example of the present embodiment, since one gait cycle is the period from when the subject's right foot lands on the ground until the right foot lands on the ground again, the gait parameter detection unit 112 detects the vertical displacement α of the waist when the right foot is in the stance phase. In the case where one gait cycle is the period from when the subject's left foot lands on the ground until the left foot lands on the ground again, the gait parameter detection unit 112 may detect the vertical displacement α of the waist when the left foot is in the stance phase.
[0214] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the displacement of the waist in the vertical direction. The cognitive function determination unit 113 determines whether the test subject has mild cognitive impairment by inputting the displacement of the waist in the vertical direction detected by the walking parameter detection unit 112 into a prediction model, which uses the displacement of the waist in the vertical direction as an input value and generates whether the test subject has mild cognitive impairment as an output value.
[0215] The memory 12 pre-stores a prediction model generated by using the displacement of the waist in the vertical direction as an input value and whether the test subject has mild cognitive impairment as an output value. The prediction model is a regression model that uses whether the test subject has mild cognitive impairment as a target variable and the time series data of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle as an explanatory variable. In particular, the memory 12 pre-stores the prediction model generated by using the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0216] Furthermore, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the displacement of the waist in the vertical direction during the predetermined period of the stance period of one leg. Specifically, the cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the displacement of the waist in the vertical direction during the period of 21% to 30% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the displacement of the waist in the vertical direction during the period of 21% to 30% of one gait cycle into the prediction model, and obtains the judgment result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0217] Moreover, in Fig.17 In the period of 21% to 30% of one walking cycle, the average waveform of the vertical displacement of the waist of the patients with mild cognitive impairment is larger than the average waveform of the vertical displacement of the waist of the healthy subjects. For this reason, the value between the average of the time series data of the vertical displacement of the waist of the patients with mild cognitive impairment obtained by the experiment and the average of the time series data of the vertical displacement of the waist of the healthy subjects during the period of 21% to 30% of one walking cycle can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the vertical displacement of the waist of the test subject during the period of 21% to 30% of one walking cycle with the pre-stored threshold.
[0218] Next, walking parameters according to a fifth modified example of the present embodiment will be described.
[0219] The walking parameter of the fifth variant of the present embodiment may also be the average value of the time series data of the first angle of the ankle joint in the first period of the stance period of one leg and the average value of the time series data of the second angle of the ankle joint in the second period of the foot suspension period of one leg.
[0220] Fig.19FIG. 5 is a schematic diagram showing the change in angle of the ankle joint of one leg during one walking cycle in the fifth variation of the present embodiment. Fig.19 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Fig.19 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of healthy subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of patients with mild cognitive impairment.
[0221] The fifth variant of the present embodiment is the same as the above-mentioned experiment, and the time series data of the angle of the ankle joint of one leg of each of the multiple test subjects is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable, and takes the average value of the angle of the ankle joint of one leg during 45% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 85% to 88% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0222] Fig. 20 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fifth modification example of the present embodiment.
[0223] The prediction model of the fifth variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the angle of the ankle joint of one leg during 45% to 50% of a gait cycle and the average value of the angle of the ankle joint of one leg during 85% to 88% of a gait cycle as explanatory variables. Fig. 20 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0224] Fig. 20 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the angle of the ankle joint of one leg during 45% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 85% to 88% of a walking cycle as explanatory variables. Fig. 20The AUC value of the ROC curve shown is 0.7024. In this case, the average value of the angle of the ankle joint of one leg during 45% to 50% of a gait cycle and the average value of the angle of the ankle joint of one leg during 85% to 88% of a gait cycle are determined as walking parameters. Furthermore, a prediction model created using the average value of the angle of the ankle joint of one leg during 45% to 50% of a gait cycle and the average value of the angle of the ankle joint of one leg during 85% to 88% of a gait cycle as explanatory variables is determined as the prediction model used by the cognitive function determination unit 113.
[0225] The walking parameter detection unit 112 detects time series data of a first angle of the ankle joint in a first period of a stance period of one leg and time series data of a second angle of the ankle joint in a second period of a foot-free period of one leg. The first period is a period from 45% to 50% of one walking cycle, and the second period is a period from 85% to 88% of one walking cycle. The walking parameter detection unit 112 detects time series data of the angle of the ankle joint of one leg in a period from 45% to 50% of one walking cycle and time series data of the angle of the ankle joint of one leg in a period from 85% to 88% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates an average value of the time series data of the angle of the ankle joint of one leg in a period from 45% to 50% of one walking cycle and an average value of the time series data of the angle of the ankle joint of one leg in a period from 85% to 88% of one walking cycle.
[0226] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the first angle and the average value of the time series data of the second angle of the ankle joint.
[0227] The memory 12 stores in advance a prediction model that uses as inputs an average value of time series data of a first angle of an ankle joint in a first period of a stance period of one leg and an average value of time series data of a second angle of an ankle joint in a second period of a foot-free period of one leg, and outputs whether the test subject has mild cognitive impairment. The memory 12 stores in advance a prediction model that uses as inputs an average value of angles of an ankle joint of one leg in a period of 45% to 50% of a gait cycle and an average value of angles of an ankle joint of one leg in a period of 85% to 88% of a gait cycle, and outputs whether the test subject has mild cognitive impairment.
[0228] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint of one leg during 45% to 50% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the angle of the ankle joint of one leg during 45% to 50% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 85% to 88% of one gait cycle into the prediction model.
[0229] Thus, the AUC value of the prediction model created using the average value of the ankle joint angle in one period alone was 0.660, while the AUC value of the prediction model created using the average value of the ankle joint angle in two periods was 0.7024. Therefore, compared with the prediction model created using the average value of the ankle joint angle in one period alone, the prediction model created using the average value of the ankle joint angle in two periods can determine the degree of cognitive function with higher accuracy.
[0230] Next, walking parameters according to a sixth modified example of the present embodiment will be described.
[0231] The walking parameter of the sixth variant of the present embodiment may also be the average value of the time series data of the angle of the knee joint during the first period of the stance phase of one leg and the average value of the time series data of the angle of the ankle joint during the second period of the foot suspension phase of one leg.
[0232] The sixth variant of the present embodiment is the same as the above-mentioned experiment, and the time series data of the angle of the knee joint of one leg of each of the multiple test subjects and the time series data of the angle of the ankle joint of one leg of each of the multiple test subjects are detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable, and takes the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0233] Fig.21 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the sixth modification example of the present embodiment.
[0234] The prediction model of the sixth variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the angle of the knee joint of one leg during 41% to 50% of a gait cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a gait cycle as explanatory variables. Fig.21 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0235] Fig.21 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. Fig.21 The AUC value of the ROC curve shown is 0.734. In this case, the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle are determined as walking parameters. Moreover, the prediction model created using the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0236] The walking parameter detection unit 112 detects time series data of the angle of the knee joint in a first period of the stance period of one leg and time series data of the angle of the ankle joint in a second period of the foot suspension period of one leg. The first period is a period from 41% to 50% of one walking cycle, and the second period is a period from 81% to 100% of one walking cycle. The walking parameter detection unit 112 detects time series data of the angle of the knee joint of one leg in a period from 41% to 50% of one walking cycle and time series data of the angle of the ankle joint of one leg in a period from 81% to 100% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the knee joint of one leg in a period from 41% to 50% of one walking cycle and the average value of the time series data of the angle of the ankle joint of one leg in a period from 81% to 100% of one walking cycle.
[0237] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the knee joint angle and the average value of the time series data of the ankle joint angle.
[0238] The memory 12 stores in advance a prediction model that uses the average value of the time series data of the angle of the knee joint in the first period of the stance period of one leg and the average value of the time series data of the angle of the ankle joint in the second period of the foot suspension period of one leg as input values, and generates as an output value whether the test subject has mild cognitive impairment. The memory 12 stores in advance a prediction model that uses the average value of the angle of the knee joint of one leg in the period of 41% to 50% of one gait cycle and the average value of the angle of the ankle joint of one leg in the period of 81% to 100% of one gait cycle as input values, and generates as an output value whether the test subject has mild cognitive impairment.
[0239] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle into the prediction model, and obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0240] Thus, the AUC values of the prediction models created using the ankle joint angle and the knee joint angle alone were 0.660 and 0.682, respectively, while the AUC value of the prediction model created using the ankle joint angle and the knee joint angle was 0.734. Therefore, compared with the prediction model created using the ankle joint angle and the knee joint angle alone, the prediction model created using the ankle joint angle and the knee joint angle can determine the degree of cognitive function with higher accuracy.
[0241] Next, walking parameters according to a seventh modified example of the present embodiment will be described.
[0242] The walking parameter of the seventh variant of the present embodiment may also be the average value of the time series data of the vertical displacement of the waist during the first period of the stance period of one leg and the average value of the time series data of the angle of the ankle joint during the second period of the foot suspending period of one leg.
[0243] The seventh variant of the present embodiment is the same as the above-mentioned experiment, and detects the time series data of the vertical displacement of the waist of each of the multiple test subjects and the time series data of the angle of the ankle joint of each of the multiple test subjects. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable, and takes the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0244] Fig. 22 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the seventh modification example of the present embodiment.
[0245] The prediction model of the seventh variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. Fig. 22 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0246] Fig. 22 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. Fig. 22 The AUC value of the ROC curve shown is 0.652. In this case, the average value of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle are determined as walking parameters. Moreover, the prediction model created using the average value of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0247] The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a first period of a stance period of one leg and time series data of angle of the ankle joint during a second period of a foot suspension period of one leg. The first period is a period from 21% to 30% of one walking cycle, and the second period is a period from 81% to 100% of one walking cycle. The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a period from 21% to 30% of one walking cycle and time series data of angle of the ankle joint of one leg during a period from 81% to 100% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of displacement of the waist in the vertical direction during a period from 21% to 30% of one walking cycle and the average value of the time series data of angle of the ankle joint of one leg during a period from 81% to 100% of one walking cycle.
[0248] The cognitive function determination unit 113 determines the degree of the cognitive function of the test subject using the average value of the time series data of the vertical displacement of the waist and the average value of the time series data of the angle of the ankle joint.
[0249] The memory 12 stores in advance a prediction model that uses as input the average value of time series data of vertical displacement of the waist during a first period of a stance period of one leg and the average value of time series data of angle of the ankle joint during a second period of a foot-free period of one leg as output values, and generates a prediction model that uses as input the average value of vertical displacement of the waist during a period of 21% to 30% of a gait cycle and the average value of angle of the ankle joint of one leg during a period of 81% to 100% of a gait cycle as output values, and generates a prediction model that uses as input the average value of vertical displacement of the waist during a period of 21% to 30% of a gait cycle and the average value of angle of the ankle joint of one leg during a period of 81% to 100% of a gait cycle as output values, and generates a prediction model that uses as output values the average value of vertical displacement of the waist during a period of 21% to 30% of a gait cycle and the average value of angle of the ankle joint of one leg as output values.
[0250] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle into the prediction model.
[0251] Thus, the AUC value of the prediction model created using only the vertical displacement of the waist is 0.587, while the AUC value of the prediction model created using the angle of the ankle joint and the vertical displacement of the waist is 0.652. Therefore, compared with the prediction model created using only the vertical displacement of the waist, the prediction model created using the angle of the ankle joint and the vertical displacement of the waist can judge the degree of cognitive function with higher accuracy.
[0252] Next, walking parameters according to an eighth modified example of the present embodiment will be described.
[0253] The walking parameter of the eighth variant of the present embodiment may also be the average value of the time series data of the vertical displacement of the waist during the first period of the stance period of one leg and the average value of the time series data of the angle of the knee joint during the second period of the stance period of one leg.
[0254] The eighth variant of the present embodiment is the same as the above-mentioned experiment, and detects the time series data of the vertical displacement of the waist of each of the multiple test subjects and the time series data of the angle of the knee joint of each of the multiple test subjects. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable, and takes the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0255] Fig.23 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the eighth modification example of the present embodiment.
[0256] The prediction model of the eighth variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as explanatory variables. Fig.23 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0257] Fig.23The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as explanatory variables. Fig.23 The AUC value of the ROC curve shown is 0.671. In this case, the average value of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle are determined as walking parameters. Moreover, the prediction model created using the average value of the displacement of the waist in the vertical direction during 21% to 30% of a walking cycle and the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function determination unit 113.
[0258] The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a first period of the stance period of one leg and time series data of the angle of the knee joint during a second period of the stance period of one leg. The first period is a period from 21% to 30% of one walking cycle, and the second period is a period from 41% to 50% of one walking cycle. The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a period from 21% to 30% of one walking cycle and time series data of the angle of the knee joint of one leg during a period from 41% to 50% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of displacement of the waist in the vertical direction during a period from 21% to 30% of one walking cycle and the average value of the time series data of the angle of the knee joint of one leg during a period from 41% to 50% of one walking cycle.
[0259] The cognitive function determination unit 113 determines the degree of the cognitive function of the test subject using the average value of the time series data of the vertical displacement of the waist and the average value of the time series data of the angle of the knee joint.
[0260] The memory 12 stores in advance a prediction model which is generated by taking as input the average value of time series data of vertical displacement of the waist during the first period of the stance period of one leg and the average value of time series data of angle of the knee joint during the second period of the stance period of one leg, and outputting whether the test subject has mild cognitive impairment. The memory 12 stores in advance a prediction model which is generated by taking as input the average value of vertical displacement of the waist during the period of 21% to 30% of one gait cycle and the average value of angle of the knee joint of one leg during the period of 41% to 50% of one gait cycle, and outputting whether the test subject has mild cognitive impairment.
[0261] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle and the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle and the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle into the prediction model, and obtains the determination result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0262] Thus, the AUC value of the prediction model created using only the vertical displacement of the waist is 0.587, while the AUC value of the prediction model created using the angle of the knee joint and the vertical displacement of the waist is 0.671. Therefore, compared with the prediction model created using only the vertical displacement of the waist, the prediction model created using the angle of the knee joint and the vertical displacement of the waist can judge the degree of cognitive function with higher accuracy.
[0263] Next, walking parameters according to a ninth modified example of the present embodiment will be described.
[0264] The walking parameters of the ninth variant of the present embodiment may also be the average value of the time series data of the vertical displacement of the waist during the first period of the stance period of one leg, the average value of the time series data of the angle of the knee joint during the second period of the stance period of one leg, and the average value of the time series data of the angle of the ankle joint during the third period of the foot hanging period of one leg.
[0265] The ninth variant of the present embodiment is the same as the above-mentioned experiment, and detects the time series data of the vertical displacement of the waist of each of the multiple test subjects, the time series data of the angle of the knee joint of each of the multiple test subjects, and the time series data of the angle of the ankle joint of each of the multiple test subjects. In addition, a prediction model is created with whether the test subject has mild cognitive impairment as the target variable, and the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle, the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle, and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0266] Fig.24 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the ninth modification example of the present embodiment.
[0267] The prediction model of the ninth variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the vertical displacement of the waist during 21% to 30% of a gait cycle, the average value of the angle of the knee joint of one leg during 41% to 50% of a gait cycle, and the average value of the angle of the ankle joint of one leg during 81% to 100% of a gait cycle as explanatory variables. Fig.24 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0268] Fig.24 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle, the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle, and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables. Fig.24 The AUC value of the ROC curve shown is 0.687. In this case, the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle, the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle, and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle are determined as walking parameters. Moreover, the prediction model generated using the average value of the vertical displacement of the waist during 21% to 30% of a walking cycle, the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle, and the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0269] The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a first period of the stance period of one leg, time series data of angle of the knee joint during a second period of the stance period of one leg, and time series data of angle of the ankle joint during a third period of the foot suspension period of one leg. The first period is a period from 21% to 30% of a walking cycle, the second period is a period from 41% to 50% of a walking cycle, and the third period is a period from 81% to 100% of a walking cycle. The walking parameter detection unit 112 detects time series data of displacement of the waist in the vertical direction during a period from 21% to 30% of a walking cycle, time series data of angle of the knee joint of one leg during a period from 41% to 50% of a walking cycle, and time series data of angle of the ankle joint of one leg during a period from 81% to 100% of a walking cycle.
[0270] Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the vertical displacement of the waist during 21% to 30% of a walking cycle, the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of a walking cycle, and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle.
[0271] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the vertical displacement of the waist, the average value of the time series data of the angle of the knee joint, and the average value of the time series data of the angle of the ankle joint.
[0272] The memory 12 stores in advance a prediction model which uses as inputs the average value of time series data of vertical displacement of the waist during the first period of the stance period of one leg, the average value of time series data of angle of the knee joint during the second period of the stance period of one leg, and the average value of time series data of angle of the ankle joint during the third period of the foot suspension period of one leg, and outputs whether the test subject has mild cognitive impairment. The memory 12 stores in advance a prediction model which uses as inputs the average value of vertical displacement of the waist during 21% to 30% of a gait cycle, the average value of angle of the knee joint of one leg during 41% to 50% of a gait cycle, and the average value of angle of the ankle joint of one leg during 81% to 100% of a gait cycle, and outputs whether the test subject has mild cognitive impairment.
[0273] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle, the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle, and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the displacement of the waist in the vertical direction during 21% to 30% of one gait cycle, the average value of the time series data of the angle of the knee joint of one leg during 41% to 50% of one gait cycle, and the average value of the time series data of the angle of the ankle joint of one leg during 81% to 100% of one gait cycle into the prediction model, and obtains the judgment result indicating whether the test subject has mild cognitive impairment from the prediction model.
[0274] Thus, the AUC values of the prediction model created by using the angle of the ankle joint, the angle of the knee joint, and the displacement of the waist in the vertical direction are 0.660, 0.682, and 0.587, respectively, and the AUC value of the prediction model created by using the angle of the ankle joint, the angle of the knee joint, and the displacement of the waist in the vertical direction is 0.687. Therefore, compared with the prediction model created by using the angle of the ankle joint, the angle of the knee joint, and the displacement of the waist in the vertical direction, the prediction model created by using the angle of the ankle joint, the angle of the knee joint, and the displacement of the waist in the vertical direction can judge the degree of cognitive function with higher accuracy.
[0275] Next, walking parameters according to a tenth modified example of the present embodiment will be described.
[0276] In a tenth variation of the present embodiment, when the test subject is a male, the walking parameter may be an average value of time series data of the angle of the ankle joint at the initial stage of the stance phase of one leg.
[0277] Fig.25 1 is a schematic diagram showing the change in angle of the ankle joint of one leg of a male test subject during one walking cycle in the tenth variation of the present embodiment. Fig.25 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Fig.25 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of male healthy subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of male patients with mild cognitive impairment.
[0278] The tenth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the angle of the ankle joint of each leg of the male test subject is detected. From the average waveform of the angle of the ankle joint of one leg of the male test subject, it can be seen that in the early stage of the stance period, that is, during 1% to 2% of a walking cycle, the angle of the ankle joint of patients with mild cognitive impairment is larger than the angle of the ankle joint of healthy people. Here, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the ankle joint of one leg during 1% to 2% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0279] Fig.26 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the tenth modification example of the present embodiment.
[0280] The prediction model of the tenth variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the ankle joint of one leg of the male test subject during 1% to 2% of a walking cycle as the explanatory variable. Fig.26 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0281] Fig.26 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the angle of the ankle joint of one leg of the male test subject during 1% to 2% of one walking cycle as an explanatory variable. Fig.26 The AUC value of the ROC curve shown is 0.7174. When the test subject is a male, the average value of the angle of the ankle joint during 1% to 2% of a walking cycle is determined as the walking parameter. Furthermore, a prediction model created using the average value of the angle of the ankle joint of one leg of the male test subject during 1% to 2% of a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0282] The processor 11 of the cognitive function evaluation device 1 of the tenth variant of the present embodiment further includes a gender recognition unit for recognizing the gender of the test subject. The gender recognition unit detects a feature value of at least one of the face and the body from the image of the test subject contained in the dynamic image data. Then, the gender recognition unit recognizes whether the test subject is male or female based on the detected feature value. In addition, the memory 12 may also pre-store user information that makes the facial image and gender of the test subject correspond to each other. In this case, the gender recognition unit may also use the user information stored in the memory 12 to recognize whether the test subject contained in the dynamic image data is male or female.
[0283] The memory 12 pre-stores a prediction model generated by using the average value of the time series data of the angle of the ankle joint of one leg of a male test subject during 1% to 2% of a gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0284] The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint at the initial stage of the stance phase of one leg when the gender recognition unit identifies the test subject as a male. The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint of one leg during 1% to 2% of one walking cycle when the gender recognition unit identifies the test subject as a male. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the ankle joint of one leg during 1% to 2% of one walking cycle.
[0285] The cognitive function determination unit 113 determines the degree of cognitive function of the male test subject using the average value of the time series data of the angle of the ankle joint of one leg during 1% to 2% of one gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the male test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the angle of the ankle joint of one leg during 1% to 2% of one gait cycle into the prediction model.
[0286] Furthermore, when the test subject is a male, at the beginning of the stance period (1% to 2% of a walking cycle), the average waveform of the angle of the ankle joint of one leg of the patients with mild cognitive impairment is higher than the average waveform of the angle of the ankle joint of one leg of the healthy subjects. For this reason, the value between the average of the time series data of the angle of the ankle joint of one leg of the male patients with mild cognitive impairment during the period of 1% to 2% of a walking cycle and the average of the time series data of the angle of the ankle joint of one leg of the male healthy subjects during the period of 1% to 2% of a walking cycle obtained by the experiment can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the angle of the ankle joint of one leg of the test subject during the period of 1% to 2% of a walking cycle with a pre-stored threshold.
[0287] As described above, when the test subject is a male, the degree of cognitive function can be determined with higher accuracy by using time series data of the angle of the ankle joint of one leg at the beginning of the stance phase (the period of 1% to 2% of one gait cycle).
[0288] Next, walking parameters according to an eleventh modified example of the present embodiment will be described.
[0289] In the eleventh variant of the present embodiment, when the test subject is a male, the walking parameter may also be the average value of the time series data of the first angle of the ankle joint in the first period of the stance period of one leg and the average value of the time series data of the second angle of the ankle joint in the second period of the foot suspension period of one leg.
[0290] Fig. 27 FIG. 1 is a schematic diagram showing the change in angle of the ankle joint of one leg of a male test subject during one walking cycle in the eleventh variation of the present embodiment. Fig. 27 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Fig. 27 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of male healthy subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of male patients with mild cognitive impairment.
[0291] The eleventh variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the angle of the ankle joint of each leg of the male test subject is detected. In addition, a prediction model is created, which uses whether the test subject has mild cognitive impairment as the target variable, and uses the average value of the angle of the ankle joint of one leg during 1% to 40% of a walking cycle and the average value of the angle of the ankle joint of one leg during 71% to 90% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0292] Fig.28 Schematic diagram showing a ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the eleventh variation of the present embodiment.
[0293] The prediction model of the eleventh variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the angle of the ankle joint of one leg of the male test subject during 1% to 40% of a walking cycle and the average value of the angle of the ankle joint of one leg of the male test subject during 71% to 90% of a walking cycle as explanatory variables. Fig.28 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0294] Fig.28 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the angle of the ankle joint of one leg of the male test subject during 1% to 40% of a walking cycle and the average value of the angle of the ankle joint of one leg of the male test subject during 71% to 90% of a walking cycle as explanatory variables. Fig.28 The AUC value of the ROC curve shown is 0.648. When the experimental subject is a male, the average value of the angle of the ankle joint during 1% to 40% of a walking cycle and the average value of the angle of the ankle joint during 71% to 90% of a walking cycle are determined as walking parameters. Moreover, a prediction model created using the average value of the angle of the ankle joint of one leg of the male experimental subject during 1% to 40% of a walking cycle and the average value of the angle of the ankle joint of one leg of the male experimental subject during 71% to 90% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0295] The processor 11 of the cognitive function evaluation device 1 according to the eleventh modified example of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0296] The walking parameter detection unit 112, when the test subject is identified as a male by the gender recognition unit, detects time series data of a first angle of the ankle joint in a first period of the stance period of one leg and time series data of a second angle of the ankle joint in a second period of the foot suspension period of one leg. The first period is a period from 1% to 40% of one walking cycle, and the second period is a period from 71% to 90% of one walking cycle. The walking parameter detection unit 112, when the test subject is identified as a male by the gender recognition unit, detects time series data of the angle of the ankle joint of one leg in a period from 1% to 40% of one walking cycle and time series data of the angle of the ankle joint of one leg in a period from 71% to 90% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates an average value of the time series data of the angle of the ankle joint of one leg in a period from 1% to 40% of one walking cycle and an average value of the time series data of the angle of the ankle joint of one leg in a period from 71% to 90% of one walking cycle.
[0297] The cognitive function determination unit 113 determines the degree of cognitive function of the male test subject using the average value of the time series data of the first angle and the average value of the time series data of the second angle of the ankle joint.
[0298] The memory 12 pre-stores a prediction model that uses the average value of time series data of a first angle of the ankle joint of a male test subject during a first period of a stance period of one leg and the average value of time series data of a second angle of the ankle joint of the male test subject during a second period of a foot-free period of one leg as input values, and uses the output value to determine whether the test subject has mild cognitive impairment. The memory 12 pre-stores a prediction model that uses the average value of angles of the ankle joint of one leg of the male test subject during a period of 1% to 40% of a walking cycle and the average value of angles of the ankle joint of one leg of the male test subject during a period of 71% to 90% of a walking cycle as input values, and uses the output value to determine whether the test subject has mild cognitive impairment.
[0299] The cognitive function determination unit 113 determines the degree of cognitive function of the male test subject using the average value of the time series data of the angle of the ankle joint of one leg during 1% to 40% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 71% to 90% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the angle of the ankle joint of one leg during 1% to 40% of one gait cycle and the average value of the time series data of the angle of the ankle joint of one leg during 71% to 90% of one gait cycle into the prediction model, and obtains a determination result indicating whether the male test subject has mild cognitive impairment from the prediction model.
[0300] As described above, when the test subject is a male, by utilizing time series data of the angles of the ankle joint of one leg during the stance phase (the period from 1% to 40% of a gait cycle) and the foot-in-the-air phase (the period from 71% to 90% of a gait cycle), the degree of cognitive function can be determined with greater accuracy.
[0301] Next, walking parameters according to a twelfth modified example of the present embodiment will be described.
[0302] In a twelfth variation of the present embodiment, when the test subject is a male, the walking parameter may be an average value of time series data of the angle of the knee joint during the foot-in-the-air period of one leg.
[0303] Fig.29 1 is a schematic diagram showing the change in the angle of the knee joint of one leg of a male test subject during one walking cycle in the twelfth variation of the present embodiment. Fig.29 In the figure, the vertical axis represents the angle of the knee joint, and the horizontal axis represents a standardized walking cycle. Fig.29 In FIG. 1 , the dotted line represents the average waveform of the angle of the knee joint of one leg of male healthy subjects, and the solid line represents the average waveform of the angle of the knee joint of one leg of male patients with mild cognitive impairment.
[0304] The twelfth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the angle of the knee joint of each leg of the male test subject is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the knee joint of one leg during 71% to 100% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0305] Fig.30Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the twelfth variation of the present embodiment.
[0306] The prediction model of the twelfth variant of the present embodiment uses whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the knee joint of one leg of the male test subject during 71% to 100% of a walking cycle as the explanatory variable. Fig.30 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0307] Fig.30 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model using the average value of the angle of the knee joint of one leg of the male test subject during 71% to 100% of one walking cycle as an explanatory variable. Fig.30 The AUC value of the ROC curve shown is 0.693. When the test subject is a male, the average value of the angle of the knee joint during 71% to 100% of a walking cycle is determined as the walking parameter. Furthermore, a prediction model created using the average value of the angle of the knee joint of one leg of the male test subject during 71% to 100% of a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0308] The processor 11 of the cognitive function evaluation device 1 according to the twelfth variation of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0309] The memory 12 pre-stores a prediction model generated by using the average value of the time series data of the angle of the knee joint of one leg of a male test subject during 71% to 100% of a gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0310] The walking parameter detection unit 112 detects the time series data of the angle of the knee joint of one leg during the foot-in-the-air period when the gender recognition unit identifies the test subject as a male. The walking parameter detection unit 112 detects the time series data of the angle of the knee joint of one leg during the period from 71% to 100% of one walking cycle when the gender recognition unit identifies the test subject as a male. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the knee joint of one leg during the period from 71% to 100% of one walking cycle.
[0311] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint of one leg during the period from 71% to 100% of one gait cycle. The cognitive function determination unit 113 determines the degree of cognitive function of the male test subject using the average value of the time series data of the angle of the knee joint of one leg during the period from 71% to 100% of one gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the male test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the angle of the knee joint of one leg during the period from 71% to 100% of one gait cycle into the prediction model.
[0312] Furthermore, when the test subject is a male, the average waveform of the angle of the knee joint of one leg of the patients with mild cognitive impairment is smaller than the average waveform of the angle of the knee joint of one leg of the healthy subjects during the foot-in-the-air period (71% to 100% of a gait cycle). For this reason, the value between the average of the time series data of the angle of the knee joint of one leg of the male patients with mild cognitive impairment during the period of 71% to 100% of a gait cycle and the average of the time series data of the angle of the knee joint of one leg of the male healthy subjects during the period of 71% to 100% of a gait cycle obtained by the experiment can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the angle of the knee joint of one leg of the test subject during the period of 71% to 100% of a gait cycle with a pre-stored threshold.
[0313] As described above, when the test subject is a male, the degree of cognitive function can be determined with higher accuracy by using time series data of the angle of the knee joint of one leg during the foot-in-the-air period (the period from 71% to 100% of one gait cycle).
[0314] Next, walking parameters according to a thirteenth modified example of the present embodiment will be described.
[0315] In the thirteenth variant of the present embodiment, when the test subject is a male, the walking parameter may be the average value of the time series data of the first displacement of the waist in the vertical direction during the first period of a walking cycle, the average value of the time series data of the second displacement of the waist in the vertical direction during the second period of a walking cycle, and the average value of the time series data of the third displacement of the waist in the vertical direction during the third period of a walking cycle.
[0316] Fig.31 FIG. 13 is a schematic diagram showing the vertical displacement of the waist of a male test subject in one walking cycle in the thirteenth variation of the present embodiment. Fig.31In the figure, the vertical axis represents the vertical displacement of the waist, and the horizontal axis represents a standardized walking cycle. Fig.31 In FIG. 1 , the dotted line represents the average waveform of the vertical displacement of the waist of male healthy subjects, and the solid line represents the average waveform of the vertical displacement of the waist of male patients with mild cognitive impairment.
[0317] The thirteenth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the vertical displacement of the waist of each male test subject is detected. In addition, a prediction model is created with whether the test subject has mild cognitive impairment as the target variable, and the average value of the first displacement of the waist in the vertical direction in the first period of 1% to 10% of a walking cycle, the average value of the second displacement of the waist in the vertical direction in the second period of 31% to 40% of a walking cycle, and the average value of the third displacement of the waist in the vertical direction in the third period of 51% to 80 of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0318] Fig.32 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the thirteenth variation of the present embodiment.
[0319] The prediction model of the thirteenth variant of the present embodiment uses whether the test subject has mild cognitive impairment as the target variable, and uses the average value of the first displacement of the waist of the male test subject in the vertical direction during the first period of 1% to 10% of a walking cycle, the average value of the second displacement of the waist of the male test subject in the vertical direction during the second period of 31% to 40% of a walking cycle, and the average value of the third displacement of the waist of the male test subject in the vertical direction during the third period of 51% to 80% of a walking cycle as explanatory variables. Fig.32 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0320] Fig.32The ROC curve shown is a curve depicting the true positive rate and false positive rate of the prediction model, which uses the average value of the first displacement of the waist of the male subject in the vertical direction during a first period of 1% to 10% of a walking cycle, the average value of the second displacement of the waist of the male subject in the vertical direction during a second period of 31% to 40% of a walking cycle, and the average value of the third displacement of the waist of the male subject in the vertical direction during a third period of 51% to 80% of a walking cycle as explanatory variables. Fig.32 The AUC value of the ROC curve shown is 0.852. When the experimental subject is a male, the average value of the first displacement of the waist in the vertical direction during the first period of 1% to 10% of a walking cycle, the average value of the second displacement of the waist in the vertical direction during the second period of 31% to 40% of a walking cycle, and the average value of the third displacement of the waist in the vertical direction during the third period of 51% to 80% of a walking cycle are defined as walking parameters. Moreover, a prediction model created using the average value of the first displacement of the waist in the vertical direction during the first period of 1% to 10% of a walking cycle of the male test subject, the average value of the second displacement of the waist in the vertical direction during the second period of 31% to 40% of a walking cycle of the male test subject, and the average value of the third displacement of the waist in the vertical direction during the third period of 51% to 80% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0321] The processor 11 of the cognitive function evaluation device 1 according to the thirteenth modified example of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0322] The walking parameter detection unit 112, when the test subject is identified as a male by the gender recognition unit, detects time series data of a first displacement of the waist in the vertical direction in a first period of a walking cycle, time series data of a second displacement of the waist in the vertical direction in a second period of a walking cycle, and time series data of a third displacement of the waist in the vertical direction in a third period of a walking cycle. The first period is a period from 1% to 10% of a walking cycle, the second period is a period from 31% to 40% of a walking cycle, and the third period is a period from 51% to 80% of a walking cycle. The walking parameter detection unit 112, when the test subject is identified as a male by the gender recognition unit, detects time series data of a first displacement of the waist in the vertical direction in a first period from 1% to 10% of a walking cycle, time series data of a second displacement of the waist in the vertical direction in a second period from 31% to 40% of a walking cycle, and time series data of a third displacement of the waist in the vertical direction in a third period from 51% to 80% of a walking cycle.
[0323] In addition, the third period can also be divided into a period from 51% to 60% of a gait cycle and a period from 61% to 80% of a gait cycle. The period from 51% to 60% of a gait cycle belongs to the foot-standing period, and the period from 61% to 80% of a gait cycle belongs to the foot-in-the-air period.
[0324] Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the first displacement of the waist in the vertical direction during the first period from 1% to 10% of a walking cycle, the average value of the time series data of the second displacement of the waist in the vertical direction during the second period from 31% to 40% of a walking cycle, and the average value of the time series data of the third displacement of the waist in the vertical direction during the third period from 51% to 80% of a walking cycle.
[0325] The cognitive function judgment unit 113 judges the degree of cognitive function of the male test subject using the average value of the time series data of the first displacement of the waist in the vertical direction, the average value of the time series data of the second displacement of the waist in the vertical direction, and the average value of the time series data of the third displacement of the waist in the vertical direction.
[0326] The memory 12 stores in advance a prediction model that uses as input the average value of time series data of first vertical displacement of the waist of the male test subject in a first period of a gait cycle, the average value of time series data of second vertical displacement of the waist of the male test subject in a second period of a gait cycle, and the average value of time series data of third vertical displacement of the waist of the male test subject in a third period of a gait cycle, and generates as an output whether the test subject has mild cognitive impairment. The memory 12 stores in advance a prediction model that uses as input the average value of first vertical displacement of the waist of the male test subject in a first period of 1% to 10% of a gait cycle, the average value of second vertical displacement of the waist of the male test subject in a second period of 31% to 40% of a gait cycle, and the average value of third vertical displacement of the waist of the male test subject in a third period of 51% to 80% of a gait cycle, and generates as an output whether the test subject has mild cognitive impairment.
[0327] The cognitive function determination unit 113 determines the degree of cognitive function of the male test subject using the average value of the time series data of the first displacement of the waist in the vertical direction during the first period of 1% to 10% of one gait cycle, the average value of the time series data of the second displacement of the waist in the vertical direction during the second period of 31% to 40% of one gait cycle, and the average value of the time series data of the third displacement of the waist in the vertical direction during the third period of 51% to 80% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the first displacement of the waist in the vertical direction during the first period of 1% to 10% of one gait cycle, the average value of the time series data of the second displacement of the waist in the vertical direction during the second period of 31% to 40% of one gait cycle, and the average value of the time series data of the third displacement of the waist in the vertical direction during the third period of 51% to 80% of one gait cycle into the prediction model, and obtains a determination result indicating whether the male test subject has mild cognitive impairment from the prediction model.
[0328] As described above, when the test subject is a male, by utilizing time series data of the vertical displacement of the waist during the first period of a gait cycle (the period from 1% to 10% of a gait cycle), the second period of a gait cycle (the period from 31% to 40% of a gait cycle), and the third period of a gait cycle (the period from 51% to 80% of a gait cycle), the degree of cognitive function can be determined with higher accuracy.
[0329] Next, walking parameters according to a fourteenth modified example of the present embodiment will be described.
[0330] In a fourteenth variation of the present embodiment, when the test subject is a female, the walking parameter may be an average value of time series data of the angle of the ankle joint during a period when the foot is in the air on one leg.
[0331] Fig.33 14 is a schematic diagram showing the change in angle of the ankle joint of one leg of a female test subject during a walking cycle in the fourteenth variation of the present embodiment. Fig.33 In the figure, the vertical axis represents the angle of the ankle joint, and the horizontal axis represents a standardized walking cycle. Fig.33 In the figure, the dotted line represents the average waveform of the angle of the ankle joint of one leg of the healthy female subjects, and the solid line represents the average waveform of the angle of the ankle joint of one leg of the female patients with mild cognitive impairment.
[0332] The fourteenth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the angle of the ankle joint of each leg of the female test subject is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the ankle joint of one leg during 81% to 100% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0333] Fig.34 Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fourteenth variation of the present embodiment.
[0334] The prediction model of the fourteenth variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and taking the average value of the angle of the ankle joint of one leg of the female test subject during 81% to 100% of a walking cycle as the explanatory variable. Fig.34 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0335] Fig.34 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the angle of the ankle joint of one leg of the female test subject during 81% to 100% of one walking cycle as an explanatory variable. Fig.34 The AUC value of the ROC curve shown is 0.803. When the experimental subject is a female, the average value of the angle of the ankle joint during 81% to 100% of a walking cycle is determined as the walking parameter. Furthermore, a prediction model created using the average value of the angle of the ankle joint of one leg of the female experimental subject during 81% to 100% of a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0336] The processor 11 of the cognitive function evaluation device 1 according to the fourteenth modified example of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0337] The memory 12 pre-stores a prediction model generated by using the average value of the time series data of the angle of the ankle joint of one leg of a female test subject during 81% to 100% of a gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0338] The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint of one leg during the foot-in-the-air period when the gender recognition unit identifies the test subject as a female. The walking parameter detection unit 112 detects the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of one walking cycle when the gender recognition unit identifies the test subject as a female. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of one walking cycle.
[0339] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the ankle joint of one leg during the foot-in-the-air period. The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of a gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the female test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the angle of the ankle joint of one leg during the period from 81% to 100% of a gait cycle into the prediction model.
[0340] Furthermore, when the test subject is a female, the average waveform of the angle of the ankle joint of one leg of the patients with mild cognitive impairment is smaller than the average waveform of the angle of the ankle joint of one leg of the healthy subjects during the foot-in-the-air period (81% to 100% of a gait cycle). For this reason, the value between the average of the time series data of the angle of the ankle joint of one leg of the female patients with mild cognitive impairment during the period of 81% to 100% of a gait cycle and the average of the time series data of the angle of the ankle joint of one leg of the healthy female subjects during the period of 81% to 100% of a gait cycle obtained by the experiment can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the angle of the ankle joint of one leg of the test subject during the period of 81% to 100% of a gait cycle with a pre-stored threshold.
[0341] As described above, when the test subject is a female, the degree of cognitive function can be determined with higher accuracy by using time series data of the angle of the ankle joint of one leg during the foot-in-the-air period (the period from 81% to 100% of one gait cycle).
[0342] Next, walking parameters according to a fifteenth modified example of the present embodiment will be described.
[0343] In a fifteenth variation of the present embodiment, when the test subject is a female, the walking parameter may be an average value of time series data of the angle of the knee joint in the stance phase of one leg.
[0344] Fig.35 15 is a schematic diagram showing the change in the angle of the knee joint of one leg of a female test subject during one walking cycle in the fifteenth variation of the present embodiment. Fig.35 In the figure, the vertical axis represents the angle of the knee joint, and the horizontal axis represents a standardized walking cycle. Fig.35 In the figure, the dotted line represents the average waveform of the knee joint angle of one leg of healthy female subjects, and the solid line represents the average waveform of the knee joint angle of one leg of female patients with mild cognitive impairment.
[0345] The fifteenth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the angle of the knee joint of each leg of the female test subject is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and takes the average value of the angle of the knee joint of one leg during 41% to 50% of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0346] Fig.36 It is a schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the fifteenth variation of the present embodiment.
[0347] The prediction model of the fifteenth variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the average value of the angle of the knee joint of one leg of the female test subject during 41% to 50% of a walking cycle as the explanatory variable. Fig.36 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0348] Fig.36 The ROC curve shown is a curve depicting the true positive rate and the false positive rate of a prediction model created using the average value of the angle of the knee joint of one leg of the female test subject during 41% to 50% of one walking cycle as an explanatory variable. Fig.36 The AUC value of the ROC curve shown is 0.733. When the test subject is a female, the average value of the angle of the knee joint during 41% to 50% of a walking cycle is determined as the walking parameter. Furthermore, a prediction model created using the average value of the angle of the knee joint of one leg of the female test subject during 41% to 50% of a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0349] The processor 11 of the cognitive function evaluation device 1 according to the fifteenth variation of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0350] The memory 12 pre-stores a prediction model generated by using the average value of the time series data of the angle of the knee joint of one leg of a female test subject during 41% to 50% of a gait cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0351] The walking parameter detection unit 112 detects the time series data of the angle of the knee joint in the stance phase of one leg when the gender recognition unit identifies the test subject as a female. The walking parameter detection unit 112 detects the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one walking cycle when the gender recognition unit identifies the test subject as a female. Furthermore, the walking parameter detection unit 112 calculates the average value of the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one walking cycle.
[0352] The cognitive function determination unit 113 determines the degree of cognitive function of the test subject using the average value of the time series data of the angle of the knee joint during the stance period of one leg. The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the average value of the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one gait cycle. The cognitive function determination unit 113 obtains a determination result indicating whether the female test subject has mild cognitive impairment from the prediction model by inputting the average value of the time series data of the angle of the knee joint of one leg during the period from 41% to 50% of one gait cycle into the prediction model.
[0353] Furthermore, when the test subject is a female, the average waveform of the angle of the knee joint of one leg of the patients with mild cognitive impairment is smaller than the average waveform of the angle of the knee joint of one leg of the healthy subjects during the stance period (41% to 50% of a gait cycle). For this reason, the value between the average of the time series data of the angle of the knee joint of one leg of the female patients with mild cognitive impairment during the period of 41% to 50% of a gait cycle and the average of the time series data of the angle of the knee joint of one leg of the healthy female subjects during the period of 41% to 50% of a gait cycle obtained by the experiment can be stored in the memory 12 as a threshold. The cognitive function determination unit 113 can also determine the degree of cognitive function by comparing the average of the time series data of the angle of the knee joint of one leg of the test subject during the period of 41% to 50% of a gait cycle with a pre-stored threshold.
[0354] As described above, when the test subject is a female, the degree of cognitive function can be determined with higher accuracy by using time series data of the angle of the knee joint of one leg during the stance phase (the period from 41% to 50% of one gait cycle).
[0355] Next, walking parameters according to a sixteenth modification of the present embodiment will be described.
[0356] In the sixteenth variant of the present embodiment, when the test subject is a female, the walking parameter may be the average value of the time series data of the first displacement of the waist in the vertical direction during the first period of the stance period of one leg and the average value of the time series data of the second displacement of the waist in the vertical direction during the second period of the foot of one leg suspended.
[0357] Fig.37 FIG. 1 is a schematic diagram showing the vertical displacement of the waist of a female test subject in one walking cycle in the sixteenth variation of the present embodiment. Fig.37 In the figure, the vertical axis represents the vertical displacement of the waist, and the horizontal axis represents a standardized walking cycle. Fig.37 In FIG. 1 , the dotted line represents the average waveform of the vertical displacement of the waist of healthy female subjects, and the solid line represents the average waveform of the vertical displacement of the waist of female patients with mild cognitive impairment.
[0358] The sixteenth variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the vertical displacement of the waist of each female test subject is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable, and takes the average value of the vertical displacement of the waist during 11% to 20% of a walking cycle and the average value of the vertical displacement of the waist during 71% to 80% of a walking cycle as explanatory variables. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0359] Fig.38 It is a schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the sixteenth variation of the present embodiment.
[0360] The prediction model of the sixteenth variant of the present embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable, and taking the average value of the vertical displacement of the waist of the female test subject during 11% to 20% of a walking cycle and the vertical displacement of the waist of the female test subject during 71% to 80% of a walking cycle as explanatory variables. Fig.38 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0361] Fig.38 The ROC curve shown is a curve depicting the true positive rate and false positive rate of a prediction model created using the average value of the vertical displacement of the female test subjects' waist during 11% to 20% of a walking cycle and the average value of the vertical displacement of the female test subjects' waist during 71% to 80% of a walking cycle as explanatory variables. Fig.38 The AUC value of the ROC curve shown is 0.813. When the test subject is a female, the average value of the vertical displacement of the waist during 11% to 20% of a walking cycle and the average value of the vertical displacement of the waist during 71% to 80% of a walking cycle are determined as walking parameters. Moreover, a prediction model created using the average value of the vertical displacement of the waist of the female test subject during 11% to 20% of a walking cycle and the average value of the vertical displacement of the waist of the female test subject during 71% to 80% of a walking cycle as explanatory variables is determined as the prediction model used by the cognitive function judgment unit 113.
[0362] The processor 11 of the cognitive function evaluation device 1 according to the sixteenth variation of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0363] The walking parameter detection unit 112, when the test subject is identified as a female by the gender recognition unit, detects time series data of a first displacement of the waist in the vertical direction during a first period of a stance period of one leg and time series data of a second displacement of the waist in the vertical direction during a second period of a foot suspension period of one leg. The first period is a period from 11% to 20% of one walking cycle, and the second period is a period from 71% to 80% of one walking cycle. The walking parameter detection unit 112, when the test subject is identified as a female by the gender recognition unit, detects time series data of a displacement of the waist in the vertical direction during a period from 11% to 20% of one walking cycle and time series data of a displacement of the waist in the vertical direction during a period from 71% to 80% of one walking cycle. Furthermore, the walking parameter detection unit 112 calculates an average value of the time series data of the displacement of the waist in the vertical direction during a period from 11% to 20% of one walking cycle and an average value of the time series data of the displacement of the waist in the vertical direction during a period from 71% to 80% of one walking cycle.
[0364] The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the average value of the time series data of the first displacement of the waist in the vertical direction during the first period of the standing period and the average value of the time series data of the second displacement of the waist in the vertical direction during the second period of the foot hanging period.
[0365] The memory 12 pre-stores a prediction model that uses the average value of time series data of first vertical displacement of the waist of the female test subject during the first period of the stance period of one leg and the average value of time series data of second vertical displacement of the waist of the female test subject during the second period of the foot suspension period of one leg as input values, and uses the output value to determine whether the test subject has mild cognitive impairment. The memory 12 pre-stores a prediction model that uses the average value of vertical displacement of the waist of the female test subject during 11% to 20% of a walking cycle and the average value of vertical displacement of the waist of the female test subject during 71% to 80% of a walking cycle as input values, and uses the output value to determine whether the test subject has mild cognitive impairment.
[0366] The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the average value of the time series data of the displacement of the waist in the vertical direction during 11% to 20% of one gait cycle and the average value of the time series data of the displacement of the waist in the vertical direction during 71% to 80% of one gait cycle. The cognitive function determination unit 113 inputs the average value of the time series data of the displacement of the waist in the vertical direction during 11% to 20% of one gait cycle and the average value of the time series data of the displacement of the waist in the vertical direction during 71% to 80% of one gait cycle into the prediction model, and obtains the determination result indicating whether the female test subject has mild cognitive impairment from the prediction model.
[0367] As described above, when the test subject is a female, by utilizing time series data of the vertical displacement of the waist during the stance period (the period from 11% to 20% of a gait cycle) and the foot-in-the-air period (the period from 71% to 80% of a gait cycle), the degree of cognitive function can be determined with greater accuracy.
[0368] Next, walking parameters according to a seventeenth modification of the present embodiment will be described.
[0369] In the seventeenth variation of the present embodiment, when the test subject is a female, the walking parameter may be the vertical displacement of the waist at a predetermined time when the foot of one leg is suspended in the air.
[0370] Fig.39 FIG. 1 is a schematic diagram showing the vertical displacement of the waist of a female test subject in one walking cycle in the seventeenth variation of the present embodiment. Fig.39 In the figure, the vertical axis represents the vertical displacement of the waist, and the horizontal axis represents a standardized walking cycle. Fig.39 In FIG. 1 , the dotted line represents the average waveform of the vertical displacement of the waist of healthy female subjects, and the solid line represents the average waveform of the vertical displacement of the waist of female patients with mild cognitive impairment.
[0371] The 17th variant of the present embodiment is different from the above-mentioned experiment, and the time series data of the vertical displacement of the waist of each female test subject is detected. In addition, a prediction model is created, which takes whether the test subject has mild cognitive impairment as the target variable and the vertical displacement of the waist at 74% of the time of a walking cycle as the explanatory variable. The prediction model is evaluated by cross-validation. Leave-one-out cross-validation is used as cross-validation. Then, the ROC curve of the prediction model is calculated. In addition, the AUC value of the ROC curve of the prediction model is calculated.
[0372] Fig.40Schematic diagram showing an ROC curve obtained from the results of judging healthy subjects and mild cognitive impairment using the prediction model of the seventeenth variation of the present embodiment.
[0373] The prediction model of the seventeenth variant of this embodiment is created by taking whether the test subject has mild cognitive impairment as the target variable and the vertical displacement of the waist of the female test subject at 74% of a walking cycle as the explanatory variable. Fig.40 In the figure, the vertical axis represents the true positive rate and the horizontal axis represents the false positive rate. The true positive rate represents the proportion of patients with mild cognitive impairment correctly judged as mild cognitive impairment by the prediction model, and the false positive rate represents the proportion of healthy people incorrectly judged as mild cognitive impairment by the prediction model.
[0374] Fig.40 The ROC curve shown in is a curve depicting the true positive rate and the false positive rate of a prediction model created using the vertical displacement of the waist of a female test subject at 74% of the time in one walking cycle as an explanatory variable. Fig.40 The AUC value of the ROC curve shown is 0.7401. When the test subject is a female, the vertical displacement of the waist at 74% of the time in a walking cycle is determined as the walking parameter. In addition, the prediction model created using the vertical displacement of the waist of the female test subject at 74% of the time in a walking cycle as an explanatory variable is determined as the prediction model used by the cognitive function determination unit 113.
[0375] The processor 11 of the cognitive function evaluation device 1 according to the seventeenth variation of the present embodiment further includes a gender recognition unit that recognizes the gender of the test subject.
[0376] The walking parameter detection unit 112 detects the vertical displacement of the waist at a predetermined time when the foot of one leg is suspended in the air, when the gender recognition unit identifies the test subject as a female. The predetermined time is 74% of a walking cycle. The walking parameter detection unit 112 detects the vertical displacement of the waist at 74% of a walking cycle, when the gender recognition unit identifies the test subject as a female.
[0377] In addition, 74% of the time of one walking cycle is near the time when the maximum displacement of the waist in the vertical direction during the period when the foot of one leg is suspended in the air is detected. Therefore, when the gender recognition unit recognizes that the test subject is a female, the walking parameter detection unit 112 can also detect the maximum displacement of the waist in the vertical direction during the period when the foot of one leg is suspended in the air.
[0378] The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the displacement of the waist in the vertical direction at a predetermined time during the period when the feet are suspended in the air.
[0379] The memory 12 stores in advance a prediction model generated by taking the vertical displacement of the waist of the female test subject at a predetermined time during a period when one leg is suspended as an input value and whether the test subject has mild cognitive impairment as an output value. The memory 12 stores in advance a prediction model generated by taking the vertical displacement of the waist of the female test subject at a time of 74% of a walking cycle as an input value and whether the test subject has mild cognitive impairment as an output value.
[0380] The cognitive function determination unit 113 determines the degree of cognitive function of the female test subject using the vertical displacement of the waist at 74% of a gait cycle. The cognitive function determination unit 113 inputs the vertical displacement of the waist at 74% of a gait cycle into the prediction model, and obtains a determination result indicating whether the female test subject has mild cognitive impairment from the prediction model.
[0381] As described above, when the test subject is a female, the degree of cognitive function can be determined with higher accuracy by utilizing the vertical displacement of the waist at a predetermined time (74% of one gait cycle) during the foot-in-the-air period.
[0382] Fig.41 It is a schematic diagram showing an example of the evaluation result screen displayed in this embodiment.
[0383] The display unit 3 displays Fig.41 The evaluation result screen includes a cognitive function evaluation prompt area 31 and an evaluation message 32 showing the past cognitive function evaluation value and the current cognitive function evaluation value. Fig.41 The cognitive function assessment prompt area 31 displays the cognitive function assessment performed once a month, the cognitive function assessment values in the past six months, and the cognitive function assessment value of this month.
[0384] The evaluation value of cognitive function is a value indicating the degree of cognitive function calculated by the prediction model. The value indicating the degree of cognitive function is expressed, for example, by 0.0 to 1.0. The evaluation result presentation unit 114 converts the value indicating the degree of cognitive function into a percentage and presents it as the evaluation value of cognitive function.
[0385] Furthermore, when displaying the evaluation value of the cognitive function in the past together with the evaluation value of the cognitive function at the present time, the cognitive function determination unit 113 stores the evaluation value of the cognitive function in the memory 12 .
[0386] Furthermore, the cognitive function evaluation prompt area 31 may also display whether the test subject has mild cognitive impairment as an evaluation result.
[0387] Furthermore, an evaluation message 32 such as "The risk of MCI has decreased compared to last month, and the condition is maintained in a good state. Please continue to live this way." is displayed. When the evaluation value of cognitive function this month is lower than the evaluation value of cognitive function last month, and the evaluation value of cognitive function this month is lower than 0.5, the evaluation result presentation unit 114 reads out the evaluation value from the memory 12. Fig.41 The evaluation message 32 shown is output to the display unit 3.
[0388] In addition, in the present embodiment, the evaluation value of the cognitive function in the past is displayed together with the evaluation value of the cognitive function in the present, but the present invention is not limited thereto, and only the evaluation value of the cognitive function in the present may be displayed. In this case, the cognitive function determination unit 113 may not store the evaluation value of the cognitive function in the memory 12.
[0389] Furthermore, the camera 2 of this embodiment may also be an anti-theft camera installed in front of the entrance, a camera sub-machine of an intercom, or a surveillance camera installed indoors. Furthermore, the display unit 3 may also be a monitor of a smart phone, a tablet computer, or an intercom.
[0390] In addition, in the present embodiment, the walking parameter detection unit 112 extracts the skeleton data based on the dynamic image data obtained from the camera 2, but the present invention is not limited to this, and the skeleton data can also be extracted using a motion capture system. The motion capture system can be any one of an optical type, a magnetic type, a mechanical type, and an inertial sensor type. For example, an optical motion capture system uses a camera to shoot a test object with labels attached to the joints, and detects the position of the label based on the shot image. The walking parameter detection unit 112 obtains the skeleton data of the test object from the position data detected by the motion capture system. As an optical motion capture system, for example, a three-dimensional motion analysis device manufactured by Inter Reha Co., Ltd. can be used.
[0391] Furthermore, the motion capture system may also include a depth sensor and a color camera, which can automatically extract the position information of the joints of the test subject from the video image and detect the test subject's posture. In this case, the test subject does not need to be labeled. Moreover, as such a motion capture system, for example, Kinect manufactured by Microsoft Corporation can be used.
[0392] When using a motion capture system to measure walking movements, it is best to extract the angle of the ankle joint, the angle of the knee joint or the vertical displacement of the waist during the walking movement based on the position coordinates, and detect the characteristic value of the walking movement based on the extracted angle or displacement.
[0393] In addition, in each of the above-mentioned embodiments, each component can be formed by dedicated hardware or can be implemented by executing a software program suitable for each component. Each component can also be implemented by allowing a program execution unit such as a CPU or a processor to read a software program recorded in a storage medium such as a hard disk or a semiconductor memory.
[0394] Part or all of the functions of the device involved in the embodiments of the present invention can be typically implemented as an integrated circuit LSI (Large Scale Integration). Part or all of these functions can be formed into chips separately, or can be formed into a chip that includes part or all of them. Moreover, the integrated circuit is not limited to LSI, and can also be implemented using a dedicated circuit or a general-purpose processor. It is also possible to use a programmable FPGA (Field Programmable Gate Array) that can be programmed after the LSI is manufactured or a reconfigurable processor that can reconstruct the connection or setting of the circuit unit inside the LSI.
[0395] Furthermore, part or all of the functions of the apparatus according to the embodiment of the present invention may be realized by causing a processor such as a CPU to execute a program.
[0396] Furthermore, the numbers used above are examples given to specifically illustrate the present invention, and the present invention is not limited to these exemplified numbers.
[0397] Moreover, the order in which the steps shown in the above flowchart are executed is only an example given to specifically illustrate the present invention, and the order other than the above can also be used within the scope of being able to obtain the same effect. Moreover, part of the above steps can also be executed simultaneously (in parallel) with other steps.
[0398] The technology related to the present invention has practical value as a technology for evaluating cognitive function based on the walking movements of a test subject because it can evaluate cognitive function simply and with high accuracy.
Claims
1. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; a detection unit that detects time series data of an angle of an ankle joint during a predetermined period of a foot-in-the-air period of one leg of the test subject from the walking data; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the angle of the ankle joint.
2. The cognitive function assessment device according to claim 1, characterized in that: The period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and when the one walking cycle is represented by 1% to 100%, The predetermined period is a period from 81% to 100% of the one walking cycle.
3. The cognitive function assessment device according to claim 2, characterized in that: The predetermined period is a period from 85% to 88% of the one walking cycle.
4. A cognitive function assessment device, which is a cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; a detection unit that detects, from the walking data, time series data of a first angle of an ankle joint during a first period of a stance period of one leg of the test subject and time series data of a second angle of the ankle joint during a second period of a foot-free period of the one leg; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the first angle of the ankle joint and an average value of the time series data of the second angle of the ankle joint.
5. The cognitive function assessment device according to claim 4, characterized in that: The period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and when the one walking cycle is represented by 1% to 100%, The first period is a period from 45% to 50% of the one walking cycle, The second period is a period from 85% to 88% of the one walking cycle.
6. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; a detection unit that detects time series data of an angle of a knee joint during a predetermined period of a stance phase of one leg of the test subject from the walking data; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the angle of the knee joint.
7. The cognitive function assessment device according to claim 6, characterized in that: The period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and when the one walking cycle is represented by 1% to 100%, The predetermined period is a period from 41% to 50% of the one walking cycle.
8. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; a detection unit that detects, from the walking data, time series data of an angle of a knee joint during a first period of a stance period of one leg of the test subject and time series data of an angle of an ankle joint during a second period of a foot-free period of the one leg; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the angle of the knee joint and an average value of the time series data of the angle of the ankle joint.
9. The cognitive function assessment device according to claim 8, characterized in that: The period from when one foot of the test subject touches the ground to when the foot touches the ground again is represented as one walking cycle, and when the one walking cycle is represented by 1% to 100%, The first period is a period from 41% to 50% of the one walking cycle, The second period is a period from 81% to 100% of the one walking cycle.
10. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; a detection unit that detects at least one of an angle of an ankle joint of one leg of the test subject and an angle of a knee joint of one leg and a displacement of a waist of the test subject in a vertical direction from the walking data; and The determination unit determines the degree of cognitive function of the test subject using at least one of the angle of the ankle joint and the angle of the knee joint and the displacement of the waist in the vertical direction.
11. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; An identification unit, for identifying the gender of the test subject; a detection unit that detects time series data of an angle of an ankle joint at an early stage of a stance phase of one leg of the test subject from the walking data when the test subject is identified as a male; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the angle of the ankle joint.
12. A cognitive function assessment device for assessing cognitive function based on the walking movements of a test subject, characterized in that: have: an acquisition unit, which acquires walking data related to the walking of the test subject; An identification unit, for identifying the gender of the test subject; a detection unit, when the test subject is identified as a male, detecting time series data of an angle of a knee joint of one leg of the test subject during a foot-in-the-air period from the walking data; as well as The determination unit determines the degree of cognitive function of the test subject using an average value of the time series data of the angle of the knee joint.
13. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; detecting time series data of the angle of the ankle joint during a prescribed period of a foot-in-the-air period of one leg of the test subject from the walking data; The degree of cognitive function of the test subject is judged using the average value of the time series data of the angle of the ankle joint.
14. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to have the following functions: acquiring walking data related to walking of the test subject; Detecting, from the walking data, time series data of a first angle of an ankle joint during a first period of a stance phase of one leg of the test subject and time series data of a second angle of the ankle joint during a second period of a foot-free phase of the one leg; The degree of cognitive function of the test subject is judged using the average value of the time series data of the first angle of the ankle joint and the average value of the time series data of the second angle of the ankle joint.
15. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; detecting time series data of the angle of the knee joint during a prescribed period of the stance phase of one leg of the test subject from the walking data; The degree of cognitive function of the test subject is determined using the average value of the time series data of the angle of the knee joint.
16. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; Detecting, from the walking data, time series data of the angle of the knee joint during a first period of the stance phase of one leg of the test subject and time series data of the angle of the ankle joint during a second period of the foot-in-the-air phase of the one leg; The degree of cognitive function of the test subject is judged using the average value of the time series data of the angle of the knee joint and the average value of the time series data of the angle of the ankle joint.
17. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; detecting at least one of an angle of an ankle joint of one leg of the test subject and an angle of a knee joint of one leg and a displacement of a waist of the test subject in a vertical direction from the walking data; The degree of cognitive function of the test subject is judged using at least one of the angle of the ankle joint and the angle of the knee joint and the displacement of the waist in the vertical direction.
18. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; Identify the sex of the test subject; When the test subject is identified as a male, detecting time series data of an angle of an ankle joint at an early stage of a stance phase of one leg of the test subject from the walking data; The degree of cognitive function of the test subject is judged using the average value of the time series data of the angle of the ankle joint.
19. A storage medium, which is a non-transitory computer-readable storage medium storing a cognitive function evaluation program for evaluating cognitive function based on walking movements of a test subject, characterized in that: The cognitive function assessment program enables the computer to: acquiring walking data related to walking of the test subject; Identify the sex of the test subject; When the test subject is identified as a male, detecting time series data of the angle of the knee joint of one leg of the test subject during a foot-in-the-air period from the walking data; The degree of cognitive function of the test subject is determined using the average value of the time series data of the angle of the knee joint.
