Knee joint replacement surgery necessity determination system, knee joint replacement surgery necessity determination method, and knee joint replacement surgery necessity determination program

Abstract

A knee joint replacement surgery feasibility determination system that can accurately determine whether knee joint replacement surgery is possible. [Solution] A knee joint replacement feasibility determination system 1 of the present invention includes a measurement device 10 worn by a user U and measuring triaxial angular velocity and triaxial acceleration, and an information processing device 20 that determines whether the user is eligible for knee joint replacement based on the triaxial angular velocity data and triaxial acceleration data. The information processing device 20 includes an acquisition unit 21 that acquires the triaxial angular velocity data and triaxial acceleration data from the measurement device 10 as gait data, an estimation unit 22 that estimates the knee joint inversion torque waveform and the knee joint extension torque waveform for one gait cycle of the user based on the acquired gait data using a knee joint torque estimation model that has been machine-learned in advance, and a determination unit 23 that determines whether the knee joint replacement is eligible based on the estimated knee joint inversion torque waveform and knee joint extension torque waveform using a knee joint torque determination model that has been machine-learned in advance.

Description

[Technical Field]

[0001] The present invention relates to the treatment of knee replacement. Necessity Knee replacement surgery Necessity Judgment system, knee replacement surgery Necessity Judgment method and knee replacement surgery Necessity This concerns the judgment program. [Background technology]

[0002] With the recent progress of the aging society, an increasing number of people are suffering from knee joint damage due to diseases such as osteoarthritis and rheumatoid arthritis, making it difficult for them to walk. When walking becomes difficult, people move less, which can lead to the risk of developing diseases such as cardiovascular disorders, diabetes, and kidney disorders, or dementia. Knee joint replacement surgery can be performed to treat knee joint damage, which can reduce the risk of developing the above diseases and dementia. For this reason, it is important to detect deterioration of the knee joint condition early and to consider the possibility of knee joint replacement surgery. Necessity The deterioration of the knee joint condition is reflected in the user's gait, and for example, techniques for determining abnormalities in a user's gait based on data acquired by a sensor while the user is walking are already known (see, for example, Patent Document 1 and Patent Document 2).

[0003] The information processing system described in Patent Document 1 acquires acceleration in the user's walking direction using a sensor attached to the user's foot, calculates feature values ​​based on the acquired acceleration, and determines whether the user has gait abnormalities based on the calculated feature values ​​and a trained model. The trained model is generated using a supervised machine learning method using a dataset including data indicating gait feature values ​​and ground truth labels containing information indicating the presence or absence of gait abnormalities and the severity of the gait abnormalities. This allows for the evaluation of the progression of the user's condition, comparisons before and after surgery, and the degree of recovery.

[0004] The gait analysis system of Patent Document 2 calculates reaction force direction information based on pressure information acquired from a pressure sensing element installed in the insole of a shoe, first knee three-dimensional angle information and second knee three-dimensional angle information acquired from inertial sensing elements installed above and below the user's knee, and a reaction force direction model. Knee joint torque is then calculated based on the pressure information, first knee three-dimensional angle information, second knee three-dimensional angle information, reaction force direction information, tibia length, and a knee joint torque model. Gait information is then determined based on the pressure information, first knee three-dimensional angle information or second knee three-dimensional angle information, and the gait model, and gait analysis results are calculated based on the gait information, knee joint torque, and gait model. The reaction force direction model, knee joint torque model, and gait model are constructed by machine learning using an analysis engine device. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Publication No. 2023-092150 [Patent Document 2] Japanese Patent Application Laid-Open No. 2017-144237 Summary of the Invention [Problem to be solved by the invention]

[0006] However, Patent Document 1 merely calculates a feature value based on the acceleration in the user's walking direction and estimates the severity of the knee condition using the calculated feature value and a trained model, and even if the severity of the knee condition can be estimated, it does not actually enable a determination as to whether or not knee joint replacement surgery is required. Furthermore, Patent Document 2 merely estimates the knee joint torque at the time of landing using pressure information acquired from a pressure sensing element and a knee joint torque model, and acquires pressure distribution information of the knee joint at the time of landing using the estimated knee joint torque at the time of landing and a gait analysis model, but it is not possible to determine whether or not knee joint replacement surgery is required based on the pressure distribution information alone. In other words, although the systems described in Patent Documents 1 and 2 evaluate and analyze gait abnormalities, they are unable to determine whether or not knee joint replacement surgery is required, that is, to determine whether or not knee joint replacement surgery is required based on gait abnormalities. Necessity There was a problem in that it was not possible to accurately determine

[0007] The present invention has been made to solve the above problems, and is a method for treating knee joint replacement. Necessity Knee replacement surgery that can accurately determine Necessity Judgment system, knee replacement surgery Necessity Judgment method and knee replacement surgery Necessity The purpose is to provide a judgment program. [Means for solving the problem]

[0008] Item 1. A measurement device that is worn by a user and measures three-axis angular velocity and three-axis acceleration while the user is walking; Based on the three-axis angular velocity data and three-axis acceleration data of the user acquired from the measuring device, Necessity and an information processing device for determining The information processing device includes: an acquisition unit that acquires the triaxial angular velocity data and the triaxial acceleration data from the measurement device as walking data of a user; an estimation unit that estimates a knee joint varus torque waveform and a knee joint extension torque waveform in one walking cycle of the user based on the walking data acquired by the acquisition unit, using a knee joint torque estimation model that has been machine-learned in advance to determine the correspondence between the triaxial angular velocity data and the triaxial acceleration data and the knee joint varus torque waveform and the knee joint extension torque waveform; The knee joint torque judgment model is pre-machined to determine the correspondence between the knee joint varus torque waveform and the knee joint extension torque waveform in one walking cycle and information on whether the knee joint is normal and does not require surgery or whether the knee joint is abnormal and requires surgery. Based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated by the estimation unit, the prediction of knee joint replacement surgery is performed. Necessity a determination unit for determining whether or not a knee joint replacement is necessary; Necessity Judging system.

[0009] Item 2. The knee joint replacement surgery according to Item 1, wherein the knee joint torque estimation model and the knee joint torque determination model are trained models generated by deep learning. Necessity Judging system.

[0010] Item 3. The knee joint replacement method according to Item 2, wherein the knee joint torque estimation model is a long short-term memory (LSTM; Long Short Term Memory) regression model. Necessity Judging system.

[0011] Item 4. The knee joint replacement surgery according to Item 2 or 3, characterized in that the knee joint torque determination model is a long short-term memory (LSTM) classification model. Necessity Judging system.

[0012] Item 5. A method for determining whether a user's knee joint replacement is performed by an information processing device having an acquisition unit, an estimation unit, and a determination unit. Necessity Knee replacement surgery Necessity A determination method comprising: an acquisition step in which the acquisition unit acquires, as walking data of the user, triaxial angular velocity data and triaxial acceleration data measured while the user is walking from a measurement device worn by the user; an estimation step in which the estimation unit estimates a knee joint varus torque waveform and a knee joint extension torque waveform for one walking cycle of the user based on the walking data acquired in the acquisition step, using a knee joint torque estimation model that has previously been machine-learned to determine a correspondence relationship between the triaxial angular velocity data and the triaxial acceleration data and the knee joint varus torque waveform and the knee joint extension torque waveform; The determination unit determines whether or not a knee joint replacement is necessary based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated in the estimation step, using a knee joint torque determination model that has been machine-learned in advance to determine a correspondence relationship between the knee joint varus torque waveform and the knee joint extension torque waveform in one walking cycle and information on whether the knee joint is normal and does not require surgery or whether the knee joint is abnormal and requires surgery. Necessity a determination step of determining whether or not the knee joint replacement Necessity Judgment method.

[0013] Section 6. User's knee replacement Necessity Knee replacement surgery Necessity A determination program, an acquisition step of acquiring, as walking data of the user, three-axis angular velocity data and three-axis acceleration data measured while the user is walking from a measurement device worn by the user; an estimation step of estimating a knee joint varus torque waveform and a knee joint extension torque waveform for one walking cycle of the user based on the walking data acquired in the acquisition step, using a knee joint torque estimation model that has been machine-learned in advance to determine the correspondence between the triaxial angular velocity data and the triaxial acceleration data and the knee joint varus torque waveform and the knee joint extension torque waveform; A knee joint torque judgment model that has been machine-learned in advance to determine the correspondence between the knee joint varus torque waveform and the knee joint extension torque waveform for one walking cycle and information on whether the knee joint is normal and does not require surgery or whether the knee joint is abnormal and requires surgery is used to determine the knee joint replacement based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated in the estimation step. Necessity and a determining step of determining whether or not the knee joint replacement is Necessity Judging program. [Effects of the Invention]

[0014] According to the present invention, Necessity can be accurately determined. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a block diagram showing the configuration of a knee joint replacement necessity determination system according to an embodiment of the present invention. [Figure 2] 2 is a diagram showing an example of a state in which a measurement device of the knee joint replacement necessity determination system of FIG. 1 is attached to a user. FIG. [Figure 3] 2 is a diagram showing the direction of torque estimated by the knee joint replacement necessity determination system of FIG. 1. FIG. [Figure 4] 1 is a flowchart showing a method for determining the necessity of knee joint replacement surgery according to one embodiment of the present invention. [Figure 5] 4 is a graph showing a torque waveform of the knee joint of a patient in the first embodiment. [Figure 6] 4 is a graph showing a knee joint torque waveform of a healthy subject in the first embodiment. [Figure 7] FIG. 10 is a diagram showing the classification results when the knee joint varus torque waveform and the knee joint extension torque waveform of the second embodiment are used. [Figure 8] FIG. 10 is a diagram showing the classification results when only the knee joint varus torque waveform is used in the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0016] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. Necessity The assessment system assesses abnormalities in the user's gait and determines whether or not knee joint replacement surgery is necessary based on the assessment results. Necessity This is a system that determines the following:

[0017] 1 to 3, a knee joint replacement procedure according to this embodiment will be described. NecessityThe determination system 1 will be described. Necessity 2 is a block diagram showing the configuration of the determination system 1. Necessity judgment 3 is a diagram showing an example of a state in which the measurement device 10 of the system 1 is attached to a user U. Necessity judgment 3A and 3B are diagrams showing the direction of torque estimated by the system 1. Fig. 3A shows the direction of knee joint varus torque, and Fig. 3B shows the direction of knee joint extension torque.

[0018] As shown in Figure 1, knee replacement surgery Necessity The determination system 1 includes a measurement device 10 and an information processing device 20. The measurement device 10 and the information processing device 20 are connected via a computer network such as the Internet or a WAN (Wide Area Network).

[0019] The measurement device 10 is a device that measures three-axis angular velocities and three-axis accelerations when the user U is walking. Specifically, the measurement device 10 measures three-axis accelerations, including acceleration in the walking direction of the user U, acceleration in a direction perpendicular to the walking direction of the user U on the horizontal plane, and acceleration in the vertical direction, as well as three-axis angular velocities, including angular velocities about the walking direction of the user U as an axis, angular velocities about a direction perpendicular to the walking direction of the user U on the horizontal plane as an axis, and angular velocities about the vertical direction as an axis.

[0020] The measurement device 10 is worn by the user U. In this embodiment, the measurement device 10 is worn on the feet of the user U. Specifically, the measurement device 10 is attached to the feet of the user U, for example, by being attached to the heel or insole of footwear worn by the user U. In FIG. 2, the measurement device 10 is worn on both feet of the user U, but it is sufficient that the measurement device 10 is attached to at least one foot of the user U that is experiencing discomfort in the knee. The measurement device 10 is, for example, a three-axis gyro sensor, a three-direction acceleration sensor, an inertial measurement unit (IMU), or the like.

[0021] The information processing device 20 determines the progress of knee joint replacement surgery for the user U based on the three-axis angular velocity data and three-axis acceleration data of the user U acquired from the measurement device 10. Necessity The information processing device 20 includes an acquisition unit 21, an estimation unit 22, a determination unit 23, a storage unit 24, and a control unit 25. The information processing device 20 is, for example, a computer such as a smartphone, a tablet, a microcomputer, a personal computer, or a server computer.

[0022] The acquisition unit 21 acquires three-axis angular velocity data and three-axis acceleration data from the measurement device 10 as walking data of the user U.

[0023] The estimation unit 22 uses the knee joint torque estimation model to estimate the knee joint varus torque waveform and the knee joint extension torque waveform of one walking cycle of the user U based on the walking data acquired by the acquisition unit 21. Here, one walking cycle is defined as one step from when one foot touches the ground until the foot on the same side touches the ground again, and the time taken for one step is defined as 100%. One walking cycle used by the estimation unit 22 is one step from when the foot that the user U feels discomfort in the knee touches the ground until it touches the ground again. Furthermore, the knee joint varus torque is a torque generated by bending the knee outward or inward when the user U is viewed from the front, as shown in FIG. 3(a). The knee joint extension torque is a torque generated by bending and straightening the knee, as shown in FIG. 3(b).

[0024] The knee joint torque estimation model is a trained model that has previously been machine-learned to determine the correspondence between the three-axis angular velocity data and three-axis acceleration data and the knee joint varus torque waveform and the knee joint extension torque waveform. Specifically, the knee joint torque estimation model is a trained model generated by deep learning, preferably a long short-term memory (LSTM) regression model. The estimation unit 22 extracts one walking cycle of walking data from the walking data acquired by the acquisition unit 21 based on the analysis results of peaks of angular velocity data and acceleration data that appear when the foot touches the ground, and performs low-pass filtering on the extracted data. At this time, the walking data for one walking cycle is distinguished as being from the right foot or the left foot based on differences in the angular velocity data and acceleration data that appear when the left and right feet move. The walking data from which high-frequency components have been removed is then input to the knee joint torque estimation model to estimate the time-series waveforms of the knee joint varus torque and the knee joint extension torque. In addition, the walking data for one walking cycle can be extracted by configuring the system so that the user U can select or input the leg on the knee side that feels uncomfortable before walking, and the walking data for one walking cycle of the leg on the knee side that feels uncomfortable can be extracted.

[0025] When multiple pieces of walking data for one walking cycle are extracted from the walking data, the estimation unit 22 selects a set of walking data in the middle of the time axis of the extracted sets of walking data for one walking cycle, and estimates the knee joint varus torque waveform and the knee joint extension torque waveform. In other words, if one walking cycle is measured five times, the estimation unit 22 uses the walking data for, for example, the third walking cycle in the middle.

[0026] The determination unit 23 determines the knee joint replacement of the user U based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated by the estimation unit 22 using the knee joint torque determination model. NecessityThe knee joint torque determination model is a trained model that has been machine-learned in advance to determine the correspondence between the knee joint varus torque waveform in one gait cycle and the knee joint extension torque waveform in one gait cycle, and information on whether the knee joint is normal and does not require surgery, or whether the knee joint is abnormal and requires surgery. Specifically, the knee joint torque determination model is a trained model generated by deep learning, and is preferably a long short-term memory (LSTM) classification model.

[0027] The knee joint torque judgment model is generated by deep learning of data collected from healthy individuals, with the knee joint varus torque waveform and knee joint extension torque waveform taken as normal examples (no surgery required), and the pre-operative knee joint varus torque waveform and knee joint extension torque waveform of individuals who have undergone knee joint replacement taken as abnormal examples (surgery required).Whether the individuals used for learning data are healthy or require knee joint replacement is determined by a doctor, taking into consideration the severity of the knee joint deformation, as well as the pain, stiffness, and inconvenience in daily life felt by the patient.The judgment unit 23 compares the knee joint varus torque waveform and knee joint extension torque waveform estimated by the estimation unit 22 in time series with the knee joint varus torque waveform and knee joint extension torque waveform whose correspondence has been learned by the knee joint torque judgment model, and determines whether the individuals are healthy or require knee joint replacement. Necessity Determine the following.

[0028] The storage unit 24 stores various information, data, programs, etc. The storage unit 24 is composed of a ROM (Read Only Memory) and a RAM (Random Access Memory). Necessity Knee replacement surgery Necessity The evaluation program is saved.

[0029] The control unit 25 is configured by a processor such as a CPU (Central Processing Unit). The control unit 25 controls the operations of the acquisition unit 21, the estimation unit 22, the determination unit 23, and the storage unit 24 by executing a program.

[0030] Next, referring to Figure 4, knee replacement NecessityThe determination system 1 determines whether the user U has a knee replacement. Necessity Knee replacement surgery Necessity The determination method will be explained. Necessity 10 is a flowchart showing a determination method. Necessity Each step of the determination method is performed by using the knee joint replacement surgery data stored in the memory unit 24 of the information processing device 20. Necessity This is performed by a decision program.

[0031] 4, while the user U is walking, the measurement device 10 attached to the feet of the user U measures three-axis angular velocity and three-axis acceleration (S10). When the user U finishes walking and measurement by the measurement device 10 is completed, the acquisition unit 21 executes an acquisition step of acquiring, from the measurement device 10, the three-axis angular velocity data and the three-axis acceleration data measured while the user U was walking, as walking data of the user U (S12). The acquired walking data is transmitted to the estimation unit 22.

[0032] When the estimation unit 22 receives the walking data from the acquisition unit 21, it executes an estimation step of estimating the knee joint varus torque waveform and knee joint extension torque waveform for one walking cycle of the user U based on the walking data acquired in the acquisition step, using the knee joint torque estimation model (S14). The estimated knee joint varus torque and knee joint extension torque are transmitted to the determination unit 23.

[0033] When the determination unit 23 receives the knee joint varus torque and the knee joint extension torque from the estimation unit 22, the determination unit 23 determines the knee joint replacement of the user U based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated in the estimation step using the knee joint torque determination model. Necessity The determination result is sent to the control unit 25 (S16).

[0034] When the control unit 25 receives the judgment result, it displays the judgment result on a display unit (for example, a display) of a user terminal (not shown) connected to the information processing device 20 via a computer network (S18). The user terminal is used to inform a user (patient, etc.) who feels discomfort in the knee joint or a person who is undergoing knee joint replacement surgery of the discomfort. Necessity The user terminal is a terminal used by a person (doctor, etc.) who judges whether or not the knee is in a good condition. The user terminal is, for example, a personal computer, a microcomputer, a tablet, a smartphone, or other device. The display unit of the user terminal displays messages such as, for example, "We recommend that you undergo knee replacement surgery for your right knee. Please consult a hospital or doctor immediately." or "Your right knee is in a condition where you should undergo knee replacement surgery. Please consult a hospital or doctor immediately." Then, Necessity Each process in the determination method is completed. [Example]

[0035] Next, knee joint replacement surgery according to the above embodiment Necessity The determination system 1 will be specifically described with reference to an example.

[0036] [First Example] A first embodiment will be described with reference to Figures 5 and 6. Figure 5 is a graph showing the knee joint torque waveform of a patient in the first embodiment. Figure 6 is a graph showing the knee joint torque waveform of a healthy subject in the first embodiment. In Figures 5 and 6, the knee joint varus torque waveform is shown in black, and the knee joint extension torque waveform is shown in gray. In addition, estimated values ​​are shown by solid lines, and actual measured values ​​are shown by dashed lines.

[0037] In the first example, the knee joint varus torque waveform and knee joint extension torque waveform are estimated or calculated during the gait cycle of a patient who feels discomfort in the knee joint and a healthy person, and the estimation accuracy of the knee joint varus torque waveform and knee joint extension torque waveform of the present invention is verified.

[0038] Example 1 shows the knee joint varus torque waveform and knee joint extension torque waveform estimated by the estimation unit 22 of the above embodiment based on the walking data acquired by having patients and healthy individuals wear the measuring device 10 and walk. Comparative Example 1 shows knee joint varus torque waveforms and knee joint extension torque waveforms calculated based on a conventional joint torque calculation method when a patient and a healthy subject walked in a space with a floor reaction force meter installed on the floor and a motion capture device installed on the ceiling. In the conventional calculation method, knee joint varus torque and knee joint extension torque are calculated from the floor reaction force measured by the floor reaction force meter and the movement of the knee joint and body inertia information captured by motion capture.

[0039] The estimation results of Example 1 and the calculation results of Comparative Example 1 are shown in Figures 5 and 6. In Figures 5 and 6, the knee joint varus torque waveform of Example 1 is shown by a black solid line, and the knee joint extension torque waveform of Example 1 is shown by a gray solid line. In addition, the knee joint varus torque waveform of Comparative Example 1 is shown by a black dashed line, and the knee joint extension torque waveform of Comparative Example 1 is shown by a gray dashed line.

[0040] 5 and 6, for both the patient and the healthy subject, the knee joint varus torque waveform and the knee joint extension torque waveform are substantially the same, with an error of less than 5% between Example 1 and Comparative Example 1. This demonstrates that the estimation unit 22 of the above embodiment accurately estimates the knee joint varus torque waveform and the knee joint extension torque waveform.

[0041] Furthermore, comparing Figures 5 and 6, it can be seen that both the knee joint varus torque waveform and the knee joint extension torque waveform are significantly different between patients and healthy subjects. Therefore, a knee joint torque determination model can be generated by performing deep learning on the knee joint varus torque waveform and the knee joint extension torque waveform in association with information on whether the subject is a patient or a healthy subject. Then, by using the knee joint torque determination model generated by deep learning, it is possible to determine the outcome of knee joint replacement surgery from the estimated knee joint varus torque waveform and knee joint extension torque waveform. Necessity It can be seen that it is possible to determine

[0042] [Second Example] A second embodiment will be described with reference to Figures 7 and 8. Figure 7 shows the classification results when a knee joint varus torque waveform and a knee joint extension torque waveform are used in the second embodiment. Figure 8 shows the classification results when only the knee joint varus torque waveform is used in the second embodiment.

[0043] In the second example, three healthy subjects and four patients (those who are to undergo knee joint replacement surgery) were asked to walk, and walking data for one walking cycle of 85 steps was obtained from the three healthy subjects, and walking data for one walking cycle of 55 steps was obtained from the four patients, and the results were used to evaluate the effectiveness of knee joint replacement surgery of the present invention based on gait abnormalities. Necessity Verify the accuracy of the judgment.

[0044] In Example 2, a healthy person and a patient are asked to wear the measurement device 10 and walk, and the determination unit 23 of the above embodiment determines whether or not the user U will undergo knee joint replacement surgery based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated from the acquired walking data. Necessity The following is determined. In Comparative Example 2, the same healthy subjects and patients as in Example 2 were asked to wear the measuring device 10 and walk, and the determination unit 23 of the above embodiment determined the knee joint varus torque waveform estimated from the acquired walking data. Necessity The following is determined. In both Example 2 and Comparative Example 2, walking data for one walking cycle of 85 steps for a healthy person and 55 steps for a patient was extracted from the acquired walking data, and the walking data for the user U during knee joint replacement surgery were used. Necessity The following is determined.

[0045] The results of the judgments in Example 2 and Comparative Example 2 are shown in Figures 7 and 8. In Figures 7 and 8, the classes into which healthy subjects and patients should actually be classified are plotted on the vertical axis "true classes," and the results of the judgments in Example 2 and Comparative Example 2 are shown in Figures 7 and 8. Necessity Knee replacement surgery performed by Judgment System 1 NecessityThe judgments are plotted on the horizontal axis, "judged class." If it is judged that knee joint replacement is necessary, it is plotted as "patient" (patient; surgery applicable), and if it is judged that knee joint replacement is not necessary, it is plotted as "healthy" (healthy). In Figures 7 and 8, if the "true class" is "patient," the "judged class" is also plotted as "patient," and if the "true class" is "healthy," the "judged class" is also plotted as "healthy." In other words, if the plots are in the upper left and lower right of Figures 7 and 8, it can be determined that the judgment has been made accurately. Figures 7 and 8 show the values ​​plotted within the corresponding boxes.

[0046] As can be seen from FIGS. 7 and 8, when only the knee joint varus torque waveform of Comparative Example 2 is used, the knee joint replacement of the present invention is not performed on a patient who actually requires knee joint replacement. Necessity There were cases where the judgment system 1 judged that knee joint replacement was unnecessary (cases plotted at the bottom left of Figure 8). Necessity There were also cases where the determination system 1 determined that knee joint replacement surgery was necessary (cases plotted in the upper right of Figure 8). In other words, when only the knee joint varus torque waveform of Comparative Example 2 was used, it was found that there were cases where it was not possible to accurately determine whether or not knee joint replacement surgery was necessary. On the other hand, when the knee joint varus torque waveform and knee joint extension torque waveform of Example 2 were used, it was possible to accurately determine whether or not knee joint replacement surgery was necessary. Necessity The results are in perfect agreement with those of Judgment System 1, and show high accuracy in the assessment of knee replacement surgery. Necessity It can be seen that the above can be determined.

[0047] As described above, the knee joint replacement surgery of the present invention Necessity The determination system 1 estimates the knee joint varus torque waveform and the knee joint extension torque waveform based on walking data acquired while the user U is walking. Then, based on the estimated knee joint varus torque waveform and knee joint extension torque waveform, the determination system 1 determines whether the user U will undergo knee joint replacement surgery. Necessity As shown in the second example, the knee joint varus torque waveform and the knee joint extension torque waveform are used to determine the degree of knee replacement. NecessityBy determining whether or not knee joint replacement surgery is necessary for a person who needs knee joint replacement surgery, which is particularly problematic, it is possible to obtain a determination result that is appropriate for the condition of the user U, and to determine whether or not knee joint replacement surgery is necessary. Necessity can be accurately determined.

[0048] Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications are possible without departing from the spirit of the present invention.

[0049] For example, in the above embodiment, when multiple walking data for one walking cycle are extracted from the walking data, the estimation unit 22 selects a set of walking data in the middle of the time axis of the extracted set of walking data for one walking cycle to estimate the knee joint varus torque waveform and the knee joint extension torque waveform, but this configuration is not necessarily required. For example, when multiple walking data for one walking cycle are extracted, the walking data for all one walking cycle may be used to estimate the waveforms of multiple knee joint replacement surgeries. Necessity In this case, the information processing device 20 may further include an evaluation unit that evaluates whether "surgery is necessary" or "surgery is not necessary" based on the multiple determination results, and the evaluation unit finally determines whether or not knee joint replacement is necessary. Necessity The evaluation unit may, for example, use the result of the evaluation using a set of walking data in the middle of the time axis as the evaluation result, and perform an evaluation by adding additional information from the other evaluation results. Specifically, if the third evaluation is "surgery required" and there is a determination that "surgery is not required" among the other evaluations, or conversely, if the third evaluation is "surgery is not required" and there is a determination that "surgery is required" among the other evaluations, an evaluation can be performed such that "improvement of walking through rehabilitation" is added.

[0050] Furthermore, in the above embodiment, the measurement device 10 is attached to the foot of the user U, but this configuration is not necessarily required. For example, the measurement device 10 may be attached to the waist or upper body of the user U. Furthermore, the measurement device 10 may be configured as a mobile terminal such as a smartphone owned by the user U, and may be attached to the user U by being placed in a pocket of the user U's clothes or hanging from the user U's neck. [Industrial Applicability]

[0051] Knee joint replacement surgery of the present invention Necessity Judgment system, knee replacement surgery Necessity Judgment method and knee replacement surgery Necessity The judgment program is used when the user feels discomfort in the knee. It measures the user's walking data and uses deep learning to determine whether the user's knee is in a condition that requires knee joint replacement, i.e., whether the user should undergo knee joint replacement surgery. Necessity The present invention has industrial applicability because it is used to determine whether a knee joint replacement is necessary and, when a message indicating that knee joint replacement surgery is required is displayed, to encourage the user to visit a doctor for knee joint replacement surgery. [Explanation of symbols]

[0052] 1 Knee replacement surgery Necessity Judgment System 10. Measuring equipment 20 Information processing equipment 21 Acquisition Department 22 Estimation part 23 Judgment section U User

Claims

1. a measurement device that is worn by a user and that measures three-axis angular velocity and three-axis acceleration while the user is walking; an information processing device that determines whether or not a user needs knee joint replacement surgery based on the user's three-axis angular velocity data and three-axis acceleration data acquired from the measurement device; The information processing device includes: an acquisition unit that acquires the triaxial angular velocity data and the triaxial acceleration data from the measurement device as walking data of a user; an estimation unit that estimates a knee joint varus torque waveform and a knee joint extension torque waveform of one walking cycle of the user based on the walking data acquired by the acquisition unit, using a knee joint torque estimation model that has been previously deep-learned to determine a correspondence relationship between triaxial angular velocity data and triaxial acceleration data and a knee joint varus torque waveform and a knee joint extension torque waveform; A knee joint replacement necessity determination system characterized by having a determination unit that determines whether knee joint replacement is necessary based on the knee joint varus torque waveform and knee joint extension torque waveform estimated by the estimation unit using a knee joint torque determination model generated by deep learning of data collected in which the knee joint varus torque waveform and knee joint extension torque waveform of healthy individuals are used as normal examples and the pre-operative knee joint varus torque waveform and knee joint extension torque waveform of individuals who have undergone knee joint replacement surgery are used as abnormal examples.

2. The knee joint replacement surgery necessity determination system according to claim 1 , wherein the knee joint torque estimation model is a long short-term memory (LSTM) regression model.

3. The knee joint replacement surgery necessity determination system according to claim 1 , wherein the knee joint torque determination model is a long short-term memory (LSTM) classification model.

4. A method for determining whether a user needs knee joint replacement surgery, the method being executed by an information processing device having an acquisition unit, an estimation unit, and a determination unit, comprising: an acquisition step in which the acquisition unit acquires, as walking data of the user, three-axis angular velocity data and three-axis acceleration data measured while the user is walking from a measurement device worn by the user; an estimation step in which the estimation unit estimates a knee joint varus torque waveform and a knee joint extension torque waveform for one walking cycle of the user based on the walking data acquired in the acquisition step, using a knee joint torque estimation model that has been previously deep-learned to determine a correspondence relationship between three-axis angular velocity data and three-axis acceleration data and a knee joint varus torque waveform and a knee joint extension torque waveform; a determination step in which the determination unit determines whether knee joint replacement is necessary based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated in the estimation step using a knee joint torque determination model generated by deep learning of data collected in which the knee joint varus torque waveform and the knee joint extension torque waveform of a healthy person are used as normal examples and the pre-operative knee joint varus torque waveform and the knee joint extension torque waveform of a person who has undergone knee joint replacement are used as abnormal examples.

5. A knee joint replacement surgery necessity determination program for determining whether a user needs knee joint replacement surgery, an acquisition step of acquiring, as walking data of the user, three-axis angular velocity data and three-axis acceleration data measured while the user is walking from a measurement device worn by the user; an estimation step of estimating a knee joint varus torque waveform and a knee joint extension torque waveform for one walking cycle of the user based on the walking data acquired in the acquisition step, using a knee joint torque estimation model that has been previously deep-learned to determine the correspondence between the three-axis angular velocity data and the three-axis acceleration data and the knee joint varus torque waveform and the knee joint extension torque waveform; a judgment step of judging the necessity of knee joint replacement based on the knee joint varus torque waveform and the knee joint extension torque waveform estimated in the estimation step, using a knee joint torque judgment model generated by deep learning of data collected in which the knee joint varus torque waveform and the knee joint extension torque waveform of a healthy person are used as normal examples and the pre-operative knee joint varus torque waveform and the knee joint extension torque waveform of a person who has undergone knee joint replacement are used as abnormal examples.

Images