JP6982324B2 - Finger operation support device and support method - Google Patents

Description

The present invention relates to a finger operation support device and a support method for supporting the improvement of skills such as operations using scissors of a barber / cosmetologist to the same level as a skilled person, so as to prevent occupational diseases caused by immature finger operations. It is the one that was made.

The following patent document discloses a system in which a patient with Parkinson's disease or the like is made to perform an action of tracking a moving target on the screen, the muscle activity of the patient at that time is detected, and the motor function of the patient is evaluated. There is. The detection of muscle activity is performed by attaching a pair of dish-shaped electrodes at intervals on the muscle belly of the muscle involved in the subject's movement and extracting a myoelectric signal from the dish-shaped electrodes. In the following patent documents, the success rate of target tracking is displayed on the y-axis and the component of muscle activity is on the x-axis, and the motor functions of normal subjects, Parkinson's disease patients and cerebral disease patients are displayed in different regions (clusters). It is stated that this label makes it possible to diagnose the disease.

Japanese Patent No. 514458

In the present invention, the muscle activity of the subject is detected in order to identify the skill of finger operation.

It is known that barbers and cosmetologists who use scissors for a long time are more likely to develop tendonitis and have a higher prevalence of carpal tunnel syndrome than other occupations.
The present inventor analyzes the muscle activities of the flexor pollicis longus muscle group (the muscle group involved in flexing the thumb) and the wrist extensor muscle group (the muscle group involved in returning the wrist) during the scissor operation. , I found that there are differences in how to use muscles, how to bend wrists, muscle tension, etc. between skilled barbers who do not have pain in their hands and beginners.

FIG. 5 shows the results of measuring the electromyogram and the opening degree of the scissors when the scissors operation for cutting the hair was simulated with the skilled person (a) and the beginner (b) as subjects. .. The electromyogram shows the muscle activity of the thenar eminence (the muscle that constitutes the bulge at the base of the thumb) in the flexor pollicis longus muscle group. The opening degree of the scissors is measured by a rotation sensor attached to the hinge of the scissors.
As is clear from FIG. 5, the scissors opening degree of the beginner (b) is not constant, and the value of the myoelectric potential is high. On the other hand, the expert (a) has a constant scissors opening degree, and the myoelectric potential is rhythmically active along the scissors opening degree.
Therefore, if the beginner's finger operation skill reaches the level of an expert, it can be expected that the anxiety of tendonitis and carpal tunnel syndrome will be released.

The present invention was devised based on such knowledge, and provides a finger operation support device that supports improvement of finger operation skill so as to prevent a disease that develops due to immaturity of finger operation, and provides a support method thereof. The purpose is to provide.

The present invention is a finger operation support device that supports improvement of the skill of finger operation performed by using scissors of a person involved in barber or beauty, and is a group of thumb flexors and hands involved in finger operation of a subject who operates scissors. Myoelectric information acquisition means for acquiring myoelectric information of joint extensor muscles, tools for detecting the opening degree information of scissors operated by the subject as tools exercise information, exercise information acquisition means, myoelectric information and tools using motion information to determine the proficiency of the subject's finger operation, and a, a skill judging means for transmitting a determination result. Then, the skill determination means is a machine learning using machine learning using the myoelectric information and tool movement information of a skilled person in the finger operation, and the myoelectric information and tool movement information of an inexperienced person as training data (teacher data). A classifier that classifies the subject and the immature person is constructed, and it is configured to judge whether the subject is an expert or an inexperienced person from the myoelectric information and the tool movement information when the subject performs a finger operation using scissors. is doing.
Therefore, the subject can recognize the degree of his / her skill, and by practicing the finger operation aiming at the skill level, the skill is improved and the occupational disease caused by immaturity can be prevented.

Further, in the finger operation support device of the present invention, the skill determination means can be configured by a support vector machine (SVM).
The SVM sets the boundary line between the area of the expert and the area of the inexperienced so that the width of the margin for separating the expert and the inexperienced person is maximized.

Further, in the hand operation support device of the present invention, feature data is extracted from myoelectric information and tool motion information by a convolutional neural network (CNN), and SVM uses the feature data to be experienced and inexperienced. And identify with.
AI (artificial intelligence) can be used to effectively distinguish between skilled and inexperienced people.

Further, the present invention provides a finger operation support method for supporting an enhanced skill in finger operation performed using those scissors involved in barber or beauty, the finger operation using scissors to those skilled and novice in finger operation The training data acquisition step for detecting the myoelectric information of the thumb flexor muscle group and the wrist extensor muscle group involved in the finger operation and the opening information of the scissors , and the myoelectric information obtained in the training data acquisition step. And the data of the opening information is input to the computer, the computer performs machine learning using the above data, and the classifier construction step of constructing the classifier for classifying the skilled person and the inexperienced person, and the finger operation using the scissors. A determination step in which data of the subject's myoelectric information and opening information are input to the computer, the computer determines whether the subject is an expert or an inexperienced person using a classifier, and the determination result is transmitted.
It is equipped with.
Therefore, the subject who recognizes the degree of his / her skill can improve his / her skill and prevent occupational diseases by practicing the finger operation aiming at the skill level.

INDUSTRIAL APPLICABILITY According to the present invention, diseases caused by immaturity of finger operation can be prevented.

A block diagram showing a configuration of a finger operation support device according to an embodiment of the present invention. The figure explaining the operation of SVM The flow chart which shows the procedure of the finger operation support method which concerns on embodiment of this invention. The figure which shows the state of practicing the finger operation by the finger operation support device of FIG. A diagram showing the difference between the electromyogram and the scissors opening of an expert and a beginner of scissors operation.

As an embodiment of the present invention, a finger operation support device that supports improvement of the skill of a barber or a beautician who operates scissors will be described.
As shown in FIG. 1, this device detects and analyzes the myoelectric information analysis device 20 that acquires and analyzes the myoelectric information of the muscles of the subject who operates the scissors, and the movement information of the scissors operated by the subject. A skill determination device 10 for determining a subject's proficiency in scissors operation using the scissors analysis device 30 and the scissors motion information analyzed by the myoelectric information analysis device 20 and the scissors motion information analyzed by the scissors analysis device 30. And have.

The myoelectric information analysis device 20 includes a flexor pollicis longus information acquisition unit 21 for acquiring the myoelectric information of the flexor pollicis longus muscle group of the subject who operates the scissors, and a hand for acquiring the myoelectric information of the wrist extensor muscle group of the subject. It includes a joint extensor muscle group information acquisition unit 22, a myoelectric information recording unit 23 that records myoelectric information, and a myoelectric information analysis unit 24 that analyzes the acquired myoelectric information.

The flexor pollicis longus information acquisition unit 21 acquires a myoelectric signal from a dish-shaped electrode attached to the muscle of the flexor pollicis longus muscle of the subject, and the wrist extensor muscle group information acquisition unit 22 acquires the wrist extensor muscle group of the subject. The myoelectric signal is acquired from the dish-shaped electrode attached to the muscle. These myoelectric signals are amplified and input to the flexor pollicis longus muscle group information acquisition unit 21 and the wrist extensor muscle group information acquisition unit 22. These myoelectric signals are once stored in the myoelectric information recording unit 23, and then analyzed by the myoelectric information analysis unit 24.

The myoelectric information analysis unit 24 is configured by artificial intelligence (AI) that analyzes myoelectric information using CNN. CNN is a tool that extracts feature points of input data by repeating "convolution processing" that condenses feature points and "pooling process" that reduces data while leaving feature information. AI can learn an appropriate filter for convolution by repeating machine learning of feature point extraction using CNN, and it becomes possible to efficiently extract feature data from input data.
The feature data of the myoelectric information extracted by the myoelectric information analysis unit 24 is recorded in the myoelectric information recording unit 23 and sent to the skill determination device 10.

The scissors analysis device 30 includes a blade motion information acquisition unit 31 that acquires information on the opening degree of the scissors operated by the subject, a blade motion information recording unit 32 that records the opening degree information of the scissors, and the acquired scissors opening degree information. It is provided with a blade motion information analysis unit 33 for analyzing the above.

When the subject operates the scissors, the blade motion information acquisition unit 31 acquires the opening degree information of the scissors from the rotation sensor attached to the hinges of the two blades of the scissors. The rotation sensor electromagnetically detects and transmits the opening degree of the scissors. This scissors opening degree information is temporarily stored in the blade motion information recording unit 32, and then analyzed by the blade motion information analysis unit 33.

The blade motion information analysis unit 33 is configured by artificial intelligence (AI) that analyzes scissors opening information using CNN. AI uses CNN to extract feature data of scissors opening information and sends it to the skill determination device 10. This feature data is also recorded in the blade motion information recording unit 32.

The skill determination device 10 includes an information recording unit 11 that records the feature data of the myoelectric information sent from the myoelectric analysis device 20 and the feature data of the scissors opening information sent from the scissors analysis device 30, and the scissors of the subject. It includes a scissors proficiency level analysis unit 12 that analyzes the proficiency level of the operation, and a scissors proficiency level display unit 13 that displays the proficiency level of the subject analyzed by the scissors proficiency level analysis unit 12.

The scissors proficiency analysis unit 12 is configured by artificial intelligence (AI) that analyzes the proficiency level of scissors operation using SVM. As shown in FIG. 2, SVM is a tool for setting a boundary line between the data belonging to class 1 and class 2 existing in the feature space so that the width of the margin is maximized.

AI uses the characteristic data of myoelectric information and the characteristic data of scissors opening information of a large number of skilled and inexperienced persons as training data, and by SVM, the margin between the expert class and the inexperienced person class is set. Repeat machine learning to set boundaries to maximum.
For example, in the feature space where the vertical axis of FIG. 2 is the feature data of the myoelectric information and the horizontal axis is the feature data of the scissors opening information, the positions of the skilled and inexperienced persons included in the training data are set in the myoelectric information. Plot based on the value of the feature data and the value of the feature data of the scissors opening information, and set the boundary line on FIG. 2 so that the margin between the expert class and the inexperienced class is maximized.
The boundary line in the feature space set by the scissors proficiency analysis unit 12 is recorded in the information recording unit 11.

After that, when the feature data of the subject's myoelectric information is sent from the myoelectric analysis device 20 and the feature data of the subject's scissors opening information is sent from the scissors analysis device 30, the scissors proficiency analysis unit 12 determines. The position of the subject on the feature space is plotted on FIG. 2 based on the value of the feature data, and if the position is in the area of the expert side of the boundary line, the subject has reached the area of the expert. It is displayed on the skill level display unit 13, and if the position on the feature space of the subject is in the area on the immature side of the boundary line, the scissors skill level display unit 13 displays that the subject is an immature person. At this time, the shortest distance from the position on the feature space of the subject to the boundary line may be displayed as a value indicating the skill level (or immaturity level).

FIG. 3 shows a flow in which a subject practices scissors operation using this finger operation support device. It is assumed that the information recording unit 11 of the skill determination device 10 records the boundary line between the skilled person and the inexperienced person in the feature space obtained by machine learning based on the prior training data.
The subject attaches electrodes to the flexor pollicis longus muscle group and the wrist extensor muscle group to extract myoelectric signals, and a rotation sensor moves the scissors attached to the hinge (step 1).
FIG. 4 shows the situation at this time.

The scissors analysis device 30 sends the feature data of the analyzed scissors opening degree information to the skill determination device 10, and the myoelectric analysis device 20 sends the feature data of the analyzed myoelectric information to the skill determination device 10 (step 2).
The skill determination device 10 records the sent data in the information recording unit 11, compares it with the boundary line of the feature space, and analyzes the proficiency level (step 3).
The scissors proficiency display unit 13 displays whether the subject's scissors operation is at the level of a skilled person or an inexperienced person, and displays a value indicating the proficiency level (step 4).

If the subject intends to practice the scissors operation (Yes in step 5), the subject attaches electrodes to the thumb flexor muscle group and the wrist extensor muscle group, and a rotation sensor is attached to the hinge as in step 1. Move the scissors (step 6).
Next, the processes of step 2, step 3, and step 4 are performed.
The subject confirms the value of the scissors skill level and adjusts the movement of the finger to be operated (step 7).

If the subject does not reach the expert's area even after practicing (No in step 8), the process returns to step 5. When the subject reaches the expert's area in step 8 (Yes in step 8), the process proceeds to step 11, and when the subject continues to practice (Yes in step 11), the process proceeds to step 5.
When the practice is finished in step 5 or step 11, the myoelectric potential and the blade movement at the time of scissors operation are recorded in the information recording unit 11 (step 12), and the practice is finished.

In this way, the subject improves his / her skill by practicing the operation of the scissors using this finger operation support device.
In addition, in step 4, when the subject is an inexperienced person, the image of the scissors operation of a skilled person may be displayed together.

Here, the operation of scissors has been described, but the present invention can also be applied to other finger operations. For example, it can be applied to the keyboard operation of a personal computer and the operation of shaking a pot when cooking.

The finger operation support device of the present invention can be used in a wide range of fields in order to improve the skill level of those involved in various finger operations.

10 Skill judgment device 11 Information recording unit 12 Scissors proficiency analysis unit 13 Scissors proficiency display unit 20 Myoelectric information analysis device 21 Mother finger flexor muscle group information acquisition unit 22 Wrist joint extensor muscle group information acquisition unit 23 Myoelectric information recording unit 24 Myoelectric information analysis unit 30 Scissors analysis device 31 Blade motion information acquisition unit 32 Blade motion information recording unit 33 Blade motion information analysis unit

Claims (4)

It is a finger operation support device that supports the improvement of the skill of finger operation performed by using scissors of a person involved in barber or beauty.
A myoelectric information acquisition means for acquiring myoelectric information of the flexor pollicis longus muscle group and the wrist extensor muscle group involved in the finger operation of the subject who operates the scissors, and
Tool movement information acquisition means for detecting information on the opening degree of the scissors operated by the subject as tool movement information, and
A skill judging means for using said myoelectric information and utensils motion information to determine the proficiency of finger manipulation of the subject, conveys a determination result,
Equipped with
The skill determination means is a machine learning using the myoelectric information and tool movement information of a skilled person in the finger operation and the myoelectric information and tool movement information of an inexperienced person as training data, and is inexperienced with a skilled person. A classifier for classifying a person is constructed, and it is determined whether the subject is an expert or an inexperienced person from the myoelectric information and the tool movement information when the subject performs a finger operation using the scissors. Hand operation support device. The finger operation support device according to claim 1.
A finger operation support device having a support vector machine (SVM) as the skill determination means. The finger operation support device according to claim 2.
The SVM is a finger operation support device that distinguishes between a skilled person and an inexperienced person by using feature data extracted from the myoelectric information and tool motion information by a convolutional neural network (CNN). It is a finger operation support method that supports the improvement of the skill of finger operation performed using scissors of a person involved in barber or beauty.
Said to perform the finger operation using scissors to those skilled and novice in the finger operation, the thumb muscle group involved in the hand operation and wrist extensor muscles of myoelectric information and detects the opening information of the scissors Training data acquisition steps and
A classifier that inputs the data of the myoelectric information and the opening degree information obtained in the training data acquisition step into a computer, and the computer performs machine learning using the data to classify skilled persons and inexperienced persons. The classifier building steps to build and
The data of the myoelectric information and the opening information of the subject who performs the finger operation using the scissors is input to the computer, and the computer determines whether the subject is an expert or an inexperienced person using the classifier. Judgment step to convey the judgment result and
Finger operation support method equipped with.

Images