CN108008821B - Performance evaluation method, device, terminal and storage medium of artificial limb action classifier - Google Patents
Performance evaluation method, device, terminal and storage medium of artificial limb action classifier Download PDFInfo
- Publication number
- CN108008821B CN108008821B CN201711336803.4A CN201711336803A CN108008821B CN 108008821 B CN108008821 B CN 108008821B CN 201711336803 A CN201711336803 A CN 201711336803A CN 108008821 B CN108008821 B CN 108008821B
- Authority
- CN
- China
- Prior art keywords
- test
- current
- action
- track
- evaluation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
- G06F3/015—Input arrangements based on nervous system activity detection, e.g. brain waves [EEG] detection, electromyograms [EMG] detection, electrodermal response detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Psychiatry (AREA)
- Multimedia (AREA)
- Social Psychology (AREA)
- Biomedical Technology (AREA)
- Dermatology (AREA)
- Neurology (AREA)
- Neurosurgery (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
- Prostheses (AREA)
Abstract
The invention discloses a performance evaluation method, a performance evaluation device, a performance evaluation terminal and a performance evaluation storage medium of a prosthesis action classifier. The method comprises the following steps: determining one test item in a plurality of preset test items as a current test item; displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time; determining a test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item; and integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes. The invention realizes the real-time evaluation of the performance of the artificial limb action classifier, and obtains a quantitative evaluation result which is more accurate.
Description
Technical Field
The embodiment of the invention relates to a prosthesis control technology, in particular to a performance evaluation method, a device, a terminal and a storage medium of a prosthesis action classifier.
Background
According to the data of national statistical office on the sampling survey of the disabled population in China in 2006, the disabled population in China is the largest, namely 2412 thousands of people, and accounts for 29.07 percent of the total disabled population. The population of disabled limbs is increasing due to accidents, congenital defects, diseases, natural disasters, etc. In the world earthquake of Wenchuan in Sichuan in 5 months of 2008, 2 million people lost their limbs. These amputees, in turn, need to wear artificial limbs to assist their daily lives and tasks. At present, the commercialized artificial limbs at home and abroad include a mechanical cable-controlled artificial limb, a myoelectric artificial limb and a myoelectric cable-controlled hybrid artificial limb according to the control mode of the artificial limb.
The traditional mechanical cable-controlled artificial limb has the problems of single function, slow operation, clumsy action, difficult maintenance and the like. Electromyogram (EMG) signals recorded from the surface of a limb have been widely used in the control of artificial limbs for a last decade. The most important components in myoelectric prostheses are electrodes and motion classifiers. The electrodes are used for collecting electric signals of muscles of the amputee, the electric signals are transmitted to the trained action classifier to be recognized and judged, then the action classifier outputs an action which is considered to be the closest action, and finally the motor drives the artificial limb to finish the action. The myoelectric artificial limb is used as an intuitive control method, an amputee can control the artificial limb like controlling own hand, and extra learning burden is not needed, so the myoelectric artificial limb is the most promising method for controlling the artificial limb at present.
After wearing the myoelectric prosthesis, since the muscle characteristics of each person are different, training of a prosthesis motion classifier in the myoelectric prosthesis is required to make the amputee fit the wearer. However, the myoelectric artificial limb in the market at present can only evaluate a classification method and a training effect by using off-line data analysis, and lacks a quantifiable real-time evaluation means, so that an amputee can only judge the control condition by data prediction and subjective observation of the control effect of the trained myoelectric artificial limb. The off-line data analysis can provide objective data, but cannot completely reflect the specific situation in the real-time myoelectric artificial limb control. In the actual use of the myoelectric prosthesis, there may be a case that the action made by the myoelectric prosthesis is not in accordance with the intention action of the amputee, or even the action that the amputee wants to complete cannot be completed at all. This can present a certain accident and risk to the amputee. In summary, the existing myoelectric artificial limb needs a model and a means for real-time, objective and effective evaluation of the training condition of the artificial limb motion classifier.
Disclosure of Invention
In view of this, embodiments of the present invention provide a performance evaluation method, apparatus, terminal and storage medium for a prosthesis motion classifier, so as to implement real-time evaluation of the performance of the prosthesis motion classifier.
In a first aspect, an embodiment of the present invention provides a performance evaluation method for a prosthesis motion classifier, where the method includes:
determining one test item in a plurality of preset test items as a current test item;
displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
determining a test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item;
and integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes.
In a second aspect, an embodiment of the present invention further provides a performance evaluation apparatus for a prosthesis motion classifier, where the apparatus includes:
the test item selection module is used for determining one test item in a plurality of preset test items as a current test item;
the action track drawing module is used for displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthetic limb action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
the test value determining module is used for determining the test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item;
and the evaluation value determining module is used for integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes.
In a third aspect, an embodiment of the present invention further provides a terminal, where the terminal includes:
one or more processors;
a storage device for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors implement the performance evaluation method of the prosthesis motion classifier according to any embodiment of the present invention.
In a fourth aspect, an embodiment of the present invention further provides a computer storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the performance evaluation method of the prosthesis motion classifier according to any embodiment of the present invention.
According to the technical scheme of the embodiment of the invention, the starting point and the test target are displayed in the current test item, the electromyographic signal of the testee is continuously acquired, the electromyographic signal is identified by the artificial limb action classifier to obtain the result action, the action track corresponding to the result action is drawn, the test time is recorded, the test value of the current test item aiming at the test evaluation index is determined according to the action track and the test time, the test values of a plurality of test items are integrated to obtain the evaluation value of the artificial limb action classifier, the real-time evaluation of the performance of the artificial limb action classifier is realized, the quantitative evaluation result is obtained, and the evaluation result is more accurate.
Drawings
FIG. 1 is a flowchart of a performance evaluation method of a prosthesis motion classifier according to an embodiment of the present invention;
FIG. 2 is a flowchart of a performance evaluation method of a prosthesis motion classifier according to a second embodiment of the present invention;
FIG. 3 is an exemplary diagram of a correspondence between preset result actions and a trajectory to be drawn in an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a performance evaluation device of a prosthesis motion classifier according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a terminal according to a fourth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some but not all of the relevant aspects of the present invention are shown in the drawings.
Example one
Fig. 1 is a flowchart of a performance evaluation method for a prosthesis motion classifier according to an embodiment of the present invention, where the embodiment is applicable to a situation of performing control performance evaluation on a prosthesis motion classifier in an electromyographic prosthesis, and the method may be executed by a performance evaluation device for the prosthesis motion classifier, where the device may be implemented by software and/or hardware, and may be generally integrated in a terminal such as a computer, and the method specifically includes the following steps:
When an amputee installs an electromyographic prosthetic for the first time, in order to enable the electromyographic prosthetic to adapt to the muscle characteristics of the amputee, action training needs to be carried out on a prosthetic action classifier in the electromyographic prosthetic. The application is the evaluation of the control performance of the trained artificial limb motion classifier.
When the control performance of the artificial limb action classifier is evaluated, a plurality of test items can be set in advance, the size and the position of a test target in each test item can be set in advance, the test items can also be randomly determined during specific test, a test subject is tested aiming at each test item to obtain the test result of the artificial limb action classifier which is trained, and finally the test results of each test item are integrated to obtain the evaluation performance of the artificial limb action classifier. Wherein the plurality of test items is at least two test items.
When the control performance of the trained artificial limb motion classifier is evaluated, one test item is selected from a plurality of preset test items to serve as a current test item. When the test items are specifically selected, the test items can be selected randomly or according to the sequence of the test items.
And 120, displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring the electromyographic signals of the testee, identifying the electromyographic signals through a prosthesis action classifier to obtain result actions, drawing action tracks corresponding to the result actions from the starting point, and recording test time.
The performance of the prosthesis motion classifier is evaluated based on Fitts' law. Among them, fitt's law is a model of human behavioral activity proposed by Paul fits in 1954, and the basic idea of the law is: the time required for any point to move to the center of the target is related to the distance and the size of the point to the target, the longer the distance is, the shorter the target is, and meanwhile, the Fitt's law gives a plurality of parameters for measuring the moving efficiency.
The coordinate axis may be a one-dimensional coordinate axis, a two-dimensional coordinate axis, or a three-dimensional coordinate axis. The number of the actions which can be identified by the trained artificial limb action classifier can be determined, and if two actions can be identified, the coordinate axis is determined to be a one-dimensional coordinate axis; when four actions can be identified, determining a coordinate axis as a two-dimensional coordinate axis; when six actions can be identified, the coordinate axis is determined to be a three-dimensional coordinate axis, which is only an example and is not limited specifically. The myoelectric artificial limb comprises an electrode, an artificial limb action classifier, a driving motor and a mechanical system; wherein, the electrode is positioned on the stump of the amputee and is used for collecting the electromyographic signal of the amputee; the artificial limb action classifier is used for identifying the electromyographic signals on the residual limb collected by the electrodes to obtain corresponding result actions; the driving motor drives the mechanical system to complete the resulting action. When the amputee wants to make an intention action, if the driving motor drives the mechanical system to finish the action, the identification of the artificial limb action classifier is successful; the prosthesis motion classifier identification fails if the drive motor drives the mechanical system to complete a false motion that is different from the intended motion. The myoelectric signals are the superposition of action potentials of movement units in a plurality of myofibers on time and space, and the surface myoelectric signals are the comprehensive effect of superficial muscle myoelectric signals and the electrical activity of a nerve trunk on the surface of skin and can reflect the activity of neuromuscular to a certain extent. The electromyographic signals in this application may be surface electromyographic signals.
The artificial limb action classifier is software, can run in the myoelectric artificial limb, and can also run in a computer and other terminals, so that the control performance of the myoelectric artificial limb action classifier is evaluated. When the prosthesis motion classifier is evaluated, the subject may be a healthy person or an amputee, and the prosthesis motion classifier must be a classifier trained on the subject.
And mapping the result action identified by the artificial limb action classifier into a corresponding track. When a current test item starts, firstly displaying a starting point and a test target position in the current test item on a coordinate axis, wherein the starting point and the test target position can be reached only when the test target position from the starting point comprises a plurality of tracks corresponding to actions, so a plurality of electromyographic signals of a testee are required to be continuously acquired, when one electromyographic signal of the testee acquired by an electrode is acquired, the electromyographic signal is input into a prosthesis action classifier to be identified to obtain a corresponding result action, a to-be-drawn track corresponding to the result action is determined, the to-be-drawn track is drawn from the starting point to obtain a drawn action track, then the next electromyographic signal is identified to obtain a corresponding result action, the corresponding to-be-drawn track is determined, the to-be-drawn track is drawn by taking an end point of the last drawn action track as the starting point, and the action tracks drawn twice before are obtained, drawing a to-be-drawn track corresponding to a result action obtained by identifying the subsequent myoelectric signal according to the drawing to obtain a final action track. In the test process of the current test item, the test time of the current test item can be recorded. The test target may be displayed as a circle, a square, or the like, and is not particularly limited herein. The subject's motion trajectory may follow a coordinate axis and have some degree of round-trip. The starting point may be any point on the coordinate axis, and for convenience of calculation, may be the origin of the coordinate. The size of the test target may be different among the plurality of test items.
And step 130, determining a test value of the current test item aiming at the test evaluation index according to the action track of the current test item and the test time.
The test evaluation index is an index for each individual test item, is set in advance, and sets a corresponding calculation rule in advance. The test evaluation index may be one or more. The test evaluation index may include at least one of a throughput rate, a path efficiency rate, a test result, an overshoot number, and a reaction time. The throughput rate is used for quantifying the completion condition of the test items, and is calculated according to the distance from the starting point to the test target, the width of the test target and the test time, and the unit is bit per second; in the current test item, the distance from the starting point to the central point of the test target is the shortest distance, and the shortest distance is divided by the length of the actual action track to be the effective rate of the path of the test item; the test result comprises test success or test failure; in the current test item, defining that the action track moves to the test target without staying for a second preset time and moves out of the test target as primary overshoot; in the current test item, the time required from the start of displaying the test target to the completion of drawing the first to-be-drawn track is the reaction time.
And after the action track of the current test item is obtained, calculating the value of the test evaluation index according to the action track and the test time of the current test item and the calculation rule of the test evaluation index to serve as a test value.
And 140, integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes.
The overall evaluation index is a comprehensive index of the artificial limb action classifier, and the overall evaluation index of the artificial limb action classifier comprises at least one of throughput rate, path effective rate, completion rate, overshoot rate and reaction time. And each overall evaluation index corresponds to an index in the test evaluation indexes one to one.
And according to a preset comprehensive rule of the test values of the plurality of test items for the test evaluation indexes, synthesizing the test values of the plurality of test items for the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier for the overall evaluation index. And if the test values of the plurality of test items aiming at the test evaluation indexes are averaged, the evaluation value of the prosthesis motion classifier aiming at the corresponding overall evaluation index is obtained.
For example, for the throughput rate, the path efficiency and the reaction time in the overall evaluation index, when test values of a plurality of test items are integrated, the corresponding evaluation indexes may be averaged, so as to obtain evaluation values of the prosthesis motion classifier for the three overall evaluation indexes. The completion rate is the result obtained by dividing the number of successful tests in the test result by the total number of the test items after the plurality of test items are all finished; the overshoot rate is a result obtained by dividing the total number of times of occurrence of overshoot by the total number of test items after the plurality of test items are all ended.
According to the technical scheme of the embodiment, the starting point and the test target are displayed in the current test item, the electromyographic signal of the testee is continuously acquired, the electromyographic signal is recognized by the artificial limb action classifier to obtain the result action, the action track corresponding to the result action is drawn, the test time is recorded, the test value of the current test item aiming at the test evaluation index is determined according to the action track and the test time, the test values of a plurality of test items are integrated to obtain the evaluation value of the artificial limb action classifier, the performance of the artificial limb action classifier is evaluated in real time, the quantitative evaluation result is obtained, the evaluation result is accurate, and compared with a traditional subjective judgment method, the method has the advantages of being direct, objective and accurate.
On the basis of the above embodiment, the method may further include:
displaying the test value of the current test item aiming at the test evaluation index in real time;
and when the evaluation value of the artificial limb action classifier for the overall evaluation index is obtained, displaying the evaluation value.
One display interface may simultaneously include a drawing interface of the track and a display interface of the evaluation index, and of course, the drawing interface of the track and the display interface of the evaluation index may also be displayed in different display interfaces respectively. In the test process of the current test item, the display interface can also comprise a virtual reality interface, the virtual display interface comprises a virtual human upper limb, when the amputee takes an intention action, the upper limb in the interface can take a corresponding action according to the result action obtained by the classifier identification, and the action can be the same as or different from the intention action, so that the correctness of the identification of the action can be intuitively reflected.
In the test process of the current test item, when the test value of the current test item aiming at the test evaluation index is obtained, the test value of the current test item aiming at the test evaluation index can be displayed in real time. And after the plurality of test items are all finished, when the evaluation value of the artificial limb action classifier for the overall evaluation index is obtained, displaying the evaluation value of the artificial limb action classifier for the overall evaluation index for the reference of the testee so as to facilitate the subsequent training to be more targeted.
Example two
Fig. 2 is a flowchart of a performance evaluation method of a prosthesis motion classifier according to a second embodiment of the present invention, which is optimized based on the second embodiment, specifically, a specific process of testing a current test item to obtain a motion trajectory is optimized, and the method specifically includes the following steps:
The prosthesis motion classifier is trained by a limited number of motions, so that in practical use, the prosthesis motion classifier fails to recognize, that is, the amputee's intended motion is mistakenly recognized as another motion.
For example, an amputee needs to train a prosthesis motion classifier therein to adapt to the characteristics of the muscles of the amputee before using the myoelectric prosthesis. The embodiment of the invention takes a two-dimensional coordinate axis as an example, and when the artificial limb action classifier is trained, four actions are trained, including: hand open, hand closed, wrist adduction, and wrist abduction. After the training of the preset times of the four actions is completed, the trained artificial limb action classifier is obtained, and the artificial limb action classifier can be used for identifying the electromyographic signals of the user to obtain corresponding result actions, namely the corresponding result actions can be provided for an amputee, and the control performance of the trained artificial limb action classifier can be evaluated through the embodiment of the invention.
When the current test item is started, the starting point and the test target are displayed, and timing is started at the same time.
In order to control the action track to reach the test target from the starting point, the subject needs to generate a plurality of intention actions so as to generate a plurality of electromyographic signals, namely, the subject needs to perform corresponding actions, so that each electromyographic signal needs to be identified separately, and the currently acquired electromyographic signal of the subject is used as the current electromyographic signal.
And 204, inputting the current electromyographic signals into a prosthesis action classifier to obtain corresponding current result actions.
And inputting the current electromyographic signals into the trained artificial limb action classifier, and obtaining corresponding current result actions through the identification of the artificial limb action classifier.
For example, fig. 3 is an exemplary diagram of a corresponding relationship between a preset result action and a track to be drawn in the embodiment of the present invention, and as shown in fig. 3, when the result action includes hand opening, hand closing, wrist adduction, or wrist abduction, the corresponding relationship between the preset result action and the track to be drawn may include: when the result is manually opened, the track to be drawn is a track drawn along the y-axis direction; when the result is that the hand is closed, the track to be drawn is a track drawn downwards along the y axis; when the result is that the wrist is abducted, the track to be drawn is a track drawn leftwards along the x axis; when the result is received as the wrist, the track to be drawn is the track drawn along the x axis. The circles of different sizes in fig. 3 represent test targets of different sizes at different positions, that is, the size and position of the test target may randomly appear according to a preset size and position in each test item.
And matching the current result action with a result action in the corresponding relation between the preset result action and the track to be drawn, and acquiring the track to be drawn corresponding to the result action as the current track to be drawn corresponding to the current result action when the matching is successful.
And step 206, controlling the cursor to move from the starting point or the end point of the historical drawing track finished according to the historical electromyographic signals to draw the current track to be drawn to form the current drawing track.
The historical electromyographic signals are electromyographic signals before the current electromyographic signals in the current test item, the historical drawing tracks are tracks formed by drawing actions according to results obtained by identifying the historical electromyographic signals by the prosthesis action classifier, and the drawing tracks corresponding to all the historical electromyographic signals are connected, so that the historical drawing tracks comprise the drawing tracks corresponding to all the historical electromyographic signals. When the current electromyographic signal is the first electromyographic signal acquired in the current test item, the current drawing track is the current track to be drawn; and when the current electromyographic signal is not the first electromyographic signal acquired in the current test item, the current drawing track comprises a historical drawing track and a current track to be drawn after drawing.
When the current electromyographic signal is the first electromyographic signal acquired in the current test item, controlling a cursor to draw the current track to be drawn from a starting point to form the current drawn track, and recording the current timing time, namely the time from the beginning of the current test item to the time for drawing the current drawn track according to the first electromyographic signal, wherein the time is also called reaction time; and when the current electromyographic signal is not the first electromyographic signal acquired in the current test item, controlling a cursor to draw a current track to be drawn from the end point of the historical drawn track which is drawn according to the historical electromyographic signal, so as to form the current drawn track. The reaction time not only refers to the reaction speed of the subject, but also includes the time from the time when the subject takes the intention action to the time when the myoelectric artificial limb recognizes the resultant action through the artificial limb action classifier, so the index can embody the recognition speed of the artificial limb action classifier.
The first preset time is the set time of the current test item, and the second preset time is the time when the current drawing track reaches the test target and needs to be reserved in the test target.
When the timing time reaches a first preset time, ending the current test item; or when the terminal point of the current drawing track reaches the test target and the time kept by the terminal point is greater than or equal to the second preset time, ending the current test item. And step 208 is executed after the current test item is finished, if the two conditions are not met, the step 203 is executed again to continue to acquire the electromyographic signals of the subject.
Optionally, when the current drawing trajectory of the current test item meets at least one of the following conditions, it is determined that the test of the current test item fails:
when the timing time reaches a first preset time, the end point of the current drawing track does not reach the test target;
the current drawn track moves to the test target without staying for a second preset time and moves out of the test target to be defined as one-time overshoot, and the overshoot times are greater than the preset times;
in the testing process, the distance of the terminal point of the current drawing track deviating from the testing target is larger than the preset distance or the feedback test of the testee fails.
In the test process, the distance that the terminal point of the current drawn track deviates from the test target is greater than the preset distance, and it can be determined that the track corresponding to the action of the result identified by the electromyographic signal generated by the subject can no longer reach the test target within the first preset time, so that it is determined that the test of the current test item fails, and this condition can also be referred to as uncontrolled jump. During the test process, if the test subject thinks that the test subject can not successfully control the cursor to move to the test target, the test failure can be actively fed back.
And 208, determining the current drawn track as the action track, and determining the current timing time as the test time.
And 209, determining a test value of the current test item aiming at the test evaluation index according to the action track of the current test item and the test time.
The test evaluation index may optionally include at least one of throughput rate, path efficiency, test result, overshoot number, and reaction time. The definitions of the test evaluation indexes are described in the above embodiments, and are not described herein.
When determining the test value of the current test item for the test evaluation index, the test value of some test evaluation indexes may be determined during the test of the current test item, such as the reaction time.
And step 210, integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes.
In the technical scheme of the embodiment, the timing is carried out when the current test item is started, the current electromyographic signals are identified by the artificial limb action classifier to obtain the corresponding current result action, determining the corresponding current track to be drawn, displaying the drawing process of the current track to be drawn in real time, when the timing time reaches a first preset time or the end point of the current drawn track reaches a test target and keeps being greater than or equal to a second preset time, determining the current drawn track as an action track, the current timing time as the test time, determining the test value of the current test item for the evaluation index according to the action track and the test time, and integrates the test values of a plurality of test items to obtain the evaluation value corresponding to the evaluation index, thereby realizing the real-time evaluation of the control performance of the artificial limb, and a quantitative evaluation result is obtained, so that the artificial limb action classifier can be evaluated safely, accurately and objectively.
EXAMPLE III
Fig. 4 is a schematic structural diagram of a performance evaluation device of a prosthesis motion classifier according to a third embodiment of the present invention, which is applicable to the case of performing control performance evaluation on a prosthesis motion classifier in an electromyographic prosthesis, and the device may be implemented by software and/or hardware, and may be generally integrated in a computer or other terminal. As shown in fig. 4, the performance evaluation device of the prosthesis motion classifier according to the present embodiment includes: a test item selection module 310, an action trajectory drawing module 320, a test value determination module 330, and an evaluation value determination module 340.
The test item selecting module 310 is configured to determine that one test item of a plurality of preset test items is a current test item;
the action track drawing module 320 is used for displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
the test value determining module 330 is configured to determine a test value of the current test item for the test evaluation indicator according to the action track of the current test item and the test time;
and the evaluation value determining module 340 is configured to synthesize the test values of the plurality of test items for the test evaluation index to obtain an evaluation value of the prosthesis motion classifier for the overall evaluation index.
Optionally, the action trajectory drawing module includes:
the starting point target display unit is used for displaying a starting point and a test target in the current test item on the coordinate axis, stopping the cursor at the starting point when starting the test and starting timing;
the electromyographic signal acquisition unit is used for acquiring a current electromyographic signal generated by the subject aiming at a test target of a current test item;
the result action determining unit is used for inputting the current myoelectric signal into the artificial limb action classifier to obtain a corresponding current result action;
the current track to be drawn determining unit is used for determining a corresponding current track to be drawn according to the current result action and the corresponding relation between the preset result action and the track to be drawn;
the track drawing unit is used for controlling a cursor to start moving from the starting point or the end point of a historical drawing track finished according to a previous historical electromyographic signal to draw the current track to be drawn to form the current drawing track, and repeatedly triggering the electromyographic signal acquisition unit until timing time reaches first preset time or the end point of the current drawing track reaches the test target and keeps more than or equal to second preset time;
and the action track determining unit is used for determining the current drawn track as the action track and determining the current timing time as the test time.
Optionally, when the current drawing trajectory of the current test item meets at least one of the following conditions, it is determined that the test of the current test item fails:
when the timing time reaches a first preset time, the end point of the current drawing track does not reach the test target;
the current drawn track moves to the test target without staying for a second preset time and moves out of the test target to be defined as one-time overshoot, and the overshoot times are greater than the preset times;
in the testing process, the distance of the terminal point of the current drawing track deviating from the testing target is larger than the preset distance or the feedback test of the testee fails.
Optionally, the test evaluation index includes at least one of throughput rate, path efficiency, test result, overshoot number, and reaction time.
Optionally, the method further includes:
the test value display module is used for displaying the test value of the current test item aiming at the test evaluation index in real time;
and the evaluation value display module is used for displaying the evaluation value of the artificial limb action classifier aiming at the overall evaluation index.
The performance evaluation device of the prosthesis motion classifier can execute the performance evaluation method of the prosthesis motion classifier provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For details of the technology not described in detail in this embodiment, reference may be made to the performance evaluation method of the prosthesis motion classifier provided in any embodiment of the present invention.
Example four
Fig. 5 is a schematic structural diagram of a terminal according to a fourth embodiment of the present invention, as shown in fig. 4, the terminal includes a processor 410, a memory 420, an input device 430, and an output device 440; the number of the processors 410 in the terminal may be one or more, and one processor 410 is taken as an example in fig. 4; the processor 410, the memory 420, the input device 430 and the output device 440 in the terminal may be connected by a bus or other means, which is exemplified in fig. 4.
The memory 420 serves as a computer-readable storage medium, and may be used for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the performance evaluation method of the prosthesis motion classifier according to the embodiment of the present invention (for example, the test item selecting module 310, the motion trajectory drawing module 320, the test value determining module 330, and the evaluation value determining module 340 in the performance evaluation apparatus of the prosthesis motion classifier). The processor 410 executes various functional applications and data processing of the terminal by executing software programs, instructions and modules stored in the memory 420, so as to implement the performance evaluation method of the prosthesis motion classifier described above.
The memory 420 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 420 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 420 may further include memory located remotely from processor 510, which may be connected to a terminal over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 430 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the terminal. The output device 440 may include a display device such as a display screen.
EXAMPLE five
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a method for performance evaluation of a prosthesis motion classifier, the method comprising:
determining one test item in a plurality of preset test items as a current test item;
displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
determining a test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item;
and integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes.
Of course, the embodiment of the present invention provides a storage medium containing computer-executable instructions, and the computer-executable instructions are not limited to the operations of the method described above, and can also perform related operations in the performance evaluation method of the prosthesis motion classifier provided in any embodiment of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the performance evaluation device of the prosthesis motion classifier, the included units and modules are only divided according to the functional logic, but not limited to the above division, as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (8)
1. A method for performance evaluation of a prosthesis motion classifier, the method comprising:
determining one test item in a plurality of preset test items as a current test item;
displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
determining a test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item;
synthesizing the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes;
displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthesis action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time, wherein the method comprises the following steps:
s1, displaying a starting point and a test target in the current test item on a coordinate axis, stopping a cursor at the starting point when starting the test, and starting timing;
s2, acquiring a current electromyographic signal generated by the subject aiming at a test target of the current test item;
s3, inputting the current myoelectric signal into a prosthesis action classifier to obtain a corresponding current result action;
s4, determining a corresponding current track to be drawn according to the current result action and the corresponding relation between the preset result action and the track to be drawn;
s5, moving a cursor from the starting point or the end point of the historical drawing track finished according to the previous historical electromyographic signal to draw the current track to be drawn to form a current drawing track, and repeatedly executing S2-S5 until the timing time reaches a first preset time or the end point of the current drawing track reaches the test target and keeps more than or equal to a second preset time;
and S6, determining the current drawing track as the action track, and determining the current timing time as the test time.
2. The method of claim 1, wherein the current test item is determined to have failed the test when the current drawing trace of the current test item satisfies at least one of the following conditions:
when the timing time reaches a first preset time, the end point of the current drawing track does not reach the test target;
the current drawn track moves to the test target without staying for a second preset time and moves out of the test target to be defined as one-time overshoot, and the overshoot times are greater than the preset times;
in the testing process, the distance of the terminal point of the current drawing track deviating from the testing target is larger than the preset distance or the feedback test of the testee fails.
3. The method of claim 1, wherein the test evaluation metrics include at least one of throughput rate, path efficiency, test results, overshoot times, and reaction time.
4. The method of claim 1, further comprising:
displaying the test value of the current test item aiming at the test evaluation index in real time;
and when the evaluation value of the artificial limb action classifier for the overall evaluation index is obtained, displaying the evaluation value.
5. A performance evaluation device for a prosthesis motion classifier, the device comprising:
the test item selection module is used for determining one test item in a plurality of preset test items as a current test item;
the action track drawing module is used for displaying a starting point and a test target in the current test item on a coordinate axis, continuously acquiring an electromyographic signal of a subject, identifying the electromyographic signal through a prosthetic limb action classifier to obtain a result action, drawing an action track corresponding to the result action from the starting point, and recording test time;
the test value determining module is used for determining the test value of the current test item aiming at the test evaluation index according to the action track and the test time of the current test item;
the evaluation value determining module is used for integrating the test values of the plurality of test items aiming at the test evaluation indexes to obtain the evaluation value of the artificial limb action classifier aiming at the overall evaluation indexes;
wherein, the action track drawing module comprises:
the starting point target display unit is used for displaying a starting point and a test target in the current test item on the coordinate axis, stopping the cursor at the starting point when starting the test and starting timing;
the electromyographic signal acquisition unit is used for acquiring a current electromyographic signal generated by the subject aiming at a test target of a current test item;
the result action determining unit is used for inputting the current myoelectric signal into the artificial limb action classifier to obtain a corresponding current result action;
the current track to be drawn determining unit is used for determining a corresponding current track to be drawn according to the current result action and the corresponding relation between the preset result action and the track to be drawn;
the track drawing unit is used for controlling a cursor to start moving from the starting point or the end point of a historical drawing track finished according to a previous historical electromyographic signal to draw the current track to be drawn to form the current drawing track, and repeatedly triggering the electromyographic signal acquisition unit until timing time reaches first preset time or the end point of the current drawing track reaches the test target and keeps more than or equal to second preset time;
and the action track determining unit is used for determining the current drawn track as the action track and determining the current timing time as the test time.
6. The apparatus of claim 5, further comprising:
the test value display module is used for displaying the test value of the current test item aiming at the test evaluation index in real time;
and the evaluation value display module is used for displaying the evaluation value of the artificial limb action classifier aiming at the overall evaluation index.
7. A terminal, characterized in that the terminal comprises:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method for performance evaluation of a prosthesis motion classifier as claimed in any one of claims 1 to 4.
8. A computer storage medium on which a computer program is stored, which program, when executed by a processor, carries out a method of performance evaluation of a prosthesis motion classifier as claimed in any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711336803.4A CN108008821B (en) | 2017-12-14 | 2017-12-14 | Performance evaluation method, device, terminal and storage medium of artificial limb action classifier |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711336803.4A CN108008821B (en) | 2017-12-14 | 2017-12-14 | Performance evaluation method, device, terminal and storage medium of artificial limb action classifier |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108008821A CN108008821A (en) | 2018-05-08 |
| CN108008821B true CN108008821B (en) | 2021-04-02 |
Family
ID=62058818
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711336803.4A Active CN108008821B (en) | 2017-12-14 | 2017-12-14 | Performance evaluation method, device, terminal and storage medium of artificial limb action classifier |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108008821B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101987048A (en) * | 2009-08-03 | 2011-03-23 | 深圳先进技术研究院 | Artificial limb control method and system thereof |
| CN106371590A (en) * | 2016-08-29 | 2017-02-01 | 华南理工大学 | High-performance motor imagery online brain-computer interface system based on OpenVIBE |
| CN106845419A (en) * | 2017-01-22 | 2017-06-13 | 深圳市唯特视科技有限公司 | A kind of method that Gait Recognition performance is assessed based on capturing movement framework |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ES2661538T3 (en) * | 2013-06-12 | 2018-04-02 | Otto Bock Healthcare Gmbh | Limb device control |
-
2017
- 2017-12-14 CN CN201711336803.4A patent/CN108008821B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101987048A (en) * | 2009-08-03 | 2011-03-23 | 深圳先进技术研究院 | Artificial limb control method and system thereof |
| CN106371590A (en) * | 2016-08-29 | 2017-02-01 | 华南理工大学 | High-performance motor imagery online brain-computer interface system based on OpenVIBE |
| CN106845419A (en) * | 2017-01-22 | 2017-06-13 | 深圳市唯特视科技有限公司 | A kind of method that Gait Recognition performance is assessed based on capturing movement framework |
Non-Patent Citations (1)
| Title |
|---|
| 基于微处理器的多功能肌电假肢控制系统;田岚等;《集成技术》;20130731;第2卷(第4期);第14-19页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108008821A (en) | 2018-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108983973B (en) | Control method of humanoid smart myoelectric artificial hand based on gesture recognition | |
| CN107378944B (en) | Multidimensional surface electromyographic signal artificial hand control method based on principal component analysis method | |
| CN110710970B (en) | Method and device for recognizing limb actions, computer equipment and storage medium | |
| JP2020034849A (en) | Work support device, work support method, and work support program | |
| WO2008097201A1 (en) | A system and method for processing brain signals in a bci system | |
| CN114707562B (en) | Electromyographic signal sampling frequency control method and device and storage medium | |
| CN103955270A (en) | Character high-speed input method of brain-computer interface system based on P300 | |
| CN109976525A (en) | A kind of user interface interaction method, apparatus and computer equipment | |
| Sree et al. | EMG controlled bionic robotic arm using artificial intelligence and machine learning | |
| CN113974589B (en) | Multi-modal behavior paradigm evaluation optimization system and cognitive ability evaluation method | |
| Chavarriaga et al. | Adaptation of hybrid human-computer interaction systems using EEG error-related potentials | |
| Chang et al. | A hierarchical hand motions recognition method based on IMU and sEMG sensors | |
| Zou et al. | A multimodal fusion model for estimating human hand force: Comparing surface electromyography and ultrasound signals | |
| CN108008821B (en) | Performance evaluation method, device, terminal and storage medium of artificial limb action classifier | |
| Hasan et al. | Human hand gesture detection based on EMG signal using ANN | |
| Nieuwoudt et al. | Investigation of real-time control of finger movements utilising surface EMG signals | |
| Lew et al. | Biofeedback Upper Limb Assessment Using Electroencephalogram, Electromyographic and Electrocardiographic with Machine Learning in Signal Classification. | |
| CN115804590A (en) | Wakefulness detection system and method | |
| CN113657243B (en) | Test method for performance influence of non-contact bionic remote control gesture | |
| Zohirov | Classification Of Some Sensitive Motion Of Fingers To Create Modern Biointerface | |
| CN111967333B (en) | Signal generation method, system, storage medium and brain-computer interface spelling device | |
| Castellini et al. | EMG-based prediction of multi-DOF activations using single-DOF training: a preliminary result | |
| CN111625098B (en) | Intelligent virtual avatar interaction method and device based on multi-channel information fusion | |
| CN114327046B (en) | Method, device and system for multi-mode human-computer interaction and intelligent state early warning | |
| KR102521180B1 (en) | Similar electrocardiography based personal authentication system, personal authentication method and computer program stored in recording medium to execute the personal authentication method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |