Evaluation training method and system based on intelligent frosted table
By dividing training data into tiny time windows on the intelligent frosted table and automatically adjusting the motor output power, the problem of inaccurate physician assessments is solved, achieving precise assessment and improved safety in rehabilitation training.
Patent Information
- Application Number
- CN202511561318.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-29
- Publication Date
- 2026-02-06
AI Technical Summary
Physicians' assessment of patients' training effectiveness may be influenced by subjective experience and fatigue levels, leading to inconsistent assessment standards, difficulty in capturing instantaneous characteristics during the training process in real time, and lag in adjusting training parameters, which affects the accuracy and safety of training effectiveness.
By dividing the training data into tiny time windows on the intelligent sanding table, the speed, acceleration, and position changes are analyzed, training evaluation coefficients and comprehensive evaluation indices are calculated, and the motor output power is automatically adjusted to match the patient's current exercise ability, thus establishing an objective and quantitative evaluation system.
It enables precise assessment and personalized adjustment of rehabilitation training, improves the safety and efficiency of training, reduces subjective bias, ensures that the training intensity matches the patient's ability, and reduces the risk of injury.
Smart Images
Figure CN121483592A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of rehabilitation training. More particularly, the present application relates to a smart sanding table-based evaluation training method and system. BACKGROUND
[0002] The sanding table is a combination of modern intelligence and traditional occupational training. It uses intelligent LED light for visual trajectory tracking, voice control guidance, and other human-computer interaction models. It combines intelligent sanding tools to perform precise upper limb function training under the guidance of an intelligent system. The training process and results are quantified. The sanding table guides users through a variety of games, colors, and graphical changes to improve the fun and compliance of training, increase proprioceptive stimulation, achieve the goal of training cognitive ability and gross motor ability of the upper limbs, and improve training efficiency. It is suitable for upper limb activity training, cognitive ability training, guided training, and control force training. In the rehabilitation training through the visual trajectory guidance of the intelligent LED light point array, straight line movement is a common and basic training mode. Straight line movement strips the complexity of movement, has lower requirements for the physical fitness of patients, and is convenient for evaluating the training quality of patients to adjust the training intensity of patients in real time.
[0003] In the prior art, a physician can make a comprehensive judgment on the training effect of a patient by combining the overall state of the patient, facial expressions, subjective feelings, and other factors, has good clinical adaptability, and can communicate with the patient in real time during the evaluation process to provide psychological support and guidance. Then the physician corrects the training intensity of the patient through experience.
[0004] However, the evaluation results of the physician may be affected by subjective experience, fatigue level, and other factors, and the evaluation standards between different physicians may differ. At the same time, real-time data changes during the training process are often fleeting, and the physician is difficult to capture and analyze these transient characteristics synchronously during the training process, resulting in that the evaluation of the physician on the training effect of the patient may not conform to the actual physiological state of the patient, and the relative separation of training and evaluation also makes the adjustment of training parameters often have a certain lag. SUMMARY
[0005] To solve the above technical problem that the evaluation of the physician on the training effect of the patient may not conform to the actual physiological state of the patient, the present application provides a solution in the following aspects.
[0006] In a first aspect, a smart sanding table-based evaluation training method includes: obtaining training data at multiple time points when a patient holds a handle of a sanding table for a training, obtaining an output power of a motor and recording it as a historical output power, wherein the training data at the first time point includes the speed and acceleration of the handle at the first time point, the speed and acceleration of the handle at the second time point, and the output power of the motor at the first time point. i i In a first aspect, a smart sanding table-based evaluation training method includes: obtaining training data at multiple time points when a patient holds a handle of a sanding table for a training, obtaining an output power of a motor and recording it as a historical output power, wherein the training data at the first time point includes the speed and acceleration of the handle at the first time point, the speed and acceleration of the handle at the second time point, and the output power of the motor at the first time point.i The relative distance between the current position and the preset initial position is used to apply resistance to the handle; the training data from multiple times are divided into multiple time windows of equal size according to the time sequence; the training evaluation coefficient of each time window is calculated, where the nth time window is the training evaluation coefficient of ... j The training evaluation coefficient of the first time window and the first time window j The system is associated with all training data within a time window; a comprehensive evaluation index is calculated, wherein the comprehensive performance index is associated with the training evaluation coefficient of each time window; the historical output power is corrected according to the comprehensive evaluation index to obtain the expected output power, and the motor output power is determined to be the expected output power when the patient holds the handle of the sanding table for training next time.
[0007] Preferably, calculate the first j The formula for the training evaluation coefficient for each time window is: .
[0008] in, M j For the first j Training evaluation coefficients for each time window. E j For the first j The mean of all relative distances within a time window. R The standard displacement scale is preset in size. σ v,j For the first j The standard deviation of all velocities within a time window μ v,j For the first j The mean of all velocities within a time window. σ a,j For the first j The standard deviation of all accelerations within a time window μ a,j For the first j The mean of all accelerations within a time window. ω 1 represents the first weight of the preset size. ω 2 represents the second weight of the preset size. ω 3 is the third weight of the preset size, and , norm () is the standard normalization function, and exp() is the exponential function with the natural logarithm e as the base.
[0009] Preferably, the first weight ω The value of 1 is 0.1, the second weight ω The value of 2 is 0.45, the third weight ω The value of 3 is 0.45.
[0010] Preferably, the calculation of the comprehensive evaluation index comprises: obtaining the first quartile and the third quartile of all training evaluation indexes; calculating a low stability index according to the training evaluation coefficients less than the first quartile; calculating a high stability index according to the training evaluation coefficients greater than the third quartile; and determining the geometric mean of the low stability index and the high stability index as the comprehensive evaluation index.
[0011] Preferably, the formula for calculating the low stability index is: .
[0012] wherein, I 1 is the low stability index, μ S is the mean of all training evaluation coefficients less than the first quartile, M s,max is the maximum of all training evaluation coefficients less than the first quartile, M s,min is the minimum of all training evaluation coefficients less than the first quartile, norm () is a standard normalization.
[0013] Preferably, the formula for calculating the high stability index is: .
[0014] wherein, I 2 is the high stability index, μ L is the mean of all training evaluation coefficients greater than the third quartile, M L,max is the maximum of all training evaluation coefficients greater than the third quartile, M L,min is the minimum of all training evaluation coefficients greater than the third quartile.
[0015] Preferably, the formula for correcting the historical output power according to the comprehensive evaluation index is: .
[0016] wherein is the expected output power, P is the historical output power, Z is the comprehensive evaluation index, Z r is a reference evaluation index of a preset size, α is a correction parameter of a preset size.
[0017] Preferably, the correction parameterα The value is 0.3.
[0018] Preferably, an evaluation and training method based on a smart frosted table further includes: issuing an alarm in response to the comprehensive evaluation index being less than a preset threshold.
[0019] In a second aspect, an assessment and training system based on an intelligent frosted table includes a processor and a memory, the memory storing a computer program, the processor executing the computer program to implement an assessment and training method based on an intelligent frosted table as described in any of the above-described inventions.
[0020] The beneficial effects of this invention are as follows: This invention establishes an objective and quantitative evaluation system by dividing the training process into small time windows and conducting multi-dimensional motion characteristic analysis. It enables personalized adaptive adjustment of training resistance, accurately identifies control defects in the movement phase, thereby improving the accuracy of rehabilitation training evaluation and significantly enhancing the safety and efficiency of upper limb functional rehabilitation training. Attached Figure Description
[0021] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a schematic flowchart illustrating the steps of an evaluation and training method based on an intelligent frosted table according to an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the structure of an evaluation and training system based on an intelligent frosted table according to this embodiment. Detailed Implementation
[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0024] Figure 1 This is a schematic flowchart illustrating the steps of an evaluation and training method based on an intelligent frosted table according to an embodiment of the present invention.
[0025] like Figure 1As shown, an evaluation and training method based on a smart frosted table includes steps S1 to S2.
[0026] Step S1: During a training session where the patient holds the handle of the sanding table, training data is obtained at multiple times to obtain the motor's output power and record it as the historical output power.
[0027] Among them, the i The training data at time step n includes: the handle at time step n. i The velocity and acceleration at a given moment, and the handle at the 1st moment. i The relative distance between the current position and the preset initial position is used by the motor to apply resistance to the handle.
[0028] It should be noted that the greater the motor's output power, the greater the resistance it exerts on the handle. Output power is a direct reflection of the motor's ability to perform work. When its output power increases, the motor can generate and maintain greater torque. This torque is converted into mechanical resistance acting on the handle through the transmission mechanism, and the patient will feel that they need to exert more force to push the handle.
[0029] Step S2: Divide the training data at multiple times into multiple time windows of the same size according to the time sequence.
[0030] It should be noted that the movements of patients gripping the handle of a sandpaper table during training are typically disordered and variable. Therefore, this invention improves the accuracy of assessing the stability of patient training movements by setting fixed-duration time windows to determine the stability of the patient's movements at a microscopic scale. Each time window serves as an independent observation sample, and the dispersion (coefficient of variation) of its internal data (i.e., the speed, acceleration, and distance from the initial position of the handle) directly reflects the patient's ability to maintain stable movement within that time period. The smaller the data fluctuation within a window, the more stable the movement control during that period. In one embodiment, the size of the time window is 0.1 seconds.
[0031] Step S3: Calculate the training evaluation coefficients for each time window.
[0032] Among them, the j The training evaluation coefficient of the first time window and the first time window j It relates to all training data within a time window.
[0033] In one embodiment, the calculation of the first j The formula for the training evaluation coefficient for each time window is: .
[0034] in, M j For the first j Training evaluation coefficients for each time window.E j For the first j The mean of all relative distances within a time window. R The standard displacement scale is preset in size. σ v,j For the first j The standard deviation of all velocities within a time window μ v,j For the first j The mean of all velocities within a time window. σ a,j For the first j The standard deviation of all accelerations within a time window μ a,j For the first j The mean of all accelerations within a time window. ω 1 represents the first weight of the preset size. ω 2 represents the second weight of the preset size. ω 3 is the third weight of the preset size, and , norm () is the standard normalization function, and exp() is the exponential function with the natural logarithm e as the base. The first weight ω The value of 1 is 0.1, the second weight ω The value of 2 is 0.45, the third weight ω The value of 3 is 0.45.
[0035] It should be noted that patients perform linear push-pull training by holding the handles.
[0036] E j For the first j The mean of all relative distances within a time window reflects the effective displacement scale completed by the patient during that time period. The larger the mean of the relative distance, the greater the patient's range of motion and the stronger the patient's current motor ability.
[0037] formula Indicates the first j The coefficient of variation for all velocities within a given time window is used to assess velocity stability during hand movements in patients. When patients exhibit velocity control dysregulation, the formula... When the value is large, the formula... The value is relatively small. Based on this, the patient's first... j The more stable the speed corresponding to the first time window, the better. j Training evaluation coefficients for each time window M j The larger.
[0038] formula Indicates the first jThe coefficient of variation of all accelerations in a time window, which reflects the fluency and smoothness of the movement. Abnormal fluctuations in acceleration usually indicate that there is a lack of precision in muscle control or there is a compensatory movement pattern. When the patient's movement is relatively smooth, the value of is large, and the value of is small. Based on this, the smoother the patient's movement in the time period corresponding to the th time window, the larger the training evaluation coefficient of the th time window. j j M j
[0039] Step S4: Calculate the comprehensive evaluation index.
[0040] The comprehensive performance index is related to the training evaluation coefficients of each time window.
[0041] In one embodiment, calculating the comprehensive evaluation index comprises: obtaining the first quartile and the third quartile of all training evaluation indexes; calculating a low stability index according to the training evaluation coefficients less than the first quartile; calculating a high stability index according to the training evaluation coefficients greater than the third quartile; and determining the geometric mean of the low stability index and the high stability index as the comprehensive evaluation index.
[0042] In one embodiment, the formula for calculating the low stability index is: .
[0043] Wherein, I 1 is the low stability index, μ S is the mean of all training evaluation coefficients less than the first quartile, M s,max is the maximum value of all training evaluation coefficients less than the first quartile, M s,min is the minimum value of all training evaluation coefficients less than the first quartile, norm () is the standard normalization.
[0044] In one embodiment, the formula for calculating the high stability index is: .
[0045] Wherein, I 2 is the high stability index, μ L is the mean of all training evaluation coefficients greater than the third quartile, M L,max is the maximum value of all training evaluation coefficients greater than the third quartile, M L,min is the minimum value among all the training evaluation coefficients greater than the third quartile.
[0046] It should be noted that a set of data is arranged from small to large and divided into four equal parts, and the values at the three division points are called quartiles. There are three quartiles, which are the lower quartile (i.e., the first quartile), the median (i.e., the second quartile), and the upper quartile (i.e., the third quartile) from small to large. The lower quartile is the quartile located at the 25% position of the ordered data, that is, the 25th percentile of the data. The median is the quartile located at the middle position of the ordered data, that is, the 50th percentile of the data. The upper quartile is the quartile located at the 75% position of the ordered data, that is, the 75th percentile of the data.
[0047] It should be noted that the training evaluation coefficients less than the first quartile correspond to training data in the time window with large fluctuations. These smaller training evaluation coefficients correspond to periods of unstable motor control in patients, which may include loss of control at the start of movement, tremor at the time of direction change, or decreased coordination due to fatigue. The training data corresponding to these training evaluation coefficients can identify motor control defects.
[0048] The training evaluation coefficients greater than the third quartile correspond to training data in the time window with small fluctuations. These larger coefficient values correspond to periods of stable motor control in patients, reflecting the best state of motor control ability. In this stage, larger coefficient values indicate better motor stability. The training data corresponding to these training evaluation coefficients can objectively evaluate the motor function potential and recovery ceiling of patients.
[0049] The present application determines the geometric mean of the low stability index and the high stability index as the comprehensive evaluation index, so that the comprehensive evaluation index can more comprehensively reflect the rehabilitation progress of the patient. When the patient makes progress in the stability of motor control, the high stability index will correspondingly increase; and when the patient's motor fluctuation is improved, the low stability index will also increase. The geometric mean can capture the improvement of both dimensions at the same time, providing a more accurate quantitative basis for the evaluation of rehabilitation effect.
[0050] Step S5: correcting the historical output power according to the comprehensive evaluation index to obtain an expected output power, and determining the output power of the motor to be the expected output power when the patient next trains the handle of the sanding table.
[0051] In one embodiment, the formula for correcting the historical output power according to the comprehensive evaluation index is: .
[0052] wherein is the expected output power,P The historical output power, Z The comprehensive evaluation index is... Z r This is a reference evaluation index for a preset size. α This is a correction parameter for the preset size.
[0053] It should be noted that the historical output power was corrected by a comprehensive evaluation index, taking into account the safety and gradual nature of the rehabilitation training process.
[0054] This invention takes into account both training continuity and personalized adjustment, respecting the patient's training history while making targeted optimizations based on the latest performance.
[0055] Adjustment factor of the formula Greater than 0 and less than 2. When the patient performs poorly (i.e. Z < Z r When the factor is less than 1, the system will appropriately reduce the output power to avoid excessive training difficulty; when the patient performs well ( Z > Z r When this factor is greater than 1, the output power is increased to provide greater training intensity for the patient, resulting in better training outcomes. This invention ensures that the training intensity is always matched to the patient's current ability level.
[0056] It should be noted that the parameters are corrected. α The larger the value, the more sensitive the response to current training performance, and the greater the correction magnitude for historical output power. In one embodiment, the correction parameter... α The value is 0.3.
[0057] In summary, this invention, by dividing continuous training data into tiny, fixed-duration windows (e.g., 0.1 seconds), enables the analysis of motion stability at a microscale. It can capture transient motion control defects (such as slight tremors during startup and instability during turning) that are difficult to detect using traditional methods, greatly improving the precision and sensitivity of the evaluation.
[0058] This invention comprehensively considers the effective displacement scale (i.e., by designing multi-dimensional training evaluation coefficients) ), speed stability (i.e. ) and motion smoothness (i.e. This transforms traditional assessments that rely on physicians' subjective experience into objective, quantitative analysis based on data. This effectively eliminates subjective bias in human assessments, making the results more consistent with patients' actual performance during training.
[0059] This invention establishes a dynamic correction relationship between a comprehensive evaluation index and motor output power, thereby automatically adjusting the resistance load for the next training session based on the patient's real-time performance in each session, ensuring that the training intensity always matches the patient's current exercise capacity. Through a closed-loop training model of "assessment-feedback-adjustment," it ensures that the patient always trains within the optimal challenge range, avoiding both poor training results due to excessively low intensity and the risk of injury or frustration caused by excessively high intensity, thus significantly improving the efficiency and safety of rehabilitation training.
[0060] In one embodiment, the present invention further includes: issuing an alarm in response to the comprehensive evaluation index being less than a preset threshold.
[0061] It should be noted that when the comprehensive assessment index remains below the threshold, it may indicate several situations: the patient may have experienced significant fatigue accumulation, leading to a significant decline in motor control; the training intensity may exceed the patient's current tolerance, posing a risk of sports injury; or the patient may be experiencing inattention, pain response, or other abnormal physiological states. These situations require timely intervention and assessment by a therapist, thus necessitating an alert.
[0062] Figure 2 This is a schematic diagram illustrating the structure of an evaluation and training system based on an intelligent frosted table according to this embodiment.
[0063] This invention also provides an evaluation and training system based on an intelligent frosted table. For example... Figure 2 As shown, the system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement an evaluation and training method based on an intelligent frosted table according to the first aspect of the present invention.
[0064] The system also includes other components well known to those skilled in the art, such as communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.
[0065] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.
[0066] In the description of this specification, "multiple" or "several" means at least two, such as two, three or more, unless otherwise explicitly specified.
[0067] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.
Claims
1. An evaluation and training method based on an intelligent frosted table, characterized in that, include: During a training session where the patient holds the handle of the sanding table, training data is obtained at multiple points in time. The output power of the motor is then recorded as the historical output power. i The training data at time step n includes: the handle at time step n. i The velocity and acceleration at a given moment, and the handle at the 1st moment. i The relative distance between the current position and the preset initial position is used by the motor to apply resistance to the handle; The training data at multiple times is divided into multiple time windows of the same size according to the time sequence; Calculate the training evaluation coefficients for each time window, where the th time window is... j The training evaluation coefficient of the first time window and the first time window j It relates to all training data within a time window; Calculate the comprehensive evaluation index, wherein the comprehensive performance index is related to the training evaluation coefficient for each time window; The historical output power is corrected based on the comprehensive evaluation index to obtain the desired output power, and the motor output power is determined to be the desired output power when the patient holds the handle of the sanding table for training next time.
2. The evaluation and training method based on an intelligent frosted table according to claim 1, characterized in that, Calculate the first j The formula for the training evaluation coefficient for each time window is: ； in, M j For the first j Training evaluation coefficients for each time window. E j For the first j The mean of all relative distances within a time window. R The standard displacement scale is preset in size. σ v,j For the first j The standard deviation of all velocities within a time window μ v,j For the first j The mean of all velocities within a time window. σ a,j For the first j The standard deviation of all accelerations within a time window μ a,j For the first j The mean of all accelerations within a time window. ω 1 represents the first weight of the preset size. ω 2 represents the second weight of the preset size. ω 3 is the third weight of the preset size, and , norm () is the standard normalization function, and exp() is the exponential function with the natural logarithm e as the base.
3. The evaluation and training method based on an intelligent frosted table according to claim 2, characterized in that, The first weight ω The value of 1 is 0.1, the second weight ω The value of 2 is 0.45, the third weight ω The value of 3 is 0.
45.
4. The evaluation and training method based on an intelligent frosted table according to claim 2, characterized in that, The calculation of the comprehensive evaluation index includes: Obtain the first and third quartiles of all training evaluation metrics; The low stability index is calculated based on the training evaluation coefficients that are less than the first quartile. The high stability index is calculated based on the training evaluation coefficients that are greater than the third quartile. The geometric mean of the low stability index and the high stability index is determined as the comprehensive evaluation index.
5. The evaluation and training method based on an intelligent frosted table according to claim 4, characterized in that, The formula for calculating the low stability index is as follows: ,in, I 1 represents the low stability index. μ S The mean of all training evaluation coefficients that are less than the first quartile. M s,max It is the maximum value among all training evaluation coefficients that are less than the first quartile. M s,min It is the minimum of all training evaluation coefficients that are less than the first quartile. norm () represents standard normalization.
6. The evaluation and training method based on an intelligent frosted table according to claim 5, characterized in that, The formula for calculating the high stability index is as follows: ,in, I 2 represents the high stability index. μ L The mean of all training evaluation coefficients that are greater than the third quartile. M L,max It is the maximum value among all training evaluation coefficients that are greater than the third quartile. M L,min It is the minimum of all training evaluation coefficients that are greater than the third quartile.
7. The evaluation and training method based on an intelligent frosted table according to claim 1, characterized in that, The formula for correcting the historical output power based on the comprehensive evaluation index is as follows: ； in For the desired output power, P The historical output power, Z The comprehensive evaluation index is... Z r This is a reference evaluation index for a preset size. α This is a correction parameter for the preset size.
8. The evaluation and training method based on an intelligent frosted table according to claim 7, characterized in that, The correction parameter α The value is 0.
3.
9. The evaluation and training method based on an intelligent frosted table according to claim 1, characterized in that, Also includes: An alarm is triggered when the comprehensive evaluation index is less than a preset threshold.
10. An evaluation and training system based on an intelligent frosted table, comprising a processor and a memory, characterized in that, The memory stores a computer program, and the processor executes the computer program to implement an evaluation and training method based on a smart frosted table as described in any one of claims 1-9.