Gaze stability training evaluation method and system

CN119069069APending Publication Date: 2024-12-03SHANGHAI ZEHNIT MEDICAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411105525.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-12-03

Smart Images

  • Figure CN119069069A_ABST
    Figure CN119069069A_ABST
Patent Text Reader

Abstract

The invention discloses a staring stability training evaluation method and system. The method comprises the following steps: receiving a training instruction of a subject; displaying a corresponding visual target on a screen based on the training instruction; if the visual target is an active visual target, determining a display position of the visual target according to a current training mode and training difficulty, and collecting motion data of the subject in a training process; if the visual target is a passive visual target, motion data of the subject in the training process are collected, and the display position of the visual target is determined according to the current training mode and the motion data of the subject; and evaluating the current staring stability training according to the motion data and the display position data of the visual target. According to the invention, the patient can carry out training suitable for himself / herself, the training is gradually progressive from easy to difficult, and the challenging of the training is increased. The whole system realizes a closed-loop process of staring stability training, so that a patient can insist in rehabilitation training for a long time, and the problem that rehabilitation training is difficult to insist is greatly solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vestibular training technology, and in particular to a gaze stability training evaluation method and system. Background Art

[0002] In the field of vestibular rehabilitation, gaze stability training is an important training method that aims to help patients improve oscillopsia caused by vestibular dysfunction, improve their ability to see clearly during head movement, and enable them to return to normal social activities. The main purpose of this training method is to keep the eyes on the target during head movement.

[0003] Generally, gaze stability training includes two classic paradigms. In paradigm one, the visual target is stationary, and the patient moves the head back and forth horizontally or vertically while maintaining visual fixation on the target. In paradigm two, the visual target moves in the opposite direction of the head horizontally or vertically, and the patient should always focus on the target. In addition to the above two paradigms, there are many different training methods.

[0004] At present, the gaze stability training commonly used in clinical practice mostly uses objects that are readily available in real life as visual targets, such as business cards, playing cards, pens and other items. The patient or the rehabilitation therapist holds the objects in a fixed or reverse position for rehabilitation training. The presentation and movement of the visual targets are variable and cannot be displayed stably, which affects the training effect. At the same time, since most rehabilitation training is done at home, the existing technology cannot monitor whether the patient is training at home and the training process. The visual target cannot be displayed stably during training, so it is impossible to study the training effect of the patient's gaze stability training at home, and it is impossible to correct the patient's wrong training movements in time, making it impossible to effectively carry out gaze stability training in a large number of patients. Summary of the invention

[0005] The purpose of the present invention is to provide a gaze stability training evaluation method and system, which calculates the position of the visual target in an active or passive mode, uses head movement data and / or eye movement data for calculation, and realizes a scientific quantitative evaluation of gaze stability training.

[0006] In order to achieve the above object, the present invention provides a gaze stability training evaluation method, comprising the following steps:

[0007] receiving a training instruction from a subject, wherein the training instruction includes a currently selected training mode and a training difficulty;

[0008] Based on the training instruction, a corresponding visual target is displayed on the screen; the visual target is an active visual target or a passive visual target; wherein the active visual target is displayed according to a preset display parameter, and the passive visual target is displayed according to the motion data of the subject;

[0009] If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; and the motion data of the subject during the training process is collected; the motion data includes head movement data and / or eye movement data;

[0010] If the visual target is a passive visual target, the motion data of the subject during the training process is collected, and the display position of the visual target is determined according to the current training mode and the motion data of the subject;

[0011] The current gaze stability training is evaluated according to the motion data and the display position data of the visual target.

[0012] Further preferably, before the gaze stability training begins, the method further includes: evaluating the training difficulty of the subject; specifically, the following steps are included:

[0013] Displaying a visual target on a screen, and prompting the subject to quickly turn his head while being able to see the visual target clearly;

[0014] collecting actual head movement data of the subject;

[0015] Calculating the head movement frequency and the head movement deflection angle of the subject according to the actual head movement data of the subject;

[0016] The training difficulty matched to the subject is evaluated according to the head movement frequency and the head movement deflection angle of the subject.

[0017] Furthermore, the step of evaluating the training difficulty of the subject further includes the following steps:

[0018] While collecting the actual head movement data of the subject, collecting the actual eye movement data of the subject through an eye tracking device;

[0019] Synchronously recording the position data of the visual target on the screen;

[0020] Calculating the gaze point position of the subject according to the actual head movement data and the actual eye movement data of the subject;

[0021] Calculating the gaze deviation between the gaze point position of the subject and the position data of the visual target at each time point;

[0022] Counting the percentage of fixation points whose fixation deviations meet the preset deviation range;

[0023] If the proportion reaches a preset ratio, the training difficulty of the subject is determined based on the actual head movement data of the subject.

[0024] Further preferably, different display parameters are set for different training difficulty levels, and the display parameters include movement frequency and deflection amplitude;

[0025] If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; specifically including:

[0026] Determining a display track of the visual target according to a current training mode;

[0027] Acquire corresponding display parameters according to the current training difficulty, wherein the display parameters include target movement frequency and target deflection amplitude;

[0028] According to the target movement frequency and the target deflection amplitude, combined with the current training mode, the display of the visual target on the screen is controlled.

[0029] Further preferably, the current gaze stability training is evaluated according to the motion data and the display position data of the visual target; specifically comprising:

[0030] Determine ideal motion data based on the display position data of the visual target and the current training mode;

[0031] Time-aligning the collected actual motion data with the ideal motion data;

[0032] Calculating the deviation between the actual motion data and the ideal motion data;

[0033] Evaluate the current gaze stability training based on the calculated deviation results.

[0034] Furthermore, the calculation of the deviation between the actual motion data and the ideal motion data specifically includes:

[0035] Performing curve fitting or tracing on the ideal motion data to obtain an ideal motion curve;

[0036] Performing curve fitting or tracing on the actual motion data to obtain an actual motion curve;

[0037] The difference between the actual motion curve and the ideal motion curve is calculated.

[0038] Furthermore, the difference between the actual motion curve and the ideal motion curve is calculated by any one of the following calculation methods:

[0039] Area difference calculation method: calculating the area between the actual motion curve and the ideal motion curve, and obtaining the difference between the actual motion curve and the ideal motion curve according to the area; or

[0040] Motion spectrum difference calculation method: performing Fourier transform on the actual motion curve and the ideal motion curve respectively to obtain the actual motion spectrum and the ideal motion spectrum;

[0041] Extracting main spectra of the actual motion spectrum and the ideal motion spectrum;

[0042] Calculate the difference between the actual motion dominant spectrum and the ideal motion dominant spectrum; or

[0043] Motion shape distance difference calculation method:

[0044] Extracting feature points at equal intervals from the actual motion curve and the ideal motion curve;

[0045] For each feature point, calculate its relative position to other points to form a shape context descriptor;

[0046] Calculating the matching cost between the shape context descriptors of the actual motion curve and the ideal motion curve;

[0047] The matching costs of all feature points are accumulated to obtain a total shape distance; the shape distance is used to represent the difference between the actual motion curve and the ideal motion curve.

[0048] Further preferably, it also includes:

[0049] After the gaze stabilization training, adjusting the training difficulty of the subject according to the evaluation results of multiple gaze stabilization trainings; specifically including:

[0050] Recording the assessment score of the subject in each gaze stability training at the current training difficulty;

[0051] When the assessment score of the subject reaches a first set score value, increasing a first cumulative count;

[0052] When the subject's assessment score is lower than a second set score value, increasing a second cumulative count; wherein the first set score value is greater than or equal to the second set score value;

[0053] If the first cumulative count is greater than a preset first cumulative number threshold, or the second cumulative count is greater than a preset second cumulative number threshold, it is determined that the current training difficulty does not match the subject; wherein the first cumulative number is the same as or different from the second cumulative number;

[0054] If the current training difficulty does not match the subject, a training difficulty matching the subject is recommended based on the subject's motion data.

[0055] Furthermore, when the evaluation score of the subject reaches a first set score value, in addition to increasing the first cumulative count, a first continuous count is also increased, and if the set score value is not reached, the first continuous count is reset to zero;

[0056] If the first cumulative count reaches a preset first cumulative number of times, and the first continuous count reaches a preset continuous number of times, the training difficulty is increased.

[0057] Furthermore, the method of recommending a training difficulty matching the subject based on the subject's exercise data specifically includes:

[0058] Calculating visual target parameters according to the motion data of the subject collected during the training process, and determining the matching training difficulty according to the visual target parameters; wherein:

[0059] If the motion data only includes head movement data, determining the head movement frequency and the head movement deflection amplitude according to the head movement data of the subject; determining the visual target parameter according to the head movement frequency and the head movement deflection amplitude to recommend a matching training difficulty;

[0060] If the motion data only includes eye movement data, the deviation between the gaze point and the visual target is calculated according to the eye movement data of the subject; if the deviation between the gaze point and the visual target is not within a set error range, it is recommended to reduce the training difficulty;

[0061] If the motion data includes the head movement data and the eye movement data, the deviation between the gaze point and the visual target is calculated based on the eye movement data; if the deviation between the gaze point and the visual target is within a set error range, the head movement frequency and the deflection amplitude are obtained based on the head movement data, and the visual target parameters are determined based on the head movement frequency and the deflection amplitude to recommend a matching training difficulty.

[0062] Further preferably, the subject's motion gain is also obtained before the gaze stability training; and

[0063] If the visual target is a passive visual target, the step of determining the display position of the visual target includes:

[0064] Calculating the actual deflection angle of the visual target according to the motion gain of the subject and the collected actual head movement data of the subject;

[0065] Calculating the display position of the visual target according to the actual deflection angle and in combination with the current training mode;

[0066] If the visual target is an active visual target, the step of determining the display position of the visual target includes:

[0067] According to the current training difficulty, obtaining the movement frequency and standard deflection amplitude of the visual target;

[0068] calculating a target deflection amplitude according to the subject's motor gain and the standard deflection amplitude;

[0069] The display position of the visual target is calculated according to the movement frequency of the visual target and the target deflection amplitude in combination with the current training mode.

[0070] Further preferably, when determining the display position of the visual target, it also includes:

[0071] obtaining the distance between the subject and the screen;

[0072] If the visual target is an active visual target:

[0073] According to the current training difficulty, corresponding display parameters are obtained, wherein the display parameters include target movement frequency and target deflection amplitude, and standard viewing distance;

[0074] Calculating the offset distance of the visual target on the screen according to the distance between the subject and the screen, the target deflection amplitude, and the standard viewing distance;

[0075] Determining the position of the visual target at each time point during the training process according to the current training mode, the offset distance of the visual target, and the target movement frequency;

[0076] If the visual target is a passive visual target:

[0077] Acquiring a head movement amplitude according to the collected head movement data of the subject;

[0078] Determining the offset distance of the visual target on the screen according to the head movement amplitude and the distance between the subject and the screen;

[0079] The display position of the visual target on the screen is determined according to the current training mode and the offset distance.

[0080] Further preferably, it also includes:

[0081] Before the training begins, the distance between the subject and the screen is obtained, and the size of the visual target is adjusted accordingly.

[0082] The present invention also provides a gaze stability training and evaluation system, comprising:

[0083] A data acquisition device, used to collect motion data of a subject during gaze stability training; the motion data includes head movement data and / or eye movement data;

[0084] A display device for displaying a visual target for gaze stability training;

[0085] One or more processors and one or more memories, wherein the one or more memories store a computer program, and the computer program is loaded and executed by the one or more processors to implement the gaze stability training and evaluation method as described in any one of the above items.

[0086] The gaze stability training and evaluation method and system disclosed in this application have at least the following advantages over the prior art:

[0087] This application implements tracking and monitoring of the patient's training process, calculates training deviations in real time, and comprehensively controls visual targets to improve the patient's concentration during training. At the same time, the visual target can be adaptively adjusted according to the patient's training performance, which is conducive to the patient's own training, gradually progressing from easy to difficult, increasing the challenge of training. The entire system realizes a closed-loop process of gaze stability training, allowing patients to persist in rehabilitation training for a long time, which greatly solves the pain point of rehabilitation training being difficult to persist in.

[0088] In addition, the hardware structure adopted in this application is simple, which is convenient for setting up a training environment and easy for implementing home training. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] Figure 1 Schematic diagram of the training paradigm mentioned in the background technology of the present invention.

[0090] Figure 2 It is a flow chart of the gaze stability training and evaluation method of the present invention.

[0091] Figure 3 It is a schematic diagram of the target position calculation theory of the present invention.

[0092] Figure 4 Schematic diagram of the training difficulty assessment process of the present invention.

[0093] Figure 5 This is a process for evaluating the difficulty of subject training according to the present invention. DETAILED DESCRIPTION

[0094] The present invention is further described in detail below through the accompanying drawings and specific embodiments.

[0095] like Figure 1 As shown, a gaze stability training evaluation method provided by an embodiment of one aspect of the present invention includes:

[0096] Example 1: This example describes in detail the gaze stability training process of the subject.

[0097] S1. receiving a training instruction from a subject, wherein the training instruction includes a currently selected training mode and training difficulty;

[0098] The training modes include the following:

[0099] (1) Fixed gaze: A visual target appears in the middle of the screen (fixed), and the subject stares at the target while making back-and-forth movements of the head (such as shaking the head or nodding). In this training mode, the visual target is always fixed in the middle of the screen, and the subject stares at the target while maintaining back-and-forth movements of the head.

[0100] (2) Saccade: The visual target moves back and forth on the screen. The subject keeps his head still and follows the movement of the visual target with his eyes. The visual target in this saccade training mode should be an active visual target. The corresponding display parameters can be obtained based on the current training mode and training difficulty, thereby controlling the display of the visual target.

[0101] (3) Reverse tracking:

[0102] If an active visual target is used, then: the visual target moves back and forth on the screen, the subject's eyes stare at the visual target, but the head moves in the opposite direction.

[0103] If a passive visual target is used, the visual target moves in the opposite direction to the subject's head movement, and the subject stares at the moving visual target; the display of the passive visual target is based on the current training mode and is controlled by the user's (subject's) motion data.

[0104] (4) Tracking in the same direction:

[0105] If an active visual target is used: the visual target moves back and forth on the screen, the subject's eyes stare at the visual target, and the head moves in the same direction as the visual target.

[0106] If a passive visual target is used: the visual target moves in the same direction as the subject's head movement, and the subject stares at the moving visual target during the movement. S2. Based on the training instruction, the corresponding visual target is displayed on the screen; the visual target is an active visual target or a passive visual target; wherein the active visual target is displayed according to the preset display parameters, and the passive visual target is displayed according to the subject's movement data;

[0107] If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; and the motion data of the subject during the training process is collected; the motion data includes head movement data and / or eye movement data;

[0108] If the visual target is a passive visual target, the motion data of the subject during the training process is collected, and the display position of the visual target is determined according to the current training mode and the motion data of the subject;

[0109] Specifically, regarding the collection of motion data, it can be collected through a head-mounted device worn by the subject, such as an eye tracker or a video nystagmus, which can collect eye movement data and head movement data; of course, it is also possible to collect only one of the head movement data or the eye movement data, such as a head-mounted device with a built-in posture sensor, which only collects head movement data. Of course, the collection of the motion data may also not be through a head-mounted device, such as by shooting the subject's training process through an external camera, especially the training image of the head (including the eyes), and then performing image processing and analysis to obtain the subject's head movement data and / or eye movement data. Therefore, this embodiment does not limit the method or device for collecting motion data.

[0110] S3. Evaluate the current gaze stability training according to the motion data and the display position data of the visual target.

[0111] Specifically include:

[0112] Determine ideal motion data based on the display position data of the visual target and the current training mode;

[0113] Time-aligning the collected actual motion data with the ideal motion data;

[0114] Calculating the deviation between the actual motion data and the ideal motion data;

[0115] Evaluate the current gaze stability training based on the calculated deviation results.

[0116] Furthermore, the deviation between the actual motion data and the ideal motion data is calculated, specifically including:

[0117] Performing curve fitting or tracing on the ideal motion data to obtain an ideal motion curve;

[0118] Performing curve fitting or tracing on the actual motion data to obtain an actual motion curve;

[0119] The system receives and processes the patient's actual head movement and / or eye movement information. The actual head movement information obtained is any one or more of the head movement deflection angle, speed, acceleration, etc., and can be recorded in the form of a curve. The eye movement information obtained is any one or more of the eye movement deflection angle, speed, acceleration, etc., and can be recorded in the form of a curve. The eye movement information can also be the patient's gaze point movement information obtained by estimating the gaze point.

[0120] Calculate the difference between the actual motion curve and the ideal motion curve. The calculation method includes but is not limited to any of the following calculation methods:

[0121] 1. Area difference calculation method: calculate the area between the actual motion curve and the ideal motion curve, and obtain the difference between the actual motion curve and the ideal motion curve according to the area;

[0122] Specifically, if the curve is represented by a set of discrete data points, a numerical integration method (such as the trapezoidal method or the Simpson method) can be used to approximate the area. If the curve is represented by an analytical function, the area can be directly calculated using symbolic integration or numerical integration methods.

[0123] 2. Motion spectrum difference calculation method: Fourier transform the actual motion curve and the ideal motion curve respectively to obtain the actual motion spectrum and the ideal motion spectrum; extract the main spectrum of the actual motion spectrum and the ideal motion spectrum respectively; calculate the difference between the actual motion main spectrum and the ideal motion main spectrum. Specifically, for example, the patient's actual head movement and ideal head movement information data are recorded in the form of curves respectively. Fourier transform the patient's head movement curve and the ideal head movement curve respectively, and compare the spectrum differences between the two. For example, the frequency component with the largest amplitude in the spectrum is taken as the main frequency, and the main frequency difference is calculated.

[0124] 3. Motion shape distance difference calculation method: Assume that we have two motion curves, one is the ideal motion curve (Ideal Motion Curve, IMC) and the other is the actual motion curve (Actual Motion Curve, AMC). Our goal is to quantify the difference between AMC and IMC. Includes the following Let's take a look at the specific processing process:

[0125] Step 1: Extract feature points at equal intervals from the actual motion curve and the ideal motion curve; for example, extract feature points at equal intervals from the ideal curve and the actual curve, such as extracting one point every 0.1 seconds.

[0126] Step 2: For each feature point, calculate its relative position with respect to other points to form a shape context descriptor; the shape context descriptor is used to describe shape information based on feature points. For example, for each feature point, calculate its relative distance and angle with other feature points on the same curve to form a shape context descriptor.

[0127] Step 3: Calculate the matching cost between the shape context descriptors of the actual motion curve and the ideal motion curve; for each pair of corresponding feature points (from IMC and AMC), calculate the difference between their shape context descriptors. This can be achieved by comparing the difference in relative distance and angle.

[0128] Step 4: Accumulate the matching costs of all feature points to obtain the total shape distance; the shape distance is used to represent the difference between the actual motion curve and the ideal motion curve. In other words, add up the matching costs of all feature points to obtain the total shape distance. This sum will quantify the overall difference between AMC and IMC. For example, when performing shape matching between the patient's head motion curve and the ideal head motion curve, the shape distance can be calculated through the shape context; the Hausdorff distance is calculated to represent the difference between the two. This total shape distance can be used to evaluate the closeness between AMC and IMC. If the total shape distance is small, it means that AMC is very close to IMC; if it is large, it means that the difference between the two is significant.

[0129] Of course, in addition to the above methods, there are many schemes for calculating the actual motion curve and the ideal motion curve, such as aligning the patient's actual head movement data with the ideal head movement data on the time axis, and calculating the difference between the patient's actual head movement information and the ideal head movement information data at each time point; or performing error analysis on the patient's actual head movement data and the ideal head movement data to calculate the mean square error or root mean square error between the two.

[0130] The above-mentioned actual motion data includes actual head movement data and / or actual eye movement data; the ideal motion data includes ideal head movement data and / or ideal eye movement data. According to the above evaluation method, the system can obtain and feedback the difference between the actual head movement data and the ideal head movement data, and / or the difference between the actual eye movement data and the ideal eye movement data in real time, so as to facilitate the patient and the rehabilitation trainer to intuitively understand the training situation and form prompts, which can be visual prompts, sound prompts, tactile prompts, etc. Patients can adjust or correct the training movements in real time according to the prompts.

[0131] Embodiment 2: This embodiment describes in detail the processing process of the training evaluation results.

[0132] After the gaze stabilization training, adjusting the training difficulty of the subject according to the evaluation results of multiple gaze stabilization trainings; specifically including:

[0133] Recording the assessment score of the subject in each gaze stability training at the current training difficulty;

[0134] When the assessment score of the subject reaches a first set score value, increasing a first cumulative count;

[0135] When the subject's assessment score is lower than a second set score value, increasing a second cumulative count; wherein the first set score value is greater than or equal to the second set score value;

[0136] If the first cumulative count is greater than a preset first cumulative number threshold, or the second cumulative count is greater than a preset second cumulative number threshold, it is determined that the current training difficulty does not match the subject; wherein the first cumulative number is the same as or different from the second cumulative number;

[0137] If the current training difficulty does not match the subject, a training difficulty matching the subject is recommended based on the subject's motion data.

[0138] Specifically, after gaze stabilization training, it can be determined whether the current training difficulty matches the subject based on the evaluation results of multiple gaze stabilization trainings. If not, the training difficulty of the subject can be adjusted in time. The evaluation results of multiple gaze stabilization trainings here can be the evaluation results of gaze stabilization training under the current mode and training difficulty of the set number of times, such as evaluating the matching degree between the subject and the reverse tracking gaze stabilization training with difficulty level 3 based on the last 5 reverse tracking gaze stabilization trainings with difficulty level 3. Of course, the basis for the difficulty matching evaluation can also be the evaluation results of gaze stabilization training within a set period or a set number of times within a set period.

[0139] If the first set score value is set to 80 points, the second set score value is set to 60 points, the first cumulative number threshold is 4, and the second cumulative number threshold is 5; in the current training mode, the training difficulty is evaluated according to the evaluation results of the last 6 gaze stability trainings. After these 6 gaze stability trainings, the training evaluation results are recorded each time. If the evaluation scores of the 6 times are statistically recorded as 78, 85, 88, 90, 93, and 89, respectively; it can be seen that the first cumulative count is 5 and the second cumulative count is 0; since the first cumulative count 5 is greater than the preset first cumulative number threshold 4, it is determined that the current difficulty does not match the subject, and a higher level of training can be recommended to the subject. Of course, if the 6 gaze stability trainings are statistically lower than 60 points, then it is obvious that the second cumulative count 6 is greater than the preset second cumulative number threshold 5, then it can also be determined that the current difficulty is too difficult for the subject and does not match the subject's current level, and it can be recommended to reduce the difficulty. Of course, if the first cumulative count is not greater than the preset first cumulative number threshold, and the second cumulative count is not greater than the preset second cumulative number threshold, it means that the current training difficulty matches the subject.

[0140] In this embodiment, by reasonably setting the threshold and score, the matching degree between the current training difficulty and the subject can be accurately determined, and in the case of mismatch, the corresponding training difficulty can be recommended based on the subject's motion data.

[0141] Preferably, based on the subject's exercise data, a training difficulty matching the subject is recommended; specifically including:

[0142] Calculating visual target parameters according to the motion data of the subject collected during the training process, and determining the matching training difficulty according to the visual target parameters; wherein:

[0143] If the motion data only includes head movement data, the head movement frequency and head movement deflection amplitude are determined according to the head movement data of the subject; the visual target parameters are determined according to the head movement frequency and head movement deflection amplitude to recommend matching training difficulty; specifically, if the head-mounted device worn by the subject cannot collect eye movement information, but has a built-in gyroscope that can collect head movement data, the training results are evaluated during the training process through the collected head movement data. Before training, the subject is prompted to keep his eyes on the visual target and move his head according to the training requirements. During the training process, the head movement data is collected through the gyroscope of the head-mounted device; finally, the head movement frequency and head movement deflection amplitude are obtained based on the most recent one or more head movement data collected, for example, the average head movement frequency and head movement deflection amplitude are calculated based on the most recent three head movement data, and then the visual target parameters are determined based on the average head movement frequency and head movement deflection amplitude, that is, the movement frequency and deflection amplitude displayed by the visual target.

[0144] The inventor discovered during the research and development test that some subjects may score low in gaze stabilization training with a low difficulty level, but score high in gaze stabilization training with a high difficulty level (faster target movement speed or higher frequency or larger deflection amplitude), and the subject may be more suitable for gaze stabilization training with a high difficulty level. This is because in low-difficulty training, the target moves slowly, and the subject may unconsciously practice at a faster head movement speed. Many subjects with good vestibular function may also show natural head acceleration and movement amplitude under slow conditions, which in turn affects the precise control of low-speed movement and affects the score. Therefore, new visual target parameters can be determined based on the head movement data of the subject during training, including but not limited to head movement speed, head movement amplitude, head movement frequency, etc., and then the target is controlled to be displayed according to the new visual target parameters.

[0145] Preferably, when making difficulty matching recommendations, matching training difficulties may also be found based on evaluation results of gaze stability training tests of subjects at different difficulty levels.

[0146] If the motion data only includes eye movement data, the deviation between the gaze point and the visual target is calculated according to the eye movement data of the subject; if the deviation between the gaze point and the visual target is not within a set error range, it is recommended to reduce the training difficulty;

[0147] If the motion data includes the head movement data and the eye movement data, the deviation between the gaze point and the visual target is calculated based on the eye movement data; if the deviation between the gaze point and the visual target is within a set error range, the head movement frequency and the deflection amplitude are obtained based on the head movement data, and the visual target parameters are determined based on the head movement frequency and the deflection amplitude to recommend a matching training difficulty.

[0148] Regarding the calculation of the gaze point, a variety of calculation methods can be used, which are not limited in this embodiment. For example, geometric operations can be used to obtain the head posture and eye rotation position and then calculate the line of sight direction. Combined with the distance between the user and the screen, the intersection of the line of sight on the screen is calculated. Of course, it is also possible to combine artificial intelligence technology and use a trained neural network model to predict the coordinates of the gaze point on the screen through the input eye movement data and head movement data. Preferably, in order to improve the accuracy of the gaze point prediction, the data input to the model includes, in addition to the head movement and eye movement data, the position data or distance data between the subject and the screen.

[0149] In another embodiment of the present application, based on the above embodiment, when the assessment score of the subject reaches a first set score value, in addition to increasing the first cumulative count, a first continuous count is also increased, and if the set score value is not reached, the first continuous count is reset;

[0150] If the first cumulative count reaches a preset first cumulative number of times, and the first continuous count reaches a preset continuous number of times, the training difficulty is increased.

[0151] Specifically, in this embodiment, if the training difficulty of the subject is to be increased, in addition to the first cumulative count reaching the preset first cumulative number of times, the first continuous count must also reach the preset continuous number of times, for example, the preset first cumulative number threshold is 5, and the preset continuous number of times is 3; then the subject needs to reach 80 points (the first preset score value) in the current difficulty of gaze stabilization training for 5 times, and at least 80 points or more in 3 consecutive gaze stabilization trainings. In this way, the recommendation of increasing difficulty can be made more accurate.

[0152] Example 3: This example describes in detail the preparation process before the subject training begins, specifically including:

[0153] Evaluate the difficulty of training, adjust the position of the visual target according to the motor gain, and adjust the size of the visual target according to the distance between the visual target and the subject;

[0154] 1) As Figure 4 As shown, evaluating the training difficulty of the subject; specifically comprising the following steps:

[0155] S01, displaying a visual target on the screen, and prompting the subject to quickly turn his head while being able to see the visual target clearly;

[0156] S02, collecting actual head movement data of the subject;

[0157] S03, calculating the subject's head movement frequency and head movement deflection angle according to the subject's actual head movement data;

[0158] S04. Determine a training difficulty matched to the subject according to the head movement frequency and the head movement deflection angle of the subject.

[0159] Specifically, the assessment of the difficulty of the training can be applied to the difficulty assessment before the formal gaze stabilization training begins. The assessment can use a passive visual target, and the movement of the visual target is determined by the subject's head movement data. The subject is told to turn his head under the premise of being able to see the visual target (visual target) clearly, the rotation angle is a given value, and the speed is limited to the fastest speed that he can tolerate. In this way, the head movement curve is fitted, and the curve frequency and the corresponding visual target parameter (difficulty parameter) are calculated.

[0160] Furthermore, if the motion data includes head movement data and eye movement data, then evaluating the training difficulty of the subject further includes the steps of:

[0161] S041. While collecting the actual head movement data of the subject, collecting the actual eye movement data of the subject through an eye tracking device;

[0162] S042, synchronously recording the position data of the visual target on the screen;

[0163] S043, calculating the gaze point position of the subject according to the actual head movement data and the actual eye movement data of the subject;

[0164] S044, calculating the gaze deviation between the gaze point position of the subject and the position data of the visual target at each time point;

[0165] S045. Counting the percentage of fixation points whose fixation deviations meet a preset deviation range;

[0166] S046. If the proportion reaches a preset ratio, determining a training difficulty matched to the subject based on the subject's actual head movement data; specifically including:

[0167] Calculating the head movement frequency and the head movement deflection angle of the subject according to the actual head movement data of the subject;

[0168] The training difficulty matched to the subject is determined according to the head movement frequency and the head movement deflection angle of the subject.

[0169] Among them, different training difficulty settings have different display parameters, including movement frequency and deflection amplitude;

[0170] If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; specifically including:

[0171] Determining a display track of the visual target according to a current training mode;

[0172] Obtain corresponding display parameters according to the current training difficulty, the display parameters including target movement frequency and target deflection amplitude;

[0173] According to the target movement frequency and the target deflection amplitude, combined with the current training mode, the display of the visual target on the screen is controlled.

[0174] 2) obtaining the subject's motor gain before gaze stability training; and

[0175] If the visual target is a passive visual target, the step of determining the display position of the visual target includes:

[0176] Calculating the actual deflection angle of the visual target according to the motion gain of the subject and the collected actual head movement data of the subject;

[0177] Calculating the display position of the visual target according to the actual deflection angle and in combination with the current training mode;

[0178] If the visual target is an active visual target, the step of determining the display position of the visual target includes:

[0179] According to the current training difficulty, obtaining the movement frequency and standard deflection amplitude of the visual target;

[0180] calculating a target deflection amplitude according to the subject's motor gain and the standard deflection amplitude;

[0181] The display position of the visual target is calculated according to the movement frequency of the visual target and the target deflection amplitude in combination with the current training mode.

[0182] When training gaze stabilization, the display position of the visual target is determined in combination with the distance between the subject and the screen, specifically including:

[0183] obtaining the distance between the subject and the screen;

[0184] If the visual target is an active visual target:

[0185] According to the current training difficulty, corresponding display parameters are obtained, wherein the display parameters include target movement frequency and target deflection amplitude, and standard viewing distance;

[0186] Calculating the offset distance of the visual target on the screen according to the distance between the subject and the screen, the target deflection amplitude, and the standard viewing distance;

[0187] Determining the position of the visual target at each time point during the training process according to the current training mode, the offset distance of the visual target, and the target movement frequency;

[0188] If the visual target is a passive visual target:

[0189] Acquiring a head movement amplitude according to the collected head movement data of the subject;

[0190] Determining the offset distance of the visual target on the screen according to the head movement amplitude and the distance between the subject and the screen;

[0191] The display position of the visual target on the screen is determined according to the current training mode and the offset distance.

[0192] Specifically, the distance between the subject and the screen may be obtained by inputting distance data by a user, or by detecting and obtaining the distance data through a distance measuring sensor on a head mounted device, or by scanning through an external distance measuring sensor or camera, and this application does not impose any restrictions on this.

[0193] If the visual target is an active visual target, the corresponding target motion frequency, target deflection amplitude and standard viewing distance can be obtained according to the current training difficulty level; if the distance between the subject and the screen is consistent with the standard viewing distance, the target motion frequency and target deflection amplitude can be directly used to control the display of the visual target. If the distance between the subject and the screen is inconsistent with the standard viewing distance (the deviation exceeds the preset range), it is also necessary to determine the new target deflection amplitude based on the ratio of the two (the maximum offset distance, a simple geometric calculation can be performed, which will not be repeated this time), and then control the visual target to be displayed according to the target motion frequency and the new target deflection amplitude.

[0194] If the visual target is a passive visual target, it is necessary to collect the subject's head movement data, including the head movement amplitude, and then determine the offset distance of the visual target based on the distance between the subject and the screen. Then, the display position of the visual target is determined in combination with the current training mode and the real-time offset distance.

[0195] In another illustrative example of the present application, the subject's motion gain data is also combined in the process of determining the display position of the visual target. Figure 3 As shown, when the difficulty gain is set to g, the patient's head turns left by an angle a, and the visual target moves right synchronously. The moving distance y is calculated based on the distance x between the screen and the patient:

[0196]

[0197] In another illustrative embodiment, the size of the visual target can be adjusted according to the distance between the subject and the screen, so that if the subject stands far away, the visual target will be enlarged accordingly, the subject can see the visual target clearly, and the training effect will not be affected.

[0198] Compared with gaze stability training conducted in hospitals, the distance between the patient and the external screen is uncontrollable during home training due to individual differences in home decoration. If the visual target remains unchanged, if the distance is too close, the patient will turn his head too much and it will be difficult for him to stare at the visual target throughout the whole process. If the distance is too far, the visual target will not be visible and the head turning angle will be too small. We add a distance measuring device to automatically measure the distance between the patient and the external screen to control the display parameters of the visual target on the external screen. Patients do not need to measure the distance by themselves, which reduces the requirements for preparation before home training. The distance measuring device can be a laser rangefinder, camera and other equipment.

[0199] At present, the main hardware components of the system include head movement information collection equipment, eye movement information collection equipment, distance measurement device and screen. Among them, the screen is required, and either the head movement information collection equipment or the eye movement information collection equipment or both are selected, and the distance measurement device is optional. The above components can exist separately or be integrated into one, such as integrated into VR equipment.

[0200] According to the above evaluation method, the system can provide real-time feedback of difference data and form prompts, which can be visual prompts, sound prompts, tactile prompts, etc. Patients can adjust or correct training movements in real time according to the prompts.

[0201] This system can process the difference data obtained after the evaluation to facilitate patients and rehabilitation therapists to intuitively understand the training situation. If a single parameter is used, the training score can be obtained by processing methods such as standardized value and threshold setting. If multiple parameters are used for comprehensive evaluation, the training score can be obtained by weighted calculation.

[0202] This system can record the training score of each training session, and patients and rehabilitation therapists can track the training status and evaluate the training effect.

[0203] The system can generate customized training reports, output difference data, single / historical training scores, head motion curves, etc. The system can also provide corresponding training guidance and analysis based on the training scores. For example, when the main frequency of the patient's head motion curve calculated during three consecutive training evaluations is significantly different from that of the ideal head motion curve, the patient and the rehabilitation therapist can be prompted to adjust the training parameters.

[0204] Obviously, the above embodiments are only examples for clear explanation, and are not intended to limit the implementation methods. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from them are still within the protection scope of the invention.

Claims

1. A gaze stability training evaluation method, It is characterized in that The following steps are involved: receiving a training instruction from a subject, wherein the training instruction includes a currently selected training mode and a training difficulty; Based on the training instruction, a corresponding visual target is displayed on the screen; the visual target is an active visual target or a passive visual target; wherein the active visual target is displayed according to a preset display parameter, and the passive visual target is displayed according to the motion data of the subject; If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; and the motion data of the subject during the training process is collected; the motion data includes head movement data and / or eye movement data; If the visual target is a passive visual target, the motion data of the subject during the training process is collected, and the display position of the visual target is determined according to the current training mode and the motion data of the subject; The current gaze stability training is evaluated according to the motion data and the display position data of the visual target.

2. The gaze stability training and evaluation method according to claim 1, It is characterized in that Before the gaze stability training begins, the method further includes: evaluating the training difficulty of the subject; specifically, the following steps are included: Displaying a visual target on a screen, and prompting the subject to quickly turn his head while being able to see the visual target clearly; collecting actual head movement data of the subject; Calculating the head movement frequency and the head movement deflection angle of the subject according to the actual head movement data of the subject; The training difficulty matched to the subject is determined according to the head movement frequency and the head movement deflection angle of the subject.

3. The gaze stability training and evaluation method according to claim 2, It is characterized in that The method of evaluating the training difficulty of the subject further comprises the steps of: While collecting the actual head movement data of the subject, collecting the actual eye movement data of the subject through an eye tracking device; Synchronously recording the position data of the visual target on the screen; Calculating the gaze point position of the subject according to the actual head movement data and the actual eye movement data of the subject; Calculating the gaze deviation between the gaze point position of the subject and the position data of the visual target at each time point; Counting the percentage of fixation points whose fixation deviations meet the preset deviation range; If the proportion reaches a preset ratio, the training difficulty matched to the subject is determined based on the actual head movement data of the subject.

4. The gaze stability training and evaluation method according to claim 1, It is characterized in that Different training difficulty settings have different display parameters, including movement frequency and deflection amplitude; If the visual target is an active visual target, the display position of the visual target is determined according to the current training mode and training difficulty; specifically including: Determining a display track of the visual target according to a current training mode; Acquire corresponding display parameters according to the current training difficulty, wherein the display parameters include target movement frequency and target deflection amplitude; According to the target movement frequency and the target deflection amplitude, combined with the current training mode, the display of the visual target on the screen is controlled.

5. The gaze stability training and evaluation method according to claim 1, It is characterized in that The step of evaluating the current gaze stability training according to the motion data and the display position data of the visual target specifically includes: Determine ideal motion data based on the display position data of the visual target and the current training mode; Time-aligning the collected actual motion data with the ideal motion data; Calculating the deviation between the actual motion data and the ideal motion data; Evaluate the current gaze stability training based on the calculated deviation results.

6. The gaze stability training and evaluation method according to claim 5, It is characterized in that The calculation of the deviation between the actual motion data and the ideal motion data specifically includes: Performing curve fitting or tracing on the ideal motion data to obtain an ideal motion curve; Performing curve fitting or tracing on the actual motion data to obtain an actual motion curve; The difference between the actual motion curve and the ideal motion curve is calculated.

7. The gaze stability training and evaluation method according to claim 6, It is characterized in that The difference between the actual motion curve and the ideal motion curve is calculated by any of the following calculation methods: Area difference calculation method: calculating the area between the actual motion curve and the ideal motion curve, and obtaining the difference between the actual motion curve and the ideal motion curve according to the area; or Motion spectrum difference calculation method: performing Fourier transform on the actual motion curve and the ideal motion curve respectively to obtain the actual motion spectrum and the ideal motion spectrum; Extracting main spectra of the actual motion spectrum and the ideal motion spectrum; Calculate the difference between the actual motion dominant spectrum and the ideal motion dominant spectrum; or Motion shape distance difference calculation method: Extracting feature points at equal intervals from the actual motion curve and the ideal motion curve; For each feature point, calculate its relative position to other points to form a shape context descriptor; Calculating the matching cost between the shape context descriptors of the actual motion curve and the ideal motion curve; The matching costs of all feature points are accumulated to obtain a total shape distance; the shape distance is used to represent the difference between the actual motion curve and the ideal motion curve.

8. The gaze stability training and evaluation method according to claim 1, It is characterized in that Also includes: After the gaze stabilization training, adjusting the training difficulty of the subject according to the evaluation results of the multiple gaze stabilization training; Specifically include: Recording the assessment score of the subject in each gaze stability training at the current training difficulty; When the assessment score of the subject reaches a first set score value, increasing a first cumulative count; When the subject's assessment score is lower than a second set score value, increasing a second cumulative count; wherein the first set score value is greater than or equal to the second set score value; If the first cumulative count is greater than a preset first cumulative number threshold, or the second cumulative count is greater than a preset second cumulative number threshold, it is determined that the current training difficulty does not match the subject; wherein the first cumulative number is the same as or different from the second cumulative number; If the current training difficulty does not match the subject, a training difficulty matching the subject is recommended based on the subject's motion data.

9. The gaze stability training and evaluation method according to claim 8, It is characterized in that When the assessment score of the subject reaches a first set score value, in addition to increasing the first cumulative count, a first continuous count is also increased, and if the set score value is not reached, the first continuous count is reset; If the first cumulative count reaches a preset first cumulative number of times, and the first continuous count reaches a preset continuous number of times, the training difficulty is increased.

10. The gaze stability training and evaluation method according to claim 8, It is characterized in that The step of recommending a training difficulty matching the subject based on the subject's exercise data specifically includes: Calculating visual target parameters according to the motion data of the subject collected during the training process, and determining the matching training difficulty according to the visual target parameters; wherein: If the motion data only includes head movement data, determining the head movement frequency and the head movement deflection amplitude according to the head movement data of the subject; determining the visual target parameter according to the head movement frequency and the head movement deflection amplitude to recommend a matching training difficulty; If the motion data only includes eye movement data, the deviation between the gaze point and the visual target is calculated according to the eye movement data of the subject; if the deviation between the gaze point and the visual target is not within a set error range, it is recommended to reduce the training difficulty; If the motion data includes the head movement data and the eye movement data, the deviation between the gaze point and the visual target is calculated based on the eye movement data; if the deviation between the gaze point and the visual target is within a set error range, the head movement frequency and the deflection amplitude are obtained based on the head movement data, and the visual target parameters are determined based on the head movement frequency and the deflection amplitude to recommend a matching training difficulty.

11. The gaze stability training and evaluation method according to claim 1, It is characterized in that The subject's motor gain is also obtained before gaze stability training; and If the visual target is a passive visual target, the step of determining the display position of the visual target includes: Calculating the actual deflection angle of the visual target according to the motion gain of the subject and the collected actual head movement data of the subject; Calculating the display position of the visual target according to the actual deflection angle and in combination with the current training mode; If the visual target is an active visual target, the step of determining the display position of the visual target includes: According to the current training difficulty, obtaining the movement frequency and standard deflection amplitude of the visual target; calculating a target deflection amplitude according to the subject's motor gain and the standard deflection amplitude; The display position of the visual target is calculated according to the movement frequency of the visual target and the target deflection amplitude in combination with the current training mode.

12. The gaze stability training and evaluation method according to claim 1, It is characterized in that When determining the display position of the visual target, it also includes: obtaining the distance between the subject and the screen; If the visual target is an active visual target: According to the current training difficulty, corresponding display parameters are obtained, wherein the display parameters include target movement frequency and target deflection amplitude, and standard viewing distance; Calculating the offset distance of the visual target on the screen according to the distance between the subject and the screen, the target deflection amplitude, and the standard viewing distance; Determining the position of the visual target at each time point during the training process according to the current training mode, the offset distance of the visual target, and the target movement frequency; If the visual target is a passive visual target: Acquiring a head movement amplitude according to the collected head movement data of the subject; Determining the offset distance of the visual target on the screen according to the head movement amplitude and the distance between the subject and the screen; The display position of the visual target on the screen is determined according to the current training mode and the offset distance.

13. The gaze stability training and evaluation method according to claim 1, It is characterized in that Also includes: Before the training begins, the distance between the subject and the screen is obtained, and the size of the visual target is adjusted accordingly.

14. A gaze stability training and evaluation system, It is characterized in that include: A data acquisition device, used to collect motion data of a subject during gaze stability training; the motion data includes head movement data and / or eye movement data; A display device for displaying a visual target for gaze stability training; One or more processors and one or more memories, wherein the one or more memories store a computer program, and the computer program is loaded and executed by the one or more processors to implement the gaze stability training and evaluation method as described in any one of claims 1-13.