Cognitive disorder evaluation system and method in combination with eye movement data
By designing task design modules, data acquisition, preprocessing and feature extraction modules, combined with eye movement data and BP neural network, cognitive impairment in subjects is evaluated, solving the problem of inaccurate evaluation in the prior art, and achieving a more accurate assessment of cognitive impairment.
Patent Information
- Application Number
- CN202510392874.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-31
- Publication Date
- 2025-07-25
AI Technical Summary
How to more accurately evaluate whether subjects have cognitive impairment in combination with eye movement data.
Design task design module, data acquisition module, data preprocessing module and feature extraction module, use eye tracking instruments to collect eye movement data of subjects when performing visual processing tasks, and evaluate cognitive impairment through BP neural network to obtain the probability value of cognitive impairment.
A more accurate assessment of subjects’ cognitive impairment is achieved, and the combination of eye movement data and BP neural networks is used to provide more accurate assessment results.
Smart Images
Figure CN120376128A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of eye movement data processing, and particularly relates to a cognitive impairment assessment system and method combining eye movement data. Background Art
[0002] Eye movement data refers to a series of data recorded through specific technologies and devices, reflecting the movement state and characteristics of the eyes. These data can reveal the eye behavior patterns of people during the processes of observing things, performing cognitive activities, etc. An eye tracker is a device used to collect eye movement data.
[0003] Cognition is a high-level function of the brain, involving multiple neuropsychological processes, such as attention, memory, language ability, executive function, visuospatial ability, etc. Neuropsychological theory holds that different cognitive functions are responsible for different regions of the brain. When these regions are diseased or damaged, the corresponding cognitive functions will be affected. By designing a series of test tasks for different cognitive domains, such as evaluating attention and short-term memory through digit span tests, and evaluating language function through word fluency tests, the function status of each region of the brain can be understood, and whether there is cognitive impairment and the specific manifestations and degrees of the impairment can be judged.
[0004] How to more accurately evaluate whether a subject has cognitive impairment by combining eye movement data is an urgent problem to be solved. For this reason, a cognitive impairment assessment system and method combining eye movement data are proposed. Summary of the Invention
[0005] The technical problem to be solved by the present invention is: how to more accurately evaluate whether a subject has cognitive impairment by combining eye movement data, and a cognitive impairment assessment system combining eye movement data is provided.
[0006] The present invention solves the above technical problem through the following technical solutions. The present invention includes a task design module, a data acquisition module, a data preprocessing module, a feature extraction module, and an evaluation module;
[0007] The task design module is used to design the visual processing tasks that the subject needs to perform;
[0008] The data acquisition module is used to collect the eye movement data of the subject when performing the visual processing tasks by using an eye tracker;
[0009] The data preprocessing module is used to preprocess some of the collected eye movement data;
[0010] The feature extraction module is used to process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network;
[0011] The evaluation module is used to evaluate the cognitive impairment of the subject by using the trained BP neural network, and obtain the probability value of the existence of cognitive impairment of the current subject.
[0012] Furthermore, the task design module includes a first task design unit, a second task design unit, and a task display unit; the first task design unit is used to design the visual search task and record the position information of the target object area on the display interface, and the second task design unit is used to design the text reading task and record the position information of each text on the display interface. The task display unit is used to display the designed visual search task and text reading task on the display interface.
[0013] Furthermore, the purpose of the visual search task is to find the target object, which is used to evaluate the attention and executive function of the subject, and the purpose of the text reading task is to track and read the text, which is used to evaluate the language processing ability of the subject.
[0014] Furthermore, the data acquisition module includes a first eye movement data acquisition unit and a second eye movement data acquisition unit; the first eye movement data acquisition unit is used to acquire the eye movement data of the subject when performing the visual search task, including the saccade speed Vs, saccade direction Ds, number of fixation points C1s, fixation point coordinates Ls C1 and the fixation duration Ts C1 , where s indicates that the corresponding data is the data when performing the visual search task; the second eye movement data acquisition unit is used to acquire the eye movement data of the subject during the text reading task, including the number of fixation points C2w, fixation point coordinates Lw C2 and the fixation duration Tw C2 , where w indicates that the corresponding data is the data when performing the text reading task.
[0015] Furthermore, the data preprocessing module includes a first preprocessing unit and a second preprocessing unit. The first preprocessing unit is used to preprocess the number of fixation points C1s, fixation point coordinates Ls C1 and the fixation duration Ts C1 to obtain the preprocessed number of fixation points C1's, fixation point coordinates Ls C1＇ and the fixation duration Ts C1＇ ; the second preprocessing unit is used to preprocess the number of fixation points C2w, fixation point coordinates Lw C2 and the fixation duration Tw C2 to obtain the preprocessed number of fixation points C2'w, fixation point coordinates Lw C2＇ and the fixation duration Tw C2＇ .
[0016] Furthermore, the specific processing process of the feature extraction module is as follows:
[0017] S41: Obtain the number of fixation points C1′s and fixation point coordinates Ls after preprocessing C1＇ And the fixation duration Ts C1＇ , according to the gaze point coordinate Ls C1＇ And the location information of the target object area on the display interface, determine the number of fixations Zs that fall into the target object area i And the fixation duration Ts corresponding to each fixation point i , calculate the fixation duration Ts i The sum of t , where i represents the i-th fixation point falling into the target object area, and the position information of the target object area on the display interface is the coordinates of each point on the contour line of the target object;
[0018] S42: Obtaining the number of gaze points C2′w and the gaze point coordinates Lw C2＇ and the fixation duration Tw C2＇ , according to the number of gaze points C2′w and the gaze point coordinates Lw C2＇ Generate the gaze point text reading trajectory L1 and calculate the gaze point duration Tw C2＇ The sum of t ; Generate the original text trajectory line L0 according to the position information of each character on the display interface; wherein the position information of each character on the display interface is the coordinate of the center point of the minimum circumscribed rectangle of each character;
[0019] S43: Calculate the similarity value between the text reading trajectory L1 and the original text trajectory line L0 based on the Euclidean distance, denoted as D L ;
[0020] S44: The number of fixations Zs during visual search tasks i , fixation duration Ts i The sum of Ts t , scanning speed Vs, scanning direction Ds, and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 when performing the text reading task L , fixation duration Tw C2＇ The sum of Tw t Normalization is performed to obtain the normalized number of fixations Zs when performing the visual search task. i '', fixation duration Ts i The sum of Ts t ', scanning speed Vs', scanning direction Ds', and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 when performing the text reading taskL '', the fixation duration Tw C2＇ sum Tw t '', and use the above six types of normalized data as the input feature parameters of the BP neural network.
[0021] Furthermore, the specific processing process of the evaluation module is as follows:
[0022] S51: Input the sum Ts of the number of fixation points Zs, i '', the fixation duration Ts i sum Ts t '', the saccade speed Vs'', the saccade direction Ds'', and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 during the text reading task L '', the fixation duration Tw C2＇ sum Tw t '' into the input layer of the BP neural network;
[0023] S52: Process the six input feature parameters in step S51 through the hidden layer in turn, and finally output the probability value of the existence of cognitive impairment of the current subject through the output layer.
[0024] The present invention also provides a cognitive impairment evaluation method combining eye movement data. Use the above-mentioned cognitive impairment evaluation system combining eye movement data to evaluate whether a subject has cognitive impairment, including the following steps:
[0025] S1: Design the visual processing tasks that the subject needs to perform;
[0026] S2: Use an eye tracker to collect the eye movement data of the subject during the visual processing task;
[0027] S3: Preprocess some of the collected eye movement data;
[0028] S4: Process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network;
[0029] S5: Combine the input feature parameters in step S4, and use the trained BP neural network to evaluate the cognitive impairment of the subject to obtain the probability value of the existence of cognitive impairment of the current subject.
[0030] The present invention has the following advantages compared with the prior art: By using an eye tracker to collect the eye movement data of a subject when performing a visual processing task, intermediate parameters such as the number of fixation points falling within the target object area and the similarity value of the trajectory line are innovatively designed, which can well characterize the state of the subject when performing the corresponding task; The BP neural network is also used to evaluate the cognitive impairment of the subject in combination with the input characteristic parameters of the component, and the probability value of the existence of the cognitive impairment of the current subject can be obtained, realizing a more accurate evaluation of whether the subject has cognitive impairment in combination with the eye movement data. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 is a schematic block diagram of the structure of a cognitive impairment evaluation system combining eye movement data in an embodiment of the present invention;
[0032] Figure 2 is a schematic flowchart of the implementation process of a cognitive impairment evaluation method combining eye movement data in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0033] The embodiments of the present invention will be described in detail below. The following embodiments are implemented on the premise of the technical solution of the present invention, and detailed implementation manners and specific operation processes are given. However, the protection scope of the present invention is not limited to the following embodiments.
[0034] As Figure 1 shown, this embodiment provides a technical solution: A cognitive impairment evaluation system combining eye movement data, including the following modules: a task design module, a data acquisition module, a data preprocessing module, a feature extraction module, and an evaluation module;
[0035] The task design module is used to design the visual processing tasks that the subject needs to perform. Among them, the visual processing tasks include visual search tasks and text reading tasks;
[0036] More specifically, the purpose of the visual search task is to find the target object, which is used to evaluate the subject's attention and executive function. The purpose of the text reading task is to track the text for reading, which is used to evaluate the subject's language processing ability.
[0037] More specifically, the task design module includes a first task design unit, a second task design unit, and a task display unit; The first task design unit is used to design the visual search task and record the position information of the target object area on the display interface. The second task design unit is used to design the text reading task and record the position information of each text on the display interface. The task display unit is used to display the designed visual search task and text reading task on the display interface.
[0038] The data acquisition module is used to collect the eye movement data of the subject when performing a visual processing task by using an eye tracker;
[0039] More specifically, the data acquisition module includes a first eye movement data acquisition unit and a second eye movement data acquisition unit; the first eye movement data acquisition unit is used to collect the eye movement data of the subject when performing a visual search task, including the saccade speed Vs, the saccade direction Ds, the number of fixation points C1s, and the fixation point coordinates Ls C1 and the fixation duration Ts C1 , where s indicates that the corresponding data is the data when performing a visual search task; the second eye movement data acquisition unit is used to collect the eye movement data of the subject during a text reading task, including the number of fixation points C2w, the fixation point coordinates Lw C2 and the fixation duration Tw C2 , where w indicates that the corresponding data is the data when performing a text reading task.
[0040] The data preprocessing module is used to preprocess some of the collected eye movement data;
[0041] More specifically, the data preprocessing module includes a first preprocessing unit and a second preprocessing unit. The first preprocessing unit is used to preprocess the number of fixation points C1s, the fixation point coordinates Ls C1 and the fixation duration Ts C1 to obtain the preprocessed number of fixation points C1's, the fixation point coordinates Ls C1＇ and the fixation duration Ts C1＇ ; the second preprocessing unit is used to preprocess the number of fixation points C2w, the fixation point coordinates Lw C2 and the fixation duration Tw C2 to obtain the preprocessed number of fixation points C2'w, the fixation point coordinates Lw C2＇ and the fixation duration Tw C2＇ .
[0042] More specifically, the preprocessing method includes cleaning the collected eye movement data to remove outliers and noise data and improve the data quality.
[0043] The feature extraction module is used to process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network;
[0044] More specifically, the specific processing process of the feature extraction module is as follows:
[0045] S41: Obtain the preprocessed number of fixation points C1's, the fixation point coordinates Ls C1＇and the fixation duration Ts C1＇ , determine the number of fixation points Zs that fall within the target object area according to the fixation point coordinates Ls C1＇ and the position information of the target object area on the display interface, as well as the fixation duration Ts corresponding to each fixation point i , calculate the sum of the fixation duration Ts i , denoted as Ts i , where i represents the i-th fixation point that falls within the target object area, and the position information of the target object area on the display interface is the coordinates of each point on the contour line of the target object;
[0046] S42: Obtain the number of fixation points C2′w, the fixation point coordinates Lw C2＇ and the fixation duration Tw C2＇ , generate the fixation point text reading trajectory L1 according to the number of fixation points C2′w and the fixation point coordinates Lw C2＇ , and calculate the sum of the fixation duration Tw C2＇ , denoted as Tw t ; Generate the original text trajectory line L0 according to the position information of each text on the display interface; among them, the position information of each text on the display interface is the coordinates of the center point of the minimum circumscribed rectangle of each text;
[0047] S43: Calculate the similarity value between the text reading trajectory L1 and the original text trajectory line L0 based on the Euclidean distance, denoted as D L ;
[0048] S44: Normalize the number of fixation points Zs i during the execution of the visual search task, the sum Ts i of the fixation duration Ts t , the saccade speed Vs, the saccade direction Ds, and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 during the execution of the text reading task L , the sum Tw C2＇ of the fixation duration Tw t , denoted as Tw i , to obtain the normalized number of fixation points Zs i '' of the visual search task, the sum Ts t '' of the fixation duration Ts L , the saccade speed Vs'', the saccade direction Ds'', and the similarity value D C2＇ '' between the text reading trajectory L1 and the original text trajectory line L0 during the execution of the text reading task t , the sum Tw C1＇ '' of the fixation duration Tw, and use the above six types of normalized data as the input feature parameters of the BP neural network.
[0049] More specifically, in the step S41, when the fixation point coordinates Ls C1＇ are within the coordinate range of the target object area, it indicates that the fixation point falls within the target object area; otherwise, it indicates that the fixation point does not fall within the target object area.
[0050] More specifically, in the step S42, the text reading trajectory L1 is generated by sequentially connecting each fixation point, and the original text trajectory line L0 is generated by sequentially connecting the center points of each minimum bounding rectangle of the text.
[0051] More specifically, in the step S43, calculating the similarity value between the trajectory lines based on the Euclidean distance is to calculate the Euclidean distance between the corresponding points on the two trajectories, and then comprehensively measure the similarity of the trajectories by a feature corresponding method. Specifically as follows:
[0052] Assume two trajectory lines T1 = (p1, p2,..., p n ) and T2 = (q1, q2,..., q n ), where p i = (x i , y i ) and q i = (x′ i , y′ i ) are the points on the trajectories.
[0053] Calculation method: First calculate the Euclidean distance of the corresponding points Then the average value of these distances can be taken As the similarity measure of the two trajectories, the smaller the D value, the more similar the trajectories.
[0054] The evaluation module is used to evaluate the cognitive impairment of the subject by using the trained BP neural network to obtain the probability value of the existence of cognitive impairment of the current subject;
[0055] More specifically, the specific processing process of the evaluation module is as follows:
[0056] S51: Input the sum Ts i '' of the normalized number of fixation points Zs i '' and fixation point duration Ts t '', saccade speed Vs'', saccade direction Ds'', and the similarity value D L '' between the text reading trajectory L1 and the original text trajectory line L0 during the execution of the text reading task, as well as the sum Tw C2＇ '' of the fixation point duration Tw t '' into the input layer of the BP neural network;
[0057] S52: Process the six input feature parameters in step S51 through the hidden layer in sequence, and finally output the probability value of the existence of cognitive impairment of the current subject through the output layer.
[0058] It should be noted that in this embodiment, the BP neural network includes an input layer, a hidden layer, and an output layer. The input layer includes six neurons, which are respectively used to input the above six input feature parameters. The hidden layer has two layers, which are respectively used to process the input feature parameters when performing visual search tasks and text reading tasks; the output layer has one neuron, which is used to output the probability value of the existence of cognitive impairment of the current subject. The probability value ranges from 0 to 1, where 0 indicates that the current existence probability value is 0, that is, there is no cognitive impairment.
[0059] As Figure 2 shown, this embodiment also provides a cognitive impairment evaluation method combining eye movement data. Use the above-mentioned cognitive impairment evaluation system combining eye movement data to evaluate whether the subject has cognitive impairment, including the following steps:
[0060] S1: Design the visual processing tasks that the subject needs to perform;
[0061] S2: Use an eye tracker to collect the eye movement data of the subject when performing visual processing tasks;
[0062] S3: Preprocess some of the collected eye movement data;
[0063] S4: Process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network;
[0064] S5: Combine the input feature parameters in step S4, and use the trained BP neural network to evaluate the cognitive impairment of the subject to obtain the probability value of the existence of cognitive impairment of the current subject.
[0065] For the details of each step in the above evaluation method, refer to the above evaluation system, which will not be elaborated here.
[0066] In summary, the cognitive impairment evaluation system combining eye movement data in the above embodiment uses an eye tracker to collect the eye movement data of the subject when performing visual processing tasks, innovatively designs intermediate parameters such as the number of fixation points falling within the target object area and the similarity value of the trajectory line, which can well characterize the state of the subject when performing the corresponding tasks; it also uses the BP neural network to combine the input feature parameters of the components to evaluate the cognitive impairment of the subject, and can obtain the probability value of the existence of cognitive impairment of the current subject, realizing the more accurate evaluation of whether the subject has cognitive impairment by combining eye movement data.
[0067] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise specifically defined.
[0068] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.
[0069] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.
Claims
1. A cognitive impairment assessment system combined with eye movement data, characterized in that, It includes a task design module, a data acquisition module, a data preprocessing module, a feature extraction module, and an evaluation module; The task design module is used to design the visual processing tasks that the subject needs to perform; The data acquisition module is used to collect the eye movement data of the subject when performing visual processing tasks by using an eye tracker; The data preprocessing module is used to preprocess some of the collected eye movement data; The feature extraction module is used to process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network; The evaluation module is used to evaluate the cognitive impairment of the subject by using the trained BP neural network to obtain the probability value of the existence of cognitive impairment of the current subject.
2. The cognitive impairment assessment system incorporating eye movement data according to claim 1, wherein The task design module includes a first task design unit, a second task design unit, and a task display unit; the first task design unit is used to design the visual search task and record the position information of the target object area on the display interface, the second task design unit is used to design the text reading task and record the position information of each text on the display interface, and the task display unit is used to display the designed visual search task and text reading task on the display interface.
3. The cognitive impairment assessment system incorporating eye movement data according to claim 2, wherein The purpose of the visual search task is to find the target object and is used to evaluate the attention and executive function of the subject. The purpose of the text reading task is to track and read the text and is used to evaluate the language processing ability of the subject.
4. The cognitive impairment assessment system combined with eye movement data according to claim 2, wherein The data acquisition module includes a first eye movement data acquisition unit and a second eye movement data acquisition unit; the first eye movement data acquisition unit is used to acquire the eye movement data of the subject when performing a visual search task, including saccade speed Vs, saccade direction Ds, number of fixation points C1s, and fixation point coordinates Ls C1 and fixation point duration Ts C1 , where s indicates that the corresponding data is the data when performing a visual search task; the second eye movement data acquisition unit is used to acquire the eye movement data of the subject during a text reading task, including the number of fixation points C2w, fixation point coordinates Lw C2 and fixation point duration Tw C2 , where w indicates that the corresponding data is the data when performing a text reading task.
5. An evaluation system for cognitive impairment combining eye movement data according to claim 4, characterized in that The data preprocessing module includes a first preprocessing unit and a second preprocessing unit. The first preprocessing unit is used to preprocess the number of fixation points C1s, the fixation point coordinates Ls C1 and the fixation point duration Ts C1 to obtain the preprocessed number of fixation points C1's, the fixation point coordinates Ls C1＇ and the fixation point duration Ts C1＇ ; the second preprocessing unit is used to preprocess the number of fixation points C2w, the fixation point coordinates Lw C2 and the fixation point duration Tw C2 to obtain the preprocessed number of fixation points C2'w, the fixation point coordinates Lw C2＇ and the fixation point duration Tw C2＇ .
6. The cognitive impairment assessment system combined with eye movement data according to claim 5, wherein The specific processing process of the feature extraction module is as follows: S41: Obtain the number of preprocessed fixation points C1′s, the fixation point coordinates Ls C1＇ and the fixation point duration Ts C1＇ , and based on the fixation point coordinates Ls C1＇ and the position information of the target object area on the display interface, determine the number of fixation points Zs i falling within the target object area and the corresponding fixation point duration Ts of each fixation point i , calculate the sum of the fixation point durations Ts i , denoted as Ts t , where i represents the i-th fixation point falling within the target object area, and the position information of the target object area on the display interface is the coordinates of each point on the target object contour; S42: Obtain the number of fixation points C2′w, the fixation point coordinates Lw C2＇ and the fixation point duration Tw C2＇ , and generate a fixation point text reading trajectory L1 based on the number of fixation points C2′w and the fixation point coordinates Lw C2＇ , and calculate the sum of the fixation point durations Tw C2＇ , denoted as Tw t ; generate an original text trajectory line L0 according to the position information of each text in the display interface; where the position information of each text in the display interface is the center point coordinates of the minimum circumscribed rectangle of each text; S43: Calculate the similarity value between the text reading trajectory L1 and the original text trajectory line L0 based on the Euclidean distance, denoted as D L ; S44: The sum Zs of the number of fixation points when performing a visual search task i , the sum Ts of the fixation durations Ts i , the saccade speed Vs, the saccade direction Ds, and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 when performing a text reading task t , the sum Tw of the fixation durations Tw L , the sum Tw of the fixation durations Tw C2＇ , the sum Tw of the fixation durations Tw t are normalized to obtain the normalized number Zs of fixation points when performing a visual search task i '', the sum Ts of the fixation durations Ts i '', the saccade speed Vs '', the saccade direction Ds '', and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 when performing a text reading task t '', the sum Tw of the fixation durations Tw L '', the sum Tw of the fixation durations Tw C2＇ '', the sum Tw of the fixation durations Tw t '', and the above six types of normalized data are used as the input feature parameters of the BP neural network.
7. The cognitive impairment assessment system incorporating eye movement data according to claim 6, characterized in that, The specific processing process of the evaluation module is as follows: S51: Input the sum Ts of the number of fixation points Zs, the fixation duration Ts, the saccade speed Vs'', the saccade direction Ds'', and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 during the execution of the text reading task, and the sum Tw of the fixation duration Tw during the execution of the visual search task after normalization into the input layer of the BP neural network; i '', the fixation duration Ts i of the sum Ts t '', the saccade speed Vs'', the saccade direction Ds'', and the similarity value D between the text reading trajectory L1 and the original text trajectory line L0 during the execution of the text reading task L '', the fixation duration Tw C2＇ of the sum Tw t into the input layer of the BP neural network; S52: Process the six input feature parameters in step S51 through the hidden layer in sequence, and finally output the probability value of the existence of cognitive impairment of the current subject through the output layer.
8. A cognitive impairment assessment method combined with eye movement data, characterized in that, Using the cognitive impairment evaluation system combined with eye movement data as described in claim 7 to evaluate whether the subject has cognitive impairment, including the following steps: S1: Design the visual processing tasks that the subject needs to perform; S2: Collect the eye movement data of the subject when performing visual processing tasks by using an eye tracker; S3: Preprocess some of the collected eye movement data; S4: Process the preprocessed part of the eye movement data and the remaining original eye movement data to obtain the input feature parameters of the BP neural network; S5: Combine the input feature parameters in step S4 and use the trained BP neural network to evaluate the cognitive impairment of the subject to obtain the probability value of the existence of cognitive impairment of the current subject.