Cognitive ability assessment method and system based on eye movement tracking and program product

Abstract

The invention relates to the field of cognitive assessment, in particular to a cognitive ability assessment method and system based on eye movement tracking and a program product. The method comprises the following steps that a testee is tested through an interactive electronic game, dominant behavior characteristics in the test are collected, and the dominant behavior characteristics comprise task completion time and accuracy; collecting a time sequence of eye movement data of the testee in the testing process; collecting the time of a key event of the testee in the testing process to obtain a time sequence of the key event, and performing time alignment and merging on the time sequence of the key event and the time sequence of the eye movement data to obtain a time sequence of a recessive behavior; and inputting the recessive behavior time sequence and the dominant behavior characteristics into a pre-trained machine learning model to obtain a comprehensive cognitive score. According to the method, on the premise that an existing interaction normal form is maintained, implicit multi-dimensional cognitive features are mined from limited behavior data streams, and the evaluation precision is improved.

Description

Technical Field

[0001] The present invention relates to the field of cognitive assessment, and in particular to a cognitive ability assessment method and system based on eye tracking, and a program product. Background Art

[0002] Clinical studies have shown that age-related cognitive decline is usually manifested as multi-dimensional cognitive impairment, including but not limited to core symptoms such as memory loss, executive function decline, visual spatial orientation impairment, impaired language fluency, abstract thinking degeneration, and learning and calculation ability decline. At present, clinical treatment is mainly based on drug treatment. Existing drugs can improve the cognitive function of the elderly to a certain extent in the early stage of the disease, but can only slightly improve the quality of life of patients and cannot prevent the decline of cognitive ability in the elderly.

[0003] In order to evaluate and train the cognitive ability of the elderly, manual testing and question-and-answer communication are usually used, or classic cognitive scales are used as evaluation tools. Although they have been verified for many years in terms of reliability and validity, they still have some shortcomings, including: some scale items rely on the subjective scoring of the evaluator, and stability and objectivity cannot be guaranteed; it is time-consuming and lengthy, making it difficult for the subjects to persist, and the time cost is high; the entire process needs to be conducted one-on-one by the evaluator, which has high labor costs.

[0004] In recent years, with the development of digital technology, cognitive assessment systems based on interactive electronic games have gradually replaced traditional paper-based assessment methods. This type of system uses preset cognitive tasks (such as memory matching, pattern recognition, reaction speed tests, etc.), which have significant advantages over subjective questionnaire assessment methods: 1) Standardized processes ensure the consistency of assessment conditions and avoid subjective biases in manual observations; 2) Digital interfaces effectively improve the participation of test subjects. Typical applications include the spatial working memory module in the Cambridge Brain Science Assessment Kit (CANTAB) and the pattern comparison test in the US NIH toolbox.

[0005] However, the existing system has significant technical bottlenecks in engineering implementation. Limited by the cognitive load theory of human-computer interaction, developers are forced to make a trade-off between task complexity and data collection dimensions: on the one hand, in order to maintain the subject's attention threshold (usually controlled at 15-20 minutes), the game mechanism must be simplified and the task level must be compressed; on the other hand, the current technical framework can only capture explicit behavioral characteristics (such as basic indicators such as task completion time and click accuracy), resulting in low efficiency in extracting multi-dimensional cognitive features.

[0006] The core of this technological paradox lies in that improving the evaluation accuracy requires increasing the data collection dimension. However, the dimension expansion will inevitably lead to an increase in interaction complexity, thus disrupting the continuity of the user experience. Existing solutions attempt to make compromises through segmented testing or modular design, but they have triggered new technical problems - it is difficult to ensure the data consistency across tasks. Therefore, there is an urgent need in this field to develop new feature extraction algorithms that can mine implicit multi-dimensional cognitive features from limited behavioral data streams while maintaining the existing interaction paradigm. Summary of the Invention

[0007] In view of the above problems, the present invention proposes the following technical solutions:

[0008] A method for evaluating cognitive ability based on eye movement tracking, comprising the following steps:

[0009] Testing the subject through an interactive electronic game, and collecting the explicit behavioral features during the test, where the explicit behavioral features include task completion time and correct rate;

[0010] Collecting the time series of eye movement data of the subject during the test;

[0011] Collecting the time of key events of the subject during the test to obtain the time series of key events, and performing time alignment and merging on the time series of key events and the time series of eye movement data, so as to obtain the implicit behavior time series;

[0012] Inputting the implicit behavior time series and the explicit behavioral features into a pre-trained machine learning model to obtain a comprehensive cognitive score.

[0013] Preferably, the eye movement data is collected by an infrared camera.

[0014] Preferably, the content of the key events includes test phase switching, appearance of interference items, appearance of numbers, and task completion.

[0015] Preferably, the eye movement data includes fixation point coordinates, pupil diameter, and eye opening and closing state.

[0016] Preferably, the machine learning model includes random forest and support vector machine.

[0017] The present invention also provides a cognitive ability evaluation system based on eye movement tracking,

[0018] including:

[0019] A memory for storing a computer program;

[0020] A processor for implementing the above method steps when executing the program stored in the memory.

[0021] The present invention also provides a computer program product, including a computer program which, when executed by a processor, implements the above method.

[0022] Beneficial effects: By combining interactive electronic games with eye movement tracking technology, an objective, efficient, and refined assessment of the cognitive function of the elderly is achieved, solving problems such as subjective bias, long time consumption, and high labor costs existing in traditional scale assessments. On the premise of maintaining the existing interaction paradigm, implicit multi-dimensional cognitive features are mined from limited behavioral data streams, improving the evaluation accuracy. Specific embodiments

[0023] To make the objectives, features, and advantages of the present invention more obvious and understandable, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0024] A cognitive ability assessment method based on eye movement tracking includes the following steps:

[0025] Test the subject through an interactive electronic game, and collect the explicit behavior characteristics during the test. The explicit behavior characteristics include task completion time and correct rate.

[0026] Collect the time series of eye movement data of the subject during the test through an infrared camera. The eye movement data includes fixation point coordinates (the horizontal and vertical coordinates of the screen position, and the time series of the screen position coordinates implies various information such as fixation duration and saccade path), pupil diameter (the change in pupil diameter reflects cognitive load), and eye opening and closing state (the blink rate reflects fatigue degree). Before the infrared camera collects the eye movement data of the subject, a personalized eye movement model is established through a 9-point calibration program (the subject sequentially fixates on 9 marked points on the screen), and the calibration error ≤ 0.5°.

[0027] Collect the time of key events during the test of the subject to obtain the time series of key events. The content of the key events includes test stage switching, appearance of interference items, appearance of numbers, task completion, etc. Align and merge the time series of key events with the time series of eye movement data to obtain the implicit behavior time series.

[0028] Input the implicit behavior time series and the explicit behavior characteristics into a pre-trained machine learning model to obtain a comprehensive cognitive score (on a scale of 0 - 100), and divide it into levels of normal (≥ 80), mild impairment (60 - 79), and moderate impairment (< 60). The machine learning model includes random forest and support vector machine.

[0029] In this embodiment, the interactive electronic game includes:

[0030] 1. Implementation method of bradykinesia test (nine - grid response speed test):

[0031] Present a 3×3 nine - grid interface on the touch screen. The system randomly lights up a number (from 1 to 9) in a certain grid, and after the number is displayed for 0.5 seconds, it automatically hides;

[0032] It is required that the subject immediately clicks on the corresponding grid to "turn off" the number after the number appears;

[0033] Record the first reaction time (the time difference from the appearance of the number to the first click), click accuracy rate (the number of correct clicks / the total number of tests), and false touch rate (the number of clicks on the blank grid);

[0034] Dynamically adjust the test difficulty: According to the performance of the subject, gradually shorten the number display duration (from 0.5 seconds to 0.2 seconds) or increase the number of simultaneously lit numbers (from single number to double numbers).

[0035] 2. Implementation method of memory decline test (dynamic digital matrix restoration test):

[0036] Display a 5×5 digital matrix in the center of the screen, where 3 - 5 numbers randomly flash brightly (duration 200ms), and the remaining numbers are gray;

[0037] After the numbers stop flashing, the matrix becomes blank, and the subject needs to click to restore the positions of the brightly lit numbers within 5 seconds based on memory;

[0038] Record the short - term memory accuracy rate (the number of correctly restored positions / the total number of brightly lit numbers), long - term memory decline rate (the percentage decrease in accuracy rate when tested again after a 60 - second delay);

[0039] Add interference items: Insert interference animations (such as flashing borders) in some tests to evaluate the anti - interference memory ability.

[0040] 3. Implementation method of spatial imagination test (three - dimensional hexahedron color reasoning test):

[0041] Show a three - dimensional rotating hexahedron model, with each of the six faces assigned high - contrast colors (red, blue, green, yellow, purple, orange), and the colors of adjacent faces are significantly different;

[0042] Hide a specified face of the hexahedron, and require the subject to select the color of the hidden face from a color palette;

[0043] Record the spatial reasoning correct rate (the number of correct selections / the total number of tests), response hesitation time (the time from the presentation of the question to the selection operation);

[0044] Dynamic rotation control: The hexahedron automatically rotates at a speed of 2 seconds per week, and the subject can manually pause the rotation to observe the structure.

[0045] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A cognitive ability assessment method based on eye movement tracking, characterized in that, It includes the following steps: Testing the subject through an interactive electronic game and collecting the overt behavioral characteristics during the test, where the overt behavioral characteristics include task completion time and accuracy rate; Collecting the time series of the subject's eye movement data during the test; Collecting the time of key events during the test of the subject to obtain the time series of key events, and performing time alignment and merging on the time series of key events and the time series of eye movement data to obtain the covert behavior time series; Inputting the covert behavior time series and the overt behavioral characteristics into a pre-trained machine learning model to obtain a comprehensive cognitive score.

2. The cognitive ability evaluation method based on eye movement tracking according to claim 1, wherein The eye movement data is collected by an infrared camera.

3. The cognitive ability assessment method based on eye movement tracking according to claim 1, characterized in that, The content of the key events includes test phase switching, appearance of interference items, appearance of numbers, and task completion.

4. The cognitive ability assessment method based on eye movement tracking according to claim 1, wherein The eye movement data includes fixation point coordinates, pupil diameter, and eye opening and closing state.

5. A method for evaluating cognitive ability based on eye movement tracking according to claim 1, characterized in that The machine learning model includes random forest and support vector machine.

6. A cognitive ability evaluation system based on eye movement tracking, characterized in that it includes: A memory for storing a computer program; A processor for implementing the method steps described in any one of claims 1-5 when executing the program stored on the memory.

7. A computer program product, comprising a computer program, characterized in that, The computer program, when executed by the processor, implements the method described in any one of claims 1 to 5.