Eye movement normal form and index evaluation method and system for cognitive disorder detection

Abstract

The invention provides an eye movement normal form and index evaluation method and system for cognitive disorder detection, and relates to the technical field of eye movement data analys.The eye movement normal form and index evaluation method comprises the steps that eye movement parameters of a subject under eye movement normal forms graded at different difficulties are obtained, and the eye movement normal forms are evaluated on the basis of the eye movement parameters under the eye movement normal forms graded at different difficulties; extracting eye movement features under different difficulty grading eye movement paradigms; the cognitive competence of the participant is evaluated, and an MoCA score is obtained; a statistical analysis method is adopted to evaluate the normality of eye movement parameters, correlation analysis is carried out on eye movement characteristics under MoCA scores and different difficulty grading eye movement normal forms, a Spearman correlation analysis method is utilized to calculate the correlation coefficient and significance between the MoCA scores and the eye movement characteristics, rank correlation graphs are drawn respectively, and the eye movement parameters are obtained. And obtaining eye movement indexes with significant differences among the participants with different degrees of cognitive impairment under each normal form and the feasibility of the eye movement indexes in cognitive competence evaluation.

Description

Technical Field

[0001] The present disclosure relates to the technical field of eye movement data analysis, and in particular to an eye movement paradigm and an indicator evaluation method and system for cognitive impairment detection. Background Art

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] Dementia is a disease characterized by cognitive deterioration, which seriously affects patients' learning and memory, language, executive function, attention, social cognition, etc. Mild cognitive impairment, as an early manifestation of dementia, is the best window for intervention.

[0004] Existing diagnostic methods for cognitive impairment mainly include neuropsychological scale assessment, biomarker detection (blood, cerebrospinal fluid) and imaging examinations. Early diagnostic technologies such as magnetic resonance imaging (MRI), positron emission tomography (PET) and biomarkers are not widely used. The diagnosis rate of early dementia in general hospital outpatient clinics is only 0.1%, and early diagnosis of dementia also faces the challenge of standardization. In order to solve the above problems, it is urgent to find new biomarkers to achieve non-invasive detection of cognitive impairment. Eye tracking technology, as a non-invasive, sensitive and objective detection method, has received increasing attention in the study of cognitive impairment in the elderly. In contrast, eye tracking technology has attracted widespread attention from researchers due to its non-invasive, sensitive and objective characteristics. However, the eye movement paradigm in current research is relatively single, and it cannot effectively conduct multi-dimensional and fine-grained analysis of patients with cognitive impairment.

[0005] In the field of cognitive impairment analysis based on eye movement data, many existing works have achieved certain research results, including exploring the differences in eye movement behavior between Alzheimer's disease (AD) and healthy people in visual search through clock reading tasks, recording the participants' eye movement interest areas, fixation indicators, etc., and calculating the proportion of correct reading. The study found that patients with cognitive impairment paid less attention to the end of the clock pointer, had a short saccade distance, and had a longer time delay in the first fixation on the pointer end area. And through the visual spatial memory task, the subjects were asked to watch a set of images, and then an object was deleted or added to this set of images to evaluate the subjects' memory ability for the changed objects and their positions, and the subjects' memory ability was evaluated by measuring the number of fixations, fixation duration, and fixation time ratio of the subjects to the changed area, and logistic regression was used to identify patients with mild cognitive impairment. It also includes a 3D comprehensive visual task, in which the subjects were asked to watch 9 3D images to collect heat maps of the visual fixation point stay position and time, and use a multi-layer comparative convolutional network (MC-CNN) to analyze the differences in visual attention between AD patients and healthy people, so as to achieve auxiliary diagnosis of AD patients. The above studies show that changes in eye movement parameters can effectively reflect the impact of the disease on the patient's cognitive ability.

[0006] Although existing studies have demonstrated the feasibility of identifying cognitive impairment based on eye movements, these studies have mainly focused on visual search tasks and have not studied in depth the changes in saccade characteristics in patients with cognitive impairment. In studies on the identification of cognitive impairment based on saccade characteristics, it was found that the reverse saccade task can effectively distinguish between memory MCI and non-memory MCI, among which patients with memory MCI (aMCI) were significantly impaired in inhibitory control, while patients with non-memory MCI (naMCI) performed closer to healthy controls. Chehrehnegar et al. used eye tracking technology to explore the differences in eye movements between patients with prodromal Alzheimer's disease (aMCI) and healthy controls. The results showed that the aMCI group showed more saccade errors, saccade omissions, and uncorrected saccades in the reverse saccade task. Studies have shown that patients with cognitive impairment have longer saccade latencies and lower accuracy in saccade tasks, and their error correction rates in reverse saccade tasks are significantly reduced. However, the scenario design of the above-mentioned cognitive impairment detection research based on eye movement characteristics is relatively simple, and only horizontal forward eye movement or horizontal reverse eye movement tasks are used to classify patients with cognitive impairment and elderly healthy controls. There is a lack of more fine-grained paradigm design to deeply analyze the changes in eye movement parameters of patients with cognitive impairment. Summary of the invention

[0007] In order to solve the above problems, the present disclosure proposes an eye movement paradigm and an index evaluation method and system for cognitive impairment detection. Through multiple difficulty-graded eye movement paradigms, the effect of task difficulty on the patient's eye movement ability is verified, and eye movement parameters with higher sensitivity for cognitive impairment analysis are explored; a large-scale cognitive impairment eye movement data set is established, and through Mann-Whitney U test analysis, eye movement indicators with significant differences between patients with different degrees of cognitive impairment under each paradigm are obtained, and further correlation analysis is performed between these indicators and Montreal Cognitive Assessment (MoCA) scores to verify the feasibility of these eye movement indicators in cognitive ability assessment.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: Eye movement paradigms and indicator assessment methods for cognitive impairment detection, including: Design eye movement paradigms with multiple difficulty levels; Obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; The participants' cognitive abilities were assessed and MoCA scores were obtained; Statistical analysis methods were used to evaluate the normality of eye movement parameters, analyze the differences in eye movement of participants at different task difficulty levels and eye movement directions, and conduct correlation analysis between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms. Spearman correlation analysis method was used to calculate the correlation coefficient and significance between the two, and rank correlation diagrams were drawn to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm. Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: Eye movement paradigm and indicator evaluation system for cognitive impairment detection, including: Task design module, used to design eye movement paradigms with multiple difficulty levels; A data acquisition module is used to obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; The evaluation and analysis module is used to grade the cognitive abilities of the participants and obtain MoCA scores. The normality of eye movement parameters is evaluated by statistical analysis methods, and the differences in eye movement of participants at different task difficulty levels and eye movement directions are analyzed. The correlation between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms is analyzed. The Spearman correlation analysis method is used to calculate the correlation coefficient and significance between the two, and rank correlation diagrams are drawn to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm. Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

[0010] Compared with the prior art, the present invention has the following beneficial effects: The eye movement paradigm and index evaluation method for cognitive impairment detection disclosed in the present invention can discover more eye movement indicators that can be used for cognitive impairment evaluation, and conduct fine-grained analysis of the cognitive level of the subject from multiple dimensions. The random forward saccade task has high sensitivity in cognitive impairment screening, especially in vertical tasks, and has significant discriminatory power, providing an important reference for early screening and evaluation.

[0011] The eye movement paradigm and index evaluation method for cognitive impairment detection disclosed in the present invention designs a series of eye movement tasks with graded difficulty based on eye movement data for cognitive impairment analysis. The eye movement tasks include forward eye movement, reverse eye movement and random forward-reverse eye movement, which are used to analyze the impact of task difficulty on the eye movement parameters of participants, further prove the feasibility of using eye movement parameters for cognitive impairment evaluation, and provide a basis for the design of fine-grained eye movement paradigm.

[0012] The eye movement paradigm and index evaluation method for cognitive impairment detection disclosed in the present invention constructs a large-scale eye movement data set to verify the relationship between eye movement indicators and the degree of cognitive impairment. The data set includes 36 young healthy controls, 15 elderly healthy controls, 23 patients with mild cognitive impairment, and 16 patients with moderate to severe dementia. This data set significantly expands the number of test samples in the study of elderly cognitive impairment assessment based on eye movement indicators, and uses statistical analysis to determine the key eye movement features that can effectively assist in the diagnosis of cognitive impairment. By analyzing the differences in eye movement indicators between different experimental groups and the correlation between eye movement indicators and MoCA scores, the effect of task difficulty on eye movement indicators was verified, and eye movement indicators with high sensitivity to cognitive impairment under different eye movement paradigms were obtained.

[0013] The eye movement paradigm and index evaluation method for cognitive impairment detection disclosed in the present invention are sensitive to task difficulty in distinguishing early cognitive impairment from healthy elderly controls in terms of the correct AOI dwell time and the first time to the correct AOI, and the correlation with the MoCA score is significantly enhanced in the random reverse saccade task. The saccade latency has the highest correlation with the MoCA score in the random reverse task in the vertical direction, indicating that the directional factor of the saccade has a significant impact on this indicator. The first fixation time and the fixation duration are significantly correlated with the MoCA score in all tasks. These results show that complex saccade tasks can effectively reflect the degree of cognitive impairment and provide an important reference for the early screening of MCI and the assessment of cognitive impairment. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The accompanying drawings constituting a part of the present disclosure are used to provide a further understanding of the present disclosure. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure.

[0015] Figure 1 Schematic diagram of a positive saccade paradigm according to an embodiment of the present disclosure; Figure 2 Schematic diagram of the reverse saccade paradigm of the embodiment of the present disclosure; Figure 3 A schematic diagram of a random forward-reverse saccade paradigm according to an embodiment of the present disclosure; Figure 4 It is a correlation diagram between MOCA and eye movement features of the embodiment of the present disclosure; in, Figure 4 (a) shows the correlation analysis between MoCA score and correct AOI dwell time under different paradigms; Figure 4 (b) is the correlation analysis between MoCA score and saccade latency; Figure 4 (c) is the correlation analysis between MoCA score and first fixation time under different paradigms; Figure 4 (d) is the correlation analysis between MoCA score and peak saccade velocity in different task paradigms; Figure 4 (e) shows the correlation analysis between MoCA score and the time to first reach the correct AOI; Figure 4 (f) shows the correlation analysis between MoCA scores and fixation duration under different paradigms. DETAILED DESCRIPTION

[0016] The present disclosure is further described below in conjunction with the accompanying drawings and embodiments.

[0017] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further explanation of the present disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present disclosure belongs.

[0018] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present disclosure. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0019] Example 1 In one embodiment of the present disclosure, an eye movement paradigm and an index evaluation method for cognitive impairment detection are provided, including: Step 1: Design an eye movement paradigm with multiple difficulty levels; Step 2: Obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; Step 3: Conduct a graded assessment of the participants’ cognitive abilities to obtain MoCA scores; Step 4: Use statistical analysis methods to evaluate the normality of eye movement parameters, analyze the eye movement differences of participants in different task difficulty levels and eye movement directions, conduct correlation analysis between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms, use Spearman correlation analysis method to calculate the correlation coefficient and significance between the two, and draw rank correlation diagrams respectively, to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm; Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

[0020] As an embodiment, the specific implementation process of the eye movement paradigm and index evaluation method for cognitive impairment detection disclosed in the present invention is as follows: Step 1: Design an eye-tracking paradigm with multiple difficulty levels Specifically, three difficulty levels of saccade task paradigms were designed, including forward saccades, reverse saccades, and random forward-reverse saccades. Each paradigm included tasks in the horizontal and vertical directions.

[0021] Step 2: Obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features of the eye movement parameters under the eye movement paradigms of different difficulty levels; Specifically, 1) Recruitment of participants: Obtain the MCI diagnostic data of the patients, and record the scores of MMSE, MoCA, Hamilton Anxiety Scale (HAMA), and Hamilton Depression Scale (HAMD). To exclude patients with anxiety and / or depression, the HAMD and HAMA scores of the participants included in the study were less than 7. A total of 23 patients with mild cognitive impairment (MCI, 20≤MMSE<24 points, 18≤MoCA<25 points), 16 patients with moderate to severe dementia (MMSE<20 points, MoCA<18 points) were included in the study, and 15 middle-aged and elderly healthy controls (age ≥50) and 36 young healthy controls (age ≤30) were matched according to age and education. All participants were recruited by the Advanced Cognitive Impairment Diagnosis and Treatment Center and the Health Examination Center of the Affiliated Hospital of Shandong University of Traditional Chinese Medicine, and their chief complaints, current medical history, past medical history, neurological examination, and neuropsychological assessment were recorded. The healthy control group was checked by the health examination center to ensure normal physical condition. The age, gender, education level and other information of the subjects are shown in Table 1. After informing the subjects of the experimental procedures and operation requirements in detail, the study obtained written informed consent from all patients and volunteers. All participants passed routine examinations and laboratory tests to ensure that each subject had normal vision or corrected to normal levels, normal hearing, and was able to fully understand the experimental scenarios and task requirements. Participants completed the eye movement tasks under the guidance of professional cognitive assessors.

[0022] Table 1 Statistics of participants’ information

[0023] 2) Task implementation and eye movement data collection. This paper uses the Tobii Pro Fusion 250 eye tracker to record the gaze point coordinates and pupil diameter data of the subjects' eyes during the experiment. Two screens were configured during the data collection process: a 24-inch display with a resolution of 1920×1080 pixels was used to present the eye saccade task, and the eye tracker was fixed below the screen; the other was an operating screen to monitor the subject's calibration status and concentration during the experiment in real time. Eye movement data analysis was performed using Tobii Pro Lab software. Before the task began, all participants underwent a five-point calibration and verification to ensure that the coverage rate of eye movement data collected during the task reached more than 90%. The specific task implementation and data collection process are as follows: (1) The subjects were asked to sit on a height-adjustable chair and underwent a five-point calibration before the eye movement test. A detailed task introduction was given before the start of each eye movement task paradigm. The present invention designed eye movement task paradigms with three difficulty levels, including forward saccades, reverse saccades, and random forward-reverse saccades.

[0024] (2) The forward saccade paradigm consists of six forward saccade tasks. At the beginning of each task, the subject is asked to look at the blue ball displayed in the center of the screen. Then, the ball disappears and appears in the peripheral visual area of ​​the screen as a target stimulus. The subject is asked to shift his gaze to the target stimulus position as quickly as possible. Figure 1 The ball was displayed for 2000 milliseconds at each position, and the time interval between two consecutive forward saccades was 1000 milliseconds.

[0025] (3) The antisaccade paradigm includes six antisaccade tasks. At the beginning of each task, the subject is asked to look at the red ball displayed in the center of the screen. Then, the ball disappears and appears in the peripheral visual area of ​​the screen as a target stimulus. At this time, the subject is required to suppress the spontaneous response and turn his eyes to the opposite direction of the target stimulus as quickly and accurately as possible. The red ball is displayed at each position for 2000ms. The antisaccade paradigm is as follows: Figure 2 The correct antisaccade process includes two key steps: one is to inhibit the reflexive gaze response to the suddenly appearing target, and the other is to actively shift the gaze to the relative position of the target.

[0026] (4) In the random forward-reverse saccade paradigm, there are 10 saccade tasks in total. The type of task each time is indicated by a ball of different colors: when the ball in the center of the screen is blue, the participant needs to perform a forward saccade task; when the ball is red, the participant needs to perform a reverse saccade task. The blue and red balls appear five times each, but the color and direction of the balls are randomly presented in each task. Figure 3 shown.

[0027] (5) Each difficulty level paradigm includes both horizontal and vertical saccade tasks. In the horizontal saccade task, the ball disappears from the center point and then randomly appears on the left or right side of the screen; in the vertical saccade task, the ball randomly appears on the top or bottom of the screen.

[0028] 3) By analyzing the eye movement parameters of subjects in tasks of different difficulty and direction, and verifying the eye movement differences of patients with cognitive impairment in different task difficulty levels and eye saccade directions, we can better achieve auxiliary diagnosis of the severity of cognitive impairment based on eye saccades.

[0029] In the above paradigm design, from the perspective of the difficulty of the saccade task: random reverse saccade task > reverse saccade task > random forward saccade task > forward saccade task, and the comprehensive difficulty of the paradigm: reverse saccade paradigm > random forward-reverse saccade paradigm > forward saccade paradigm. In the subsequent analysis, the eye movement characteristics of the two tasks, random forward saccade and random reverse saccade, will be calculated and analyzed separately for the eye movement parameters related to the task. As for the eye movement indicators related only to the paradigm (fixation duration and first fixation time), the comprehensive discussion of random forward-reverse shows that the comprehensive difficulty of its saccade execution is lower than that of the reverse saccade paradigm and higher than that of the forward saccade paradigm.

[0030] 4) Extract eye movement features of eye movement parameters under eye movement paradigms of different difficulty levels.

[0031] Specifically, the present disclosure uses Tobii Pro Lab software to calculate and count the eye movement features of the eye movement data in the saccade and antisaccade tasks. The calculated eye movement features include saccade accuracy (SA), saccade correction rate (SCR), time to first enter the correct area (AOI) (TFCA), time to stay in the correct area (AOI) (DTCA), saccade latency (SL), saccade peak velocity (PV), fixation duration (TDF) and first fixation time (TFF).

[0032] Among them, saccade refers to the process of moving between fixation points. The saccade accuracy rate is the ratio of the number of times the first saccade after the stimulus occurs to the correct AOI in the total number of tasks. The saccade error correction rate indicates the proportion of incorrect saccades that are corrected. The time to enter the correct AOI for the first time refers to the time interval from the appearance of the stimulus to the entry of the fixation point into the correct AOI. The residence time in the correct AOI area refers to the total residence time of the fixation point in the correct AOI. The saccade latency is the time interval between the appearance of the stimulus and the beginning of the correct saccade response of the eyes. The peak saccade velocity is the highest velocity during the saccade process. In addition, the first fixation time and fixation duration in the center area of ​​the screen at the beginning of each paradigm were calculated. The first fixation time refers to the time from the beginning of the paradigm to the occurrence of the first fixation, and the fixation duration refers to the total fixation time in the center area of ​​the screen within 2000ms after the beginning of the paradigm.

[0033] Step 3: Score the participants’ cognitive abilities and construct a multi-difficulty level eye movement paradigm and eye movement index evaluation method for cognitive impairment detection; The present disclosure uses a set of comprehensive neuropsychological tests to grade the cognitive abilities of participants, and the test duration is about 90 minutes. It mainly includes the Activities of Daily Living (ADL) scale, the Montreal Cognitive Assessment (MoCA) scale and the Mini-mental State Examination (MMSE) scale. The ADL scale is used to assess the self-care ability of the elderly or people with chronic diseases and disabilities in daily life. The scale contains six assessment indicators, namely eating, bathing, dressing, going to the toilet, moving and defecation control. The MoCA scale is used to quickly screen for cognitive dysfunction, and its assessment criteria cover eight cognitive areas and eleven examination items, including attention and concentration, executive function, memory, language, visual structure skills, abstract thinking, calculation ability and orientation and other cognitive function areas. The MoCA scale also includes scoring guidance for specific projects, such as alternating connection test, visual structure skill test (drawing cubes and clocks), naming and memory, etc. The MMSE scale is used to assess the individual's intellectual state and the degree of cognitive function defects. This scale mainly evaluates the cognitive functions of the subjects in terms of time orientation, place orientation, immediate memory, attention and calculation, delayed memory, language ability, visual-spatial ability, etc. Doctors and cognitive assessors conduct a comprehensive assessment of the patient's degree of cognitive impairment based on the scores of the above scales, combined with the patient's medical history, outpatient records and various examination results.

[0034] Step 4: Statistical analysis methods were used to evaluate the normality of eye movement parameters, analyze the eye movement differences of participants at different task difficulty levels and eye movement directions, and conduct correlation analysis on the eye movement characteristics of MoCA scores and eye movement parameters under different difficulty level eye movement paradigms. The Spearman correlation analysis method was used to calculate the correlation coefficient and significance between the two, and rank correlation diagrams were drawn respectively to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm and the feasibility of eye movement indicators in cognitive ability assessment.

[0035] Specifically, all statistical analyses were performed using SPSS 26.0. First, the normality of demographic data and eye movement parameters was assessed using the Kolmogorov-Smirnov test. The Kolmogorov-Smirnov test was performed by calculating the empirical cumulative distribution function of the sample data. Theoretical normal cumulative distribution function The maximum absolute difference between To measure the deviation between the two, and to determine whether the data conforms to the normal distribution. Subsequently, the Mann-Whitney U test was used to analyze the differences in eye movement characteristics between the young healthy control group and the elderly healthy control group, the elderly healthy control group and the mild cognitive impairment group, and the mild cognitive impairment group and the moderate to severe dementia group. The p value of the Mann-Whitney U test was recorded for each analysis, and the rank sum of the two groups of samples was calculated to evaluate whether there was a significant difference in their distribution. Assume that the two groups of samples are and , the sample sizes are m and n respectively. First, merge and sort the two groups of samples, and calculate the rank of each sample value, and find the sum of the ranks of the two groups and Then, according to the formula and Calculate the U statistic for each group and take the smaller value As a test statistic. Finally, the corresponding p-value is calculated by looking up a table or using statistical software to determine whether there is a significant difference between the two groups of data.

[0036] In addition, correlation analysis was performed on MoCA scores and eye movement characteristics in various task scenarios with different difficulty levels. The Spearman correlation analysis method was used to calculate the correlation coefficient and significance between the two, and rank correlation diagrams were drawn. Table 2 records the number, age, gender, and MoCA scores of the young healthy control group (YC), the middle-aged and elderly healthy control group (EC), the mild cognitive impairment patients (MCI), and the moderate to severe dementia patients (Dementia), and the differences in demographics and neuropsychological test scores were analyzed for the three elderly groups (EC / MCI / Dementia). There was no statistically significant difference in age, education level, and gender among the three elderly groups of EC, MCI, and Dementia (p>0.05).

[0037] Table 2. Information of the participants

[0038] Furthermore, the difference analysis of eye movement parameters of each control group in the horizontal forward saccade paradigm is shown in Table 3. The elderly healthy control group and the young healthy control group showed significant differences in multiple eye movement parameters, including the time to the correct AOI for the first time (p=0.024), the correct AOI dwell time (p=0.004), the saccade latency (p=0.024), the fixation duration (p=0.004), and the first fixation time (p<0.001). The eye movement ability of the healthy elderly group declined significantly, indicating that the attention and reaction speed of the subjects were significantly weakened with age. The peak saccade velocity of patients with mild cognitive impairment was significantly lower than that of the elderly healthy control group (p=0.018), indicating that mild cognitive impairment had a negative impact on saccade velocity. However, the eye movement characteristics between patients with mild cognitive impairment and patients with moderate to severe dementia did not show significant differences.

[0039] Table 3. Analysis of differences in eye movement parameters among control groups in the horizontal forward saccade paradigm

[0040] Furthermore, in the vertical forward saccade paradigm, the differences in eye movement characteristics of each control group are shown in Table 4. There were also significant differences between the young healthy control group and the elderly healthy control group in the time to the correct AOI for the first time (p=0.004), the time to stay in the correct AOI (p=0.003), and the latency of saccades (p=0.004). In addition, there were significant differences between the elderly healthy control group and patients with mild cognitive impairment in the time to the first fixation (p=0.033), but there was no significant difference in the eye movement characteristics between patients with mild cognitive impairment and patients with moderate to severe dementia.

[0041] Table 4. Analysis of differences in eye movement parameters among the control groups in the vertical forward saccade paradigm

[0042] Although there were significant differences in eye movement characteristics between the young healthy control group and the elderly healthy control group in the forward saccade task, there were fewer or even no eye movement characteristics with significant differences between the other control groups. This suggests that the forward saccade task is not powerful enough in the analysis of cognitive impairment and cannot effectively assess the severity of cognitive impairment, and a more sensitive test method needs to be developed.

[0043] Furthermore, in the random forward-reverse saccade paradigm, the forward saccade task and the reverse saccade task were analyzed separately. The difference analysis of eye movement characteristics of each control group in the random level forward saccade task is shown in Table 5. As in the original forward saccade task, there were significant differences between the young healthy control group and the elderly healthy control group in multiple eye movement indicators, including the time to the first correct AOI (p = 0.016), saccade latency (p = 0.007), fixation duration (p = 0.003), and the sum of accuracy and error correction rate (p = 0.003). There was still no significant difference in eye movement characteristics between the elderly healthy control group and the mild cognitive impairment group. However, in the comparison between mild cognitive impairment patients and moderate to severe dementia patients, the time to the first correct AOI (p = 0.026), saccade latency (p = 0.019) and fixation duration (p = 0.043) showed significant differences.

[0044] Table 5. Analysis of differences in eye movement parameters among control groups in the random horizontal positive saccade paradigm

[0045] The difference analysis of eye movement parameters of each control group in the random vertical forward saccade task is shown in Table 6. There were significant differences in the first fixation time (p=0.003) and fixation duration (p=0.003) between the young healthy control group and the elderly healthy control group. There were significant differences in the time to the correct AOI for the first time (p=0.005), the time to stay in the correct AOI (p=0.022), and the saccade latency (p=0.007) between the elderly healthy control group and the mild cognitive impairment group. In addition, there were significant differences in accuracy (p=0.009), fixation duration (p=0.043), and the sum of accuracy and error correction rate (p=0.009) between mild cognitive impairment patients and moderate to severe dementia patients.

[0046] Table 6. Analysis of differences in eye movement parameters among the control groups in the random vertical positive saccade paradigm

[0047] From the above analysis, it can be seen that the random forward saccade task showed a higher cognitive impairment assessment effect than the original forward saccade task. For patients with mild cognitive impairment and moderate to severe dementia, there was no difference in the eye movement characteristics in the original horizontal forward saccade task, while multiple eye movement characteristics (SA / TFCA / TDF / SCR) in the random forward saccade task showed obvious differences, which shows that increasing the difficulty of the forward saccade paradigm is helpful to improve the ability to distinguish mild cognitive impairment and moderate to severe dementia. In the random vertical forward saccade task, more saccade parameters (TFCA, DTCA, SL) showed significant differences between the healthy elderly control group and the mild cognitive impairment group, indicating that the vertical saccade task showed higher sensitivity to the analysis of mild cognitive impairment than the horizontal saccade task.

[0048] Furthermore, the difference analysis of eye movement characteristics of each control group in the horizontal reverse saccade paradigm is shown in Table 7. The comparison results between the young healthy control group and the elderly healthy control group showed that, except for the peak velocity and saccade latency, other eye movement characteristics showed significant differences. In addition, there was a significant difference in the time from the first time to the correct AOI between the healthy elderly group and the mild cognitive impairment patients (p=0.010). The average time from the first time to the correct AOI in the healthy elderly group was 537.67 ms, while the mild cognitive impairment group was significantly extended to 749.04 ms, which was significantly higher than the healthy elderly group. This indicates that the attention of patients with mild cognitive impairment is impaired and the efficiency of task execution is reduced. There were also significant differences in accuracy (p=0.045), fixation duration (p=0.045), first fixation time (p=0.009) and the sum of accuracy and error correction rate (p=0.001) between patients with mild cognitive impairment and patients with moderate to severe dementia.

[0049] Table 7. Analysis of differences in eye movement parameters among the control groups in the horizontal antisaccade paradigm

[0050] The difference analysis of eye movement characteristics of each control group in the vertical countersaccade paradigm is shown in Table 8. Similar to the horizontal countersaccade, the comparison results between the young healthy control group and the elderly healthy control group showed that, except for the peak velocity and saccade latency, other eye movement characteristics showed significant differences. When further analyzing the differences between the elderly healthy control group and patients with mild cognitive impairment, it was found that there was a significant difference in the correct AOI dwell time (p=0.022). The correct AOI dwell time of patients with mild cognitive impairment (1097.86ms) was significantly lower than that of the elderly healthy control group (1289.27ms), indicating that the attention of patients with mild cognitive impairment in this task was reduced. In addition, patients with mild cognitive impairment showed significant differences in accuracy (p=0.024), fixation duration (p=0.011), first fixation time (p=0.024), and the sum of accuracy and error correction rate (p=0.009) compared with patients with moderate and severe dementia.

[0051] Table 8. Analysis of differences in eye movement parameters among the control groups in the vertical countersaccade paradigm

[0052] Compared with all forward saccade tasks, the original reverse saccade paradigm had more eye movement features with significant differences between young controls and healthy elderly controls, and between patients with mild cognitive impairment and patients with moderate to severe dementia. However, for healthy elderly controls and mild cognitive impairment, the original reverse saccade paradigm had lower sensitivity and its discriminative power was still obviously insufficient.

[0053] Furthermore, the difference analysis of eye movement characteristics of each control group in the random horizontal reverse saccade paradigm is shown in Table 9. There are significant differences in other eye movement characteristics except peak velocity between the young healthy control group and the elderly healthy control group. There are significant differences in accuracy (p=0.021), correct AOI dwell time (p=0.021), and the sum of accuracy correction rate (p=0.012) between the elderly healthy control group and the mild cognitive impairment group. There are significant differences in accuracy (p=0.010), time to the first correct AOI (p=0.009), correct AOI dwell time (p=0.001), fixation duration (p=0.043), and the sum of accuracy correction rate (p=0.004) between mild cognitive impairment patients and moderate to severe dementia patients.

[0054] Table 9. Analysis of differences in eye movement parameters among the control groups in the random horizontal antisaccade paradigm

[0055] The difference analysis of eye movement characteristics of each control group in the random vertical countersaccade task is shown in Table 10. Similar to the random horizontal countersaccade task, there were significant differences in eye movement characteristics except peak velocity between the young healthy control group and the elderly healthy control group. There were significant differences in accuracy (p=0.038), time to the first correct AOI (p=0.044), time to stay in the correct AOI (p=0.007), and the sum of accuracy and error correction rate (p=0.011) between the elderly healthy control group and the mild cognitive impairment group. In the comparison between mild cognitive impairment patients and moderate to severe dementia patients, there were significant differences in eye movement characteristics except peak velocity.

[0056] Table 10. Analysis of differences in eye movement parameters among the control groups in the random vertical countersaccade paradigm

[0057] According to the above analysis results, compared with the original antisaccade task, more eye movement features showed significant differences when using the random antisaccade task to analyze cognitive impairment. For example, for the healthy elderly control group and patients with mild cognitive impairment, only one eye movement feature (TFCA) in the original antisaccade task was significantly different, while four eye movement features (SA, TFCA, DTCA, SCR) in the random antisaccade task were significantly different. In addition, compared with the random horizontal antisaccade task, the random vertical antisaccade task showed a stronger sensitivity to cognitive impairment. Especially between patients with mild cognitive impairment and patients with moderate to severe dementia, more eye movement features in the random vertical antisaccade task were significantly different.

[0058] In summary, the saccade task showed significant differences in different control groups. First, the saccade task showed high sensitivity in distinguishing young and elderly healthy individuals, verifying the effect of age on eye movement ability, which indicates that it is feasible to analyze age-related cognitive and physiological changes based on eye movement characteristics. Second, as the difficulty of the saccade task increased, the differences in eye movement characteristics between the elderly healthy control group, the MCI group, and the moderate to severe dementia group gradually increased. In particular, for the elderly healthy group and the MCI group, there were significant differences in eye movement characteristics only in the random horizontal forward saccade, random horizontal reverse saccade, and random vertical reverse saccade tasks. Finally, the vertical saccade task was more sensitive to the detection of cognitive impairment, which may be related to its task complexity and differences in eye movement mechanisms in different directions. Therefore, saccade tasks with multiple difficulty levels have important application prospects in the early screening of cognitive impairment and the assessment of cognitive function.

[0059] Furthermore, in order to further verify the effectiveness of the proposed paradigm in the assessment of cognitive impairment, the present disclosure conducted a correlation analysis between MoCA scores and eye movement parameters in various task scenarios. Using the Spearman correlation analysis method, the correlation coefficient and significance between the two were calculated, and correlation graphs were drawn to explore the impact of task difficulty on the patient's eye movement performance. The correlation graph between MoCA scores and eye movement characteristics is shown in Figure 2. Figure 4 As shown. Because the accuracy and error correction rates are discretely distributed and cannot be correlated with the MoCA score, this disclosure only analyzes the remaining 6 eye movement features. Including: (1) Correct AOI dwell time Correlation analysis between MoCA scores and correct AOI dwell time under different paradigms Figure 4 As shown in (a) in the figure. As can be seen from the figure, the slope in the random reverse saccade task is the largest (horizontal random reverse saccade task r=0.552, p<0.000; random vertical reverse saccade task r=0.513, p<0.000), indicating that in these two tasks, the correlation between the participants' MoCA score and the correct AOI dwell time is the strongest. In addition, these two task scenes are located at the bottom of the figure, which shows that the difficulty of these two tasks is relatively high, and all participants generally stay in the correct AOI for a shorter time. In the positive saccade paradigm, the MoCA score and the correct AOI dwell time show a lower correlation (horizontal positive correlation coefficient r=0.193, p=0.166, vertical positive correlation coefficient r=0.309, p=0.024).

[0060] (2) Saccadic latency Correlation analysis between MoCA scores and saccade latency Figure 4 As shown in (b) in the figure. As can be seen from the figure, the slope of random vertical reverse saccade is the largest (correlation coefficient r=-0.570, significant p<0.000), which is located at the top of the graph. This shows that in this task, the saccade latency of all participants is longer and has a stronger correlation with the MoCA score. In addition, it can be clearly seen from the figure that only the saccade latency of the random vertical reverse saccade task shows a significant correlation. This result shows that the saccade latency, an eye movement feature, performs differently in the vertical and horizontal directions, which is of great significance for analyzing different degrees of cognitive impairment.

[0061] (3) First fixation time Correlation analysis between MoCA scores and first fixation time in different paradigms Figure 4As shown in (c) in the figure. The figure shows that the slopes of the horizontal antisaccade task and the vertical antisaccade task are larger (horizontal antisaccade task r = -0.430, p=0.001; vertical antisaccade task r=-0.409, p=0.002), showing a higher correlation with the MoCA score, indicating that these two tasks have a stronger ability to distinguish different degrees of cognitive impairment. In addition, the first fixation time in all tasks is significantly correlated with the MoCA score.

[0062] (4) Peak saccade velocity Figure 4 (d) shows the correlation analysis between MoCA scores and peak saccade velocity in different task paradigms. The results show that in each saccade task, the correlation between peak saccade velocity and MoCA scores is weak and presents irregular positive and negative correlations. This result indicates that the application value of peak saccade velocity in assessing the degree of cognitive impairment is limited.

[0063] (5) First arrival at the correct AOI time like Figure 4 As shown in (e), the correlation analysis results between MoCA scores and the time to the correct AOI for the first time show that the slope of the random reverse saccade task is the largest (horizontal random reverse saccade r = -0.397, p = 0.003; vertical random reverse saccade r = -0.456, p = 0.001), and it is located at the top of the graph. This shows that these two tasks are more difficult, and the time for all participants to reach the correct AOI for the first time is relatively long, and the correlation with MoCA scores is more significant.

[0064] (6) Center fixation duration Figure 4 (f) in the figure shows the correlation analysis between MoCA scores and fixation duration in different paradigms. As can be seen from the figure, the slope of the reverse saccade paradigm is larger (horizontal reverse saccade paradigm r=0.486, p<0.000; vertical reverse saccade paradigm r=0.394, p=0.003), which is located at the bottom of the chart, indicating that the fixation duration of these two tasks has a high correlation with the MoCA score, showing a stronger discrimination in distinguishing different degrees of cognitive impairment. In addition, the fixation duration in all paradigms is significantly correlated with the MoCA score.

[0065] In summary, the multi-level difficulty saccade paradigm designed in the present disclosure effectively analyzes the relationship between the subject's eye movement performance and task difficulty, and provides a basis for quantitative evaluation of the degree of cognitive impairment based on eye movement parameters. First, the two eye movement characteristics of the correct AOI residence time and the first time to the correct AOI are more sensitive to the difficulty of the saccade paradigm, and there are obvious differences between the forward saccade and reverse saccade paradigms. In the forward saccade task, the correlation between these two saccade characteristics and the MoCA score is low, or even non-correlated. In the random reverse saccade task, the correlation between these two eye movement parameters and the MoCA score is also significantly higher than that of the original reverse saccade task. Secondly, the saccade latency only shows a high correlation with the MoCA score in the random vertical reverse saccade task, which indicates that the saccade latency is greatly affected by the saccade direction. The peak saccade velocity is not related to the MoCA score under all paradigms, indicating that it cannot be used for cognitive impairment assessment. Finally, the two eye movement parameters of the first fixation time and the fixation duration have a certain correlation with the MoCA score under each task. However, these two eye movement features were most highly correlated with MoCA scores in the original reverse saccade paradigm, least highly correlated in the original forward saccade paradigm, and moderately correlated in the two paradigms containing random forward-reverse saccades. Because the difficulty level of the paradigm is: reverse saccade paradigm > random forward-reverse saccade paradigm > forward saccade paradigm, these two eye movement features are also affected by the difficulty of the saccade paradigm. The greater the difficulty of the paradigm, the worse the performance of the subjects and the stronger the sensitivity of the eye movement features.

[0066] The results of this public experiment show that the random reverse saccade task has a high sensitivity in the screening of cognitive impairment, especially the random reverse saccade task in the vertical direction shows the most significant discriminative power. Analysis of the eye movement characteristics in different paradigms found that the correct AOI dwell time and the time to the correct AOI for the first time were more sensitive to the task difficulty. In the random reverse saccade task, the correlation between these two eye movement parameters and MoCA scores was significantly enhanced; the correlation between saccade latency and MoCA scores in the random vertical reverse saccade task was significantly higher than that in the random horizontal reverse saccade task, proving that the saccade direction had a significant effect on this eye movement feature; the first fixation time and fixation duration were significantly correlated with MoCA scores in all tasks. However, unlike other indicators, these two indicators had the highest correlation with MoCA scores in the reverse saccade task, which may be due to the higher difficulty of the reverse saccade paradigm than the random forward-reverse saccade paradigm. In general, the multi-difficulty graded saccade paradigm designed in this study can discover more eye movement indicators that can be used for the assessment of cognitive impairment, conduct a fine-grained analysis of the cognitive level of the subjects from multiple dimensions, and provide an important reference for the early screening and assessment of cognitive impairment.

[0067] Example 2 In one embodiment of the present disclosure, an eye movement paradigm and index evaluation system for cognitive impairment detection is provided, comprising: Task design module, used to design eye movement paradigms with multiple difficulty levels; A data acquisition module is used to obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; The evaluation and analysis module is used to evaluate the cognitive ability of the participants and obtain MoCA scores; the normality of eye movement parameters is evaluated by statistical analysis methods, the differences in eye movement of participants at different task difficulty levels and eye movement directions are analyzed, and the correlation between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms is analyzed. The Spearman correlation analysis method is used to calculate the correlation coefficient and significance between the two, and the rank correlation diagrams are drawn respectively to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm; Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

[0068] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0069] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0070] Although the above describes the specific implementation methods of the present disclosure in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present disclosure. Technical personnel in the relevant field should understand that on the basis of the technical solution of the present disclosure, various modifications or variations that can be made by those skilled in the art without creative work are still within the scope of protection of the present disclosure.

Claims

1. An eye movement paradigm and index evaluation method for cognitive impairment detection, characterized in that: include: Design eye movement paradigms with multiple difficulty levels; Obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; The participants' cognitive abilities were assessed and MoCA scores were obtained; Statistical analysis methods were used to evaluate the normality of eye movement parameters, analyze the differences in eye movement of participants at different task difficulty levels and eye movement directions, and conduct correlation analysis between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms. Spearman correlation analysis method was used to calculate the correlation coefficient and significance between the two, and rank correlation diagrams were drawn to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm. Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

2. The eye movement paradigm and index evaluation method for cognitive impairment detection according to claim 1, characterized in that: Designing an eye movement paradigm with multiple difficulty levels means designing eye movement task paradigms with three difficulty levels, including forward saccades, reverse saccades, and random forward-reverse saccades. Each paradigm includes tasks in the horizontal and vertical directions.

3. The eye movement paradigm and index evaluation method for cognitive impairment detection according to claim 1, characterized in that: By testing the participants' performance in tasks of different difficulty levels, the differences in eye movement ability between different control groups were verified, and a high-precision eye tracker was used to collect eye movement data of people in different categories of control groups, including the gaze point coordinates and pupil diameter data of both eyes.

4. The eye movement paradigm and index evaluation method for cognitive impairment detection according to claim 3, characterized in that: The forward saccade paradigm consisted of six forward saccade tasks. At the beginning of each task, participants were asked to look at a blue ball displayed in the center of the screen. The ball disappeared and appeared as a target stimulus in the peripheral visual area of ​​the screen. Participants were asked to shift their gaze to the target stimulus position as quickly as possible. The display time of the ball at each position and the time interval between two consecutive forward saccade tasks were set.

5. The eye movement paradigm and index evaluation method for cognitive impairment detection according to claim 3, characterized in that: By calculating the empirical cumulative distribution function of the sample data Theoretical normal cumulative distribution function The maximum absolute difference between To measure the deviation between samples, and to determine whether the data conforms to the normal distribution, then, by calculating the rank sum of the two groups of samples to evaluate whether there is a significant difference in their distribution, assuming that the two groups of samples are and , the sample sizes are m and n respectively. First, merge and sort the two groups of samples, and calculate the rank of each sample value to find the sum of the ranks of the two groups. and , then, according to the formula and Calculate the U statistic for each group and take the smaller value As a test statistic, the corresponding p-value is finally calculated to determine whether there is a significant difference between the two groups of data.

6. The eye movement paradigm and index evaluation method for cognitive impairment detection according to claim 1, characterized in that: Eye movement characteristics include saccade accuracy, saccade error correction rate, time to first enter the correct AOI, time spent in the correct AOI, saccade latency, first fixation time, and center point continuous fixation time.

7. Eye movement paradigm and index evaluation system for cognitive impairment detection, characterized in that: include: Task design module, used to design eye movement paradigms with multiple difficulty levels; A data acquisition module is used to obtain the eye movement parameters of the participants under the eye movement paradigms of different difficulty levels, and based on the eye movement parameters under the eye movement paradigms of different difficulty levels, extract the eye movement features under the eye movement paradigms of different difficulty levels; The evaluation and analysis module is used to evaluate the cognitive ability of the participants and obtain MoCA scores; the normality of eye movement parameters is evaluated by statistical analysis methods, the differences in eye movement of participants at different task difficulty levels and eye movement directions are analyzed, and the correlation between MoCA scores and eye movement characteristics under different difficulty level eye movement paradigms is analyzed. The Spearman correlation analysis method is used to calculate the correlation coefficient and significance between the two, and the rank correlation diagrams are drawn respectively to obtain eye movement indicators with significant differences between participants with different degrees of cognitive impairment under each paradigm; Cognitive ability assessment was achieved based on eye movement indicators that showed significant differences among participants with different degrees of cognitive impairment under each paradigm.

8. The eye movement paradigm and index evaluation system for cognitive impairment detection according to claim 7, characterized in that: Designing an eye movement paradigm with multiple difficulty levels means designing eye movement task paradigms with three difficulty levels, including forward saccades, reverse saccades, and random forward-reverse saccades. Each paradigm includes tasks in the horizontal and vertical directions.

9. The eye movement paradigm and index evaluation system for cognitive impairment detection according to claim 7, characterized in that: By testing the participants' performance in tasks of different difficulty levels, the differences in eye movement ability between different control groups were verified, and a high-precision eye tracker was used to collect eye movement data of people in different categories of control groups, including the gaze point coordinates and pupil diameter data of both eyes.

10. The eye movement paradigm and index evaluation system for cognitive impairment detection according to claim 8, characterized in that: The forward saccade paradigm consisted of six forward saccade tasks. At the beginning of each task, participants were asked to look at a blue ball displayed in the center of the screen. The ball disappeared and appeared as a target stimulus in the peripheral visual area of ​​the screen. Participants were asked to shift their gaze to the target stimulus position as quickly as possible. The display time of the ball at each position and the time interval between two consecutive forward saccade tasks were set.

Images