Exploring the Connection between Eye Movement Parameters and Eye Fatigue

Eye fatigue, a prominent symptom of computer vision syndrome (CVS), has gained significant attention in various domains due to the increasing diversification of electronic display devices and their widespread usage scenarios. The COVID-19 pandemic has further intensified the reliance on these devices, leading to prolonged screen time. This study aimed to investigate the effectiveness of utilizing eye movement patterns in discriminating fatigue during the usage of electronic display devices. Eye movement data was collected from subjects experiencing different levels of fatigue, and their fatigue levels were recorded using the T/CVIA-73-2019 scale. The analysis revealed that features related to the pupils demonstrated a high level of confidence and reliability in distinguishing fatigue, especially related to pupil size. However, features associated with fixations, such as fixation duration and frequency, did not significantly contribute to fatigue discrimination. Furthermore, the study explored the influence of subjective awareness on fatigue discrimination. By modifying the experimental settings and considering the subjects’ subjective perception, it was observed that individual consciousness and self-awareness played a crucial role in fatigue discrimination. The implications of these findings extend beyond the field of computer vision syndrome, offering potential applications in developing interventions and strategies to alleviate eye fatigue and promote eye health among individuals who extensively use electronic display devices.


Introduction
Visual fatigue, also known as eye fatigue, refers to the excessive strain experienced by the visual system due to prolonged or intensive visual tasks [1] [2].It is a common issue faced by individuals who engage in activities such as prolonged computer usage, reading for extended periods, or driving long distances.The demands of modern society, where screens and digital devices are ubiquitous, have led to an increase in cases of visual fatigue.
The problem of visual fatigue is a major manifestation of computer vision syndrome [3], which is influenced by both accommodation and intraocular mechanisms [4].Accommodation mechanisms can lead to blurred vision, diplopia, presbyopia, myopia, and slow focus change.Prolonged use of digital devices can significantly affect accommodation lag.According to the American Optometric Association, using electronic devices for more than two hours greatly increases the risk of computer vision syndrome [5].The effects on accommodation, convergence, and pupil size are primarily caused by the near working condition rather than the screen itself [6] [7].In response to these mechanisms, studies on blue light filtering to reduce eyestrain have been a hot topic in the field [8].Another aspect of visual fatigue is related to the ocular surface mechanism, which can cause symptoms such as dryness, redness, gritty sensation, and burning of the eyes after prolonged electronic display use.Blinking plays a crucial role in maintaining normal ocular surface water circulation throughout the tear secretion cycle, ocular surface wetting, ocular surface evaporation, and tear expulsion [9].Studies have shown a significant decrease in blink frequency during computer use, indicating the onset of visual fatigue [10].To analyse the fatigue state, many studies have used blink frequency as a primary indicator.These studies have observed a decrease in blink frequency from 18.4 blinks/min to 3.6 blinks/min during computer use.Another study reported a decrease from 22 blinks/min to 7 blinks/min.Related reports show that compared with the pre-COVID-19 era, the incidence of this symptom has soared from 5% to 65% to 84% to 90%, and the prevalence in children has increased to 50% to 60% [11].
One essential aspect of studying visual fatigue is examining the relationship between visual fatigue and eye movements [12].Eye movements, such as fixations, saccades, and pupil movement, play a vital role in visual perception and attention [13].Previous research suggests that changes in eye movements may indicate visual fatigue, as fatigue can impact the efficiency and accuracy of these movements [14].Investigating this relationship can provide valuable insights into the underlying mechanisms of visual fatigue and contribute to more effective diagnostic and preventive measures.
While numerous studies have been conducted on fatigue detection in the driving field, researchers have focused on exploring the impact of different signal sources, detection methods, and evaluation methodologies.Research mainly includes video-based methods [15][16][17] [18], retinal thicknessbased methods [19], SSVEP based methods [20], electrooculography-based methods [21] and electrocardiogram-based methods [22][23] [24].The study of electrooculography has been a key aspect in the research of eye fatigue.By calculating the amplitude changes of the electromyography of the eye muscle and the horizontal and vertical electromyography of the eye, the velocity of eye movement was measured, and the time-effectiveness was compared longitudinally, so as to judge the degree of eye fatigue [25].Electrooculography plays a crucial role in the research of reconstructing eye movement trajectories [26].Most fatigue monitoring algorithms based on electrical signals use blink frequency and average eye closing time as indicators to determine the state.Some studies also perform time-frequency conversion on the acquired EOG signals, extract features from the frequency domain, and perform fatigue analysis [27] [28].However, because this method is nonlinear in nature, it has its own limitations [29].In addition, there are some multi-modal fusion algorithms for eye fatigue discrimination, such as fatigue monitoring algorithms evaluated by electrocardiogram sensors, skin reflected current, skin temperature and other indicators [30] and by Bayesian network based on EEG and β wave power of brain stem reflection [31].
In recent years, eye-tracking technology has emerged as a powerful tool for studying visual fatigue.By precisely and non-invasively measuring eye movements, eye-tracking provides detailed insights into visual behaviour and attention during various tasks [32], Compared to traditional self-report methods and observation techniques, eye-tracking offers several advantages, mainly including objective and quantitative assessments of visual fatigue.This technology enables researchers to gain a deeper understanding of visual fatigue and develop more effective strategies for detecting and managing it.
The purpose of this paper is to explore the relationship between visual fatigue and eye movements in the daily use of electronic display terminals, and to focus on the advantages of using eye-tracking technology to detect and evaluate visual fatigue.Through the design of a fatigue measurement experiment and the extraction of relevant characteristic indexes based on the principles of eye movement, our aim was to objectively assess the feasibility of using eye-tracking technology to measure fatigue level and identify effective indicators.

Setup
In this study, we employed a self-developed infrared eye-tracking system to accurately record the eye movements of the participants.This system enables real-time tracking of eye positions and pupil size, offering a resolution of 3840 by 2160 and a sampling rate of 25Hz.To ensure the reliability of the collected eye movement images, participants were instructed to face the camera as much as possible during the experiment.The distance between participants and the camera was strictly controlled within a range of 45-55cm, and a nine-point calibration was conducted prior to the commencement of formal tasks to guarantee data accuracy.To capture the data characteristics of the participants under various levels of fatigue, this test required them to undergo an assessment both at the beginning and at the end of the test day.

Participants
The experiment was carried out for three months, and a total of 47 valid sets of experimental data were collected.Thirteen subjects, consisting of seven males and six females, were selected for the study.The inclusion criteria for participant selection were as follows: a) Refractive errors in the primary and nonprimary eyes that met specific requirements for each group, with a cylindrical dioptre less than 0.50D and an equivalent spherical dioptre difference between the eyes of less than 1.00D.b) Corrected binocular visual acuity of 1.0 or higher.c) Students aged between 22 and 26 are enrolled in a master's program.d) Normal performance on visual acuity tests.
Participants with the following conditions were excluded from the study: a) Use of medications that may affect changes in pupil size or mechanisms of ocular surface regulation.b) Non-myopic eye diseases, such as retinal detachment and strabismus, that can be corrected with lenses.c) Presence of hereditary developmental disorders or cognitive impairments.d) History of alcohol consumption, drug abuse, or use of medications that affect cognition, such as antiepileptics, antipsychotics, or anticholinergics.

Fatigue discrimination questionnaire(T/CVIA-73-2019)
Prior to the experiment, participants were required to complete an eye fatigue questionnaire to assess their fatigue levels during the study.The questionnaire utilized the Scale evaluation method, which is the second part of the visual fatigue test and evaluation method released by the CHINA VIDEO INDUSTRY ASSOCIATION (CVIA) on July 19, 2019.This scale has been validated through research conducted both domestically and internationally, demonstrating accuracy and precision in measuring the level of eye fatigue.
The questionnaire primarily focuses on assessing various aspects of eye fatigue, including dry eyes, blurred vision, diplopia, tearing eyes, burning eyes, eye pain, tearing eyes, eye tightness, and eye irritation.Additionally, it evaluates head fatigue (e.g., headache, dizziness) and limb fatigue resulting from the use of display terminals.The questionnaire provides five options for each question, with a score of 0, 25, 50, 75, and 100 points.At the end of the questionnaire, the mean score was calculated as the current scale score.Score below 15 indicates a low fatigue level in practical application, while a score of 15 or higher indicates a high fatigue level [33].

Visual acuity test.
Upon completion of the questionnaire, a visual acuity test was conducted consisting of two parts.In the first part, a circular picture was displayed on a blank screen for a brief duration.The circular pictures appeared randomly in either red, blue, or green, and participants were instructed to select the color that had just appeared within the subsequent five seconds.The second part of the test involved identifying a picture with stripes among four identical solid-colored square pictures.The number of stripes varied from 3 to 15.The objective of this experiment was to assess the participants' visual acuity.In the first and second parts of the test, participants were required to achieve 95% and 90% accuracy, respectively, to demonstrate normal visual acuity.

Scintillation test.
In this study, a flash task was utilized to examine potential differences in eye tracking metrics across varying levels of fatigue.Previous research has indicated that the characteristics of blink and pupil changes are more pronounced in scenes with significant variations [34].If the ability of the eye muscles to adjust during fatigue is diminished, the range of pupil variation should narrow, and the response rate may decrease in an environment with rapidly changing stimuli.
The study employed a four-part scintillation test, which consisted of the following components: the fast flicker test (FF), slow flicker test (SF), luminance gradient test (LG), and colour flicker test (CF).Initially, participants underwent five fast flash tests.In each trial, the screen rapidly blinked at a frequency of 5Hz for ten seconds, with five seconds of rest between groups.This procedure was repeated for a total of five sessions.Subsequently, the slow flash test was conducted using the same timing and intervals, but with the screen blinking at a frequency of 1Hz.
The third part of the task involved stimuli that were significantly reduced compared to the previous two parts.In this phase, the screen brightness changed from 0 to 255 at a frequency of 20Hz.Each experiment was repeated five times, with five seconds of rest between groups.The final part of the task examined the response of the eyes to different colours.The screen cycled through the colours blue, green, red, yellow, magenta, and cyan at a frequency of 1Hz.Throughout the experimental phase, participants were instructed to focus their gaze on a cross positioned at the centre of the screen.

Visual search test.
The study aimed to investigate the impact of eye fatigue on subjects' task response levels through a visual search test (VS).This test comprised three distinct steps, each serving a specific purpose: Step 1 (Resting State): During this initial phase, the screen was completely black, except for a small yellow dot positioned at the centre.Subjects were instructed to fixate their gaze on the yellow dot for a duration of two seconds.
Step 2 (Preparing State): This phase assessed the subjects' ability to resist interference.They were required to maintain their fixation on the yellow dot, while six red rings with a radius of 30 pixels appeared on the screen.These rings were evenly distributed around the yellow dot, forming a circle with a radius of 400 pixels.This stage lasted for two seconds.
Step 3 (Searching State): In the final stage, one of the six rings was filled with grey.Subjects were considered successful in their search if they relocated the position of their fixation point within the greyfilled ring within one second.Each trial was repeated eight times, with the first four trials serving as non-interference trials, and the remaining four trials involving the same circle positioned at different locations to interfere with the subjects' target search.
In the third part of the experiment, we intentionally shortened the time allotted for the visual search process.This was done to minimize subjectively driven movements of the fixation point towards the target area.Our aim was to encourage subconscious search behaviours to objectively assess the current level of eye fatigue.(c)

Reading test.
To replicate the typical usage of electronic display terminals in daily life, we incorporated a free reading test (RT) at the conclusion of our experiment.During this reading test, participants were given one minute to read an unlimited number of paragraphs.The excerpts were taken from the renowned novel "Rickshaw Boy" and ranged between 450 and 500 words in length.The reading material was presented in a full-screen format, with black text displayed on a white background.The test was repeated three times, with a five-second break between each trial.

Feature calculation
Based on the experimental design and task requirements, various features were extracted from different experiments to capture the overall characteristics.These characteristics can be categorized into three main aspects: Pupil Characteristics: number of blinks (BN), average duration of eye closures (BD), maximum pupil sizes (PMAX), minimum pupil sizes (PMIN), mean velocity of pupil change (PV), response time for the pupil to adjust to screen changes (PL).
Fixation Characteristics: number of fixations (FN), average duration of fixations (FD), fixation as a percentage of total time (FR), number of saccades (SN), average distance covered during saccades (SD).
Task-Type Characteristics: proportion of fixations on the target area of visual search task (TFR), response time for the pupil to search the target area of visual search task (TSL), success rate of visual search tasks (TSR), proportion of fixations on the centre of the screen for tasks requiring (CFR).

Statistical analysis
To analyze the data, we employed IBM SPSS Statistics 22.The goal of the analysis was to investigate the impact of fatigue levels on eye movement characteristics in different task environments.To assess group differences across various fatigue levels, a two-way ANOVA method was employed.In the analysis process of this study, feature difference under analysis was performed according to different interval segments of different tasks, such as the blinking task to distinguish black and white changes on the screen, and the visual search task to distinguish different task stages.

Results
The statistical analysis revealed several significant findings regarding the different eye movement characteristics: Pupil Characteristics: The maximum and minimum pupil sizes exhibited significant differences across all task scenarios, demonstrating a noticeable increase in the mean level.Among the other pupil characteristics, the mean pupil change velocity displayed significant differences among different fatigue groups within the fast flash task.Additionally, there was a significant difference in the response latency of pupil changes during the scene with the screen turning white in the slow flicker task.However, the blink number and mean eye closure duration did not show any significant differences across all tasks.
Fixation Features: Only the proportion of total fixation duration showed significant differences in the colour mutation task, while the average saccade distance exhibited significant differences in the reading task.No significant differences were observed in other tasks and features.
Task Characteristics: There were no significant differences in the proportion of attention concentration across all tasks.However, there were significant differences in the reaction time, accuracy, and proportion of target fixation within the visual search task.The data indicated that the low fatigue level corresponded to significantly higher accuracy and proportion of target fixation compared to the high fatigue level, while the response time was significantly lower in the low fatigue level.
These findings provide valuable insights into the relationship between eye movement characteristics and fatigue levels in different task environments.The visualization results of the data are drawn into a box plot, which is shown in Fig. 2 and Table1-Table3.In these figures and tables, we use abbreviations to simplify the names of the tests and features with significant differences are highlighted in red.

Discussion
In this study, we utilized the T/CVIA-73-2019 fatigue analysis scale to identify fatigue status and analyse the differences in fatigue levels across various tasks.The purpose of this study was to verify the effectiveness of eye movement features in identifying eye fatigue in the daily use of electronic displays, and to test the feasibility of using eye movement features to classify eye fatigue.When considering overall feature categories, gaze behaviour exhibited poor differentiation, while pupil-related features performed well, indicating the significant influence of subjective consciousness on fixation behaviour.This aligns with the notion that individuals may not subjectively perceive eye fatigue under conditions of high mental stress, even if eye fatigue is present.Therefore, future research should focus on minimizing the influence of subjective consciousness in experimental design or selecting objective features that accurately reflect fatigue levels.Additionally, further investigation is necessary to explore the inherent association between specific trait changes and fatigue levels, as well as unravel the underlying mechanisms.By refining experimental design and selecting objective features, future studies can enhance the accuracy and reliability of fatigue assessments.
Pupil characteristics were found to be effective in analysing differences between groups.Regardless of whether there was a large area brightness change or a small local change on the screen, the maximum and minimum pupil area exhibited significant differences in response to fatigue levels.Notably, when the brightness change was rapid, the speed of pupil changes also significantly varied.Moreover, there were significant differences in pupillary response latency when transitioning from black to white at a slower frequency.These findings demonstrate that the accommodative mechanism influenced by eye fatigue performs noticeably during actual measurements.The pupil's accommodative ability serves as a direct reflection of the degree of eye muscle fatigue, indicating its high reliability.
While most fixation-related features did not show significant differences, the mean saccade distance exhibited significant variations during the reading task.This phenomenon can be explained that fixation behaviour is heavily influenced by the subjective thoughts of the subjects, whereas reading habits are only slightly affected by the degree of fatigue.However, the average saccade distance objectively reflects the subjects' current information receiving speed during reading, which decreased as fatigue levels increased.In the task involving colour mutation, the subjects' fixation frequency significantly decreased.This proves that although there were no significant differences in the performance of this feature under relatively mild stimuli, it became challenging to maintain fixation behaviour under relatively strong stimuli, resulting in a significant difference.Therefore, this finding provides further evidence.By controlling the experimental design and task, the influence of subjects' subjective consciousness on the experimental results can be mitigated by increasing the stimulus intensity.
Among the task features, all the task features of the visual search test displayed significant differences.This finding confirms the significant impact of experimental design on the discriminative characteristics of eye fatigue.In the experimental design stage, the reaction time of the subjects in the search stage was intentionally compressed, leading to a bias in the subjects' reaction process, which was guided by the objective visual level.As a result, the subjective search consciousness and subjective driving ability of the subjects were reduced, resulting in more favourable experimental results.As fatigue increased, the proportion of target gaze, reaction time, and accuracy of the subjects decreased to varying degrees, aligning with the experimental expectations.

Conclusion
The purpose of this study is to verify the feasibility of using eye movement indicators to identify eye fatigue during daily electronic screen use and to establish a reliable fatigue discrimination system by using fatigue scales and fatigue tests.Our findings revealed distinct responses to fatigue across different feature categories, with their underlying patterns being related, to varying degrees, to the causes of eye fatigue and attention mechanisms.To gain a deeper understanding of this association, future studies should focus on exploring the influence of subjective factors on fatigue detection.The objective evaluation index and characteristic system of eye fatigue were thoroughly examined in this study, highlighting the potential of eye tracking technology for objective assessment of eye fatigue.

Figure 1 .
Figure 1.(a) Scatter plot of fatigue scale.(b) Schematic diagram of the experiment.(c) Flowchart of the visual search task.

Figure 2 .
Figure 2. Distribution of features with significant differences

Table 1 .
Data Analysis of Pupil Characteristics.

Table 2 .
Data Analysis of Fixation Characteristics.

Table 3 .
Data Analysis of Task Characteristics.