The Short- and Long-Term Perceptual Learning of Clinical Dynamic Visual Acuity Test

Abstract

Purpose: The purpose of this study was to investigate the short- and long-term learning effect of dynamic visual acuity (DVA) tests.

Methods: Participants between 18 and 30 years with corrected to normal visual acuity were enrolled in this study. Three repeated sessions were performed, with 15 minutes and 15 days’ intervals between sessions. Each session included 9 DVA tests of horizontal, vertical, and diagonal motions of E optotypes at 20, 40, and 80 degrees per second (dps). The short- and long-term learning effects were analyzed from repeated DVA tests.

Results: Of the 58 enrolled participants, the mean age was 23.1 ± 2.1 years. DVA significantly varied among motion types and velocities (P < 0.05, respectively). There was a significant short-term learning effect for 20 (P = 0.004), 40 (P < 0.001), and 80 (P = 0.014) dps DVA test of horizontal motion, 40 dps DVA test of vertical (P = 0.003), and diagonal motion (P = 0.036). The long-term learning effect was detected in the 40 dps diagonal motion DVA test (P = 0.015). The short- and long-term learning effects were positively associated with initial DVA in most combinations of motion type and velocity tests (P < 0.05, respectively). The short- (P = 0.031) and long-term (P = 0.024) learning effect of 80 dps horizontal motion DVA test was greater in male than female participants.

Conclusions: There is a significant short-term learning effect in the DVA test of various motion types, but the long-term learning effect was rarely observed, and it is greater in participants with worse initial DVA.

Dynamic visual acuity (DVA) refers to the ability to identify the details of objects that have relative motion to observers, which is crucial to daily task performance.¹ In life scenarios, the object and the observer are often in relative motion. Therefore, DVA may manifest the real-life visual function better and be more sensitive to visual disturbance than static visual acuity.^2–4 There are two main types of DVA tests in clinical settings: DVA tests with static or moving optotypes.⁵ DVA test with static optotypes evaluated vestibular-ocular reflex.^6,7 In contrast, DVA tests with moving optotypes present standardized optotypes mechanically or digitally and are used to assess dynamic visual function.^8–10 Currently, DVA is a promising indicator for real-life functional vision in clinical ophthalmology, including cataract, refractive surgery, and dry eye.^11–14 During the DVA test, the patients were required to identify the opening direction of the moving optotype of a certain speed, and the minimum size that the patients could identify correctly was recorded as DVA. 

Perceptual learning defines performance improvement in specific tasks after training or practice, which occurs in almost all visual tasks.¹⁵ It could improve the accuracy and reduce response times in determining motion direction, orientation, and spatial frequency in experiment settings.^16–18 However, it may decrease the reliability of the first measurement of some subjective examinations, like visual field tests.^19,20 Thus, investigating the learning effect of the DVA test is crucial for its clinical application and result interpretation. 

Using the mechanically horizontal moving Landolt-C target, previous research demonstrated a significant short learning effect on DVA after repeated practice with the task.^21,22 In addition, in two tests across a 2-week interval, DVA measurement showed good test-retest repeatability in horizontal, vertical, and oblique motion types,²³ which provides indirect clues that there is no long-term learning effect. There are critical limitations in previous studies. Static visual acuity significantly affects DVA,²⁴ but the static visual acuity was not normalized before the DVA test in previous research. Additionally, the short- and long-term learning effect and its influential factors in DVA tests of different motion types have not been investigated and compared before. 

Thus, the present research enrolled participants with corrected to normal static visual acuity. The DVA of horizontal, vertical, and diagonal motion across different velocities was measured repeatedly in short- and long-term intervals. The short- and long-term learning effects were calculated accordingly, and gender and sports experience were investigated as potential factors that could influence the learning effect. The research provides the basis for optimizing the DVA test procedure to obtain more reliable results and interpretation of test results. 

Methods

Participants

Fifty-eight adult participants were recruited for the study, and men accounted for 46.6% of the participants. The participants were eligible for inclusion in the study if they were 18 to 30 years of age with normal or corrected to normal visual acuity. The exclusion criteria were as follows: (a) high myopia (noncycloplegic autorefraction ≤ −6.00 diopters [D]); (b) history of severe ocular diseases, including severe ocular surface disease, glaucoma, and retinal diseases; eye alignment, or extraocular movement problem; and (c) cognitive disorders. The research followed the tenets of the Declaration of Helsinki, and the protocol was approved by the local review board. All participants provided written informed consent before enrollment after an explanation of the aim and procedure of the study. 

Dynamic Visual Acuity Test

The DVA test conducted in the present research evaluates moving objects’ DVA. The test stimuli, apparatus, and procedures were identical to our previously published article.²⁵ The test optotypes and paradigm were programmed with the Psychtoolbox plugin MATLAB 2017b (MathWorks, Natick, MA, USA). The stimulus was the letter E from the standard logarithmic visual chart. The presenting size of the optotype was calculated according to the test distance and the viewing angle of the same size in the standard logarithmic visual chart. The moving velocity was quantified as the viewing angle changed per second. The test optotypes were presented on a 24-inch in-plane switching screen. The screen refresh rate was 165 hertz (Hz), and the response time was 1 ms. There were three motion types of the optotypes. For horizontal motion, the letter E appeared in the middle of the screen's left side and moved horizontally to the right side. For vertical motion, the letter E appeared in the middle of the screen's upper side and moved vertically down to the lower side. For diagonal motion, the letter E appeared randomly in one of the four corners of the screen to move diagonally to the other corner. The diagonal motion task was included to test unpredictable motion, compared with predictable motion. 

The test distance was set at 2.5 meters. During the test, the participant was required to judge the opening direction of the moving letter E in a four alternative forced choice regime and responded with the arrow keys of the keyboard. The opening direction of the letter E was randomized in each trial. The test started with pretraining to make the participant aware of the test procedure and optotype, followed by the formal test. During the pretraining, five moving optotypes of four sizes bigger (0.2 to 0.4 LogMAR) than the static visual acuity with random opening direction were presented consecutively. The following formal test began with the same size as the pretraining. Eight optotypes of the same size were presented one by one. If five out of eight optotypes could be identified correctly, it would be changed to one size smaller until less than five optotypes were identified of the same size. The result was calculated according to the principle of the LogMAR visual chart.²⁵ 

Procedure

Before the test, all enrolled subjects underwent a static visual acuity test (LogMAR visual chart) with daily wear spectacles, automatic refraction (Topcon, KR-800) and fundus examination (Topcon, TRC-NW400CN). Subjective refraction was performed based on automatic refraction for patients with static visual acuity worse than 0 (LogMAR). The sphere and cylinder were tuned with the trial lens to obtain no worse than 0 (LogMAR) static visual acuity in each eye. The DVA test was performed with the naked eye if the static visual acuity was 0 (LogMAR) or better. Otherwise, the DVA was evaluated with the patient wearing corrected distance visual acuity eyeglasses to obtain no worse than 0 (LogMAR) corrected static visual acuity. 

The study flow chart is demonstrated in Figure 1. Three repeated sessions were performed, with 15-minute intervals between the first and second sessions and 15 days ± 2 days intervals between the second and third sessions. The day before each session, the participants were not allowed to drink stimulating beverages or eat stimulating food, which was confirmed before the test. Each test session was performed at the same time of day for individual participants. Each session included 9 DVA tests with different motion types and velocities, with a 2-minute interval between each test. The test sequence was the same across 3 sessions, in the following order: horizontal motion of 20 degrees per second (dps), horizontal motion of 40 dps, horizontal motion of 80 dps, vertical motion of 20 dps, vertical motion of 40 dps, vertical motion of 80 dps, diagonal motion of 20 dps, diagonal motion of 40 dps, diagonal motion of 80 dps. Each session took about 30 minutes. 

Figure 1.

 

The study flow chart. dps, degrees per second.

Statistical Analysis

The statistical analysis was performed with SPSS (version 26.0). The Kolmogorov-Smirnov test was performed to determine the normality of the distribution. Descriptive statistics were reported as mean and standard deviation (SD) for continuous variables, whereas number and percentage were for categorical variables. Considering the correlation of repeated tests obtained from the same participants, a linear mixed model was applied to include random intercept when comparing the DVA of different motion types, velocities, and sessions. Multiple comparison corrections were conducted using the Bonferroni method. The short-term learning effect of DVA was calculated as the result of the second session minus the first session, and the long-term learning effect was calculated as the result of the third session minus the first session. The relationship between the learning effect and DVA of the first session was analyzed using linear regression. The participants were given a survey on sports and duration prior to the test. Subgroup analysis was performed by gender and sports experience, and a single-factor linear model was applied to compare the learning effect between subgroups. The sports experience was defined as “Yes” if the participant exercised regularly, and the static sports were excluded, including anaerobic fitness and yoga. A P value of < 0.05 was denoted as statistically significant. 

Results

Of the 58 enrolled individuals who completed 3 test sessions, the mean age was 23.1 ± 2.1 years, and men accounted for 46.6% of the participants. The mean binocular spherical equivalence was −2.42 ± 1.85 D. 

DVA of Different Motion Types and Velocities

The DVA of different motion types and velocities is demonstrated in Table 1. For the baseline measurement, the DVA were significantly varied among different velocities in horizontal (P < 0.001), vertical (P = 0.005), and diagonal motion (P < 0.001). The DVA at 20 dps was significantly better than that at 40 and 80 dps in different motion types (P < 0.05, respectively). No significant difference was observed between 40 dps and 80 dps DVA in horizontal (P = 0.075) or vertical (P = 1.000), but not diagonal motion (P < 0.001). 

Table 1.

 

The DVA of Different Motion Types and Velocities

 

When compared among different motion types, the DVA of horizontal motion was significantly better than vertical (P < 0.001) and diagonal (P < 0.001) motion at the velocity of 20 dps. There was no significant difference between vertical and diagonal motion DVA at 20 dps (P = 0.286). At 40 and 80 dps, the DVA of diagonal motion was worse than horizontal motion (P < 0.001 for 40 dps and P = 0.001 for 80 dps) and vertical motion (P < 0.001 for 40 dps and P = 0.003 for 80 dps). No significant difference was observed between horizontal and vertical motion DVA at 40 (P = 0.614) or 80 dps (P = 1.000). 

Except for the vertical motion DVA test, the results for the short- and long-term repeat measurements were similar to the baseline measurement. There was no significant difference among DVA of different velocities for vertical motion during short- (P = 0.288) and long-term (P = 0.088) repeat measurements. 

DVA of Different Sessions

The DVA of different sessions is compared and demonstrated in Figure 2. There was a significant short-term learning effect for 20 (P = 0.004), 40 (P < 0.001), and 80 (P = 0.014) dps DVA test of horizontal motion, but no significant long-term learning effect was observed (P > 0.05 for all 3 velocities). For the vertical motion, there was a significant short-term learning effect for the 40 dps DVA test (P = 0.003) but not for the 20 (P < 0.001) or 80 dps (P = 0.460) DVA test. The long-term learning effect was not detected for the vertical motion DVA test (P > 0.05 for all 3 velocities). For the diagonal motion DVA test, short- (P = 0.036) and long-term (P = 0.015) learning effects were shown in the 40 dps test. There was no significant short- or long-term learning effect in the 20 or 80 dps diagonal motion DVA test (P > 0.05, respectively). 

Figure 2.

 

Baseline, short-, and long-term repeated measurements for dynamic visual acuity (DVA) of difference motion types and velocities. (A) Horizontal motion of 20 dps. (B) Horizontal motion of 40 dps. (C) Horizontal motion of 80 dps. (D) Vertical motion of 20 dps. (E) Vertical motion of 40 dps. (F) Vertical motion of 80 dps. (G) Diagonal motion of 20 dps. (H) Diagonal motion of 40 dps. (I) Diagonal motion of 80 dps. *P < 0.05. dps, degrees per second.

Short- and Long-Term Learning Effects

The histogram of short- and long-term learning effects is demonstrated in Figure 3. For the second session test of horizontal motion, 77.6%, 70.7%, and 70.7% of the participants had the same or better 20, 40, and 80 dps DVA than the first session test. Similarly, 72.4% of participants performed the same or better in 40 dps vertical motion DVA in the second session than the first session test, and the number was 62.1% for the 40 dps diagonal motion test. 

Figure 3.

 

Histogram for short- and long-term learning effect of different motion types and velocities. Short-term learning effect for horizontal motion of 20 dps (A), horizontal motion of 40 dps (B), horizontal motion of 80 dps (C), vertical motion of 20 dps (D), vertical motion of 40 dps (E), vertical motion of 80 dps (F), diagonal motion of 20 dps (G), diagonal motion of 40 dps (H), and diagonal motion of 80 dps (I). Long-term learning effect for horizontal motion of 20 dps (J), horizontal motion of 40 dps (K), horizontal motion of 80 dps (L), vertical motion of 20 dps (M), vertical motion of 40 dps (N), vertical motion of 80 dps (O), diagonal motion of 20 dps (P), diagonal motion of 40 dps (Q), and diagonal motion of 80 dps (R). dps, degrees per second.

The linear regression analyzing the relationship between the learning effect and baseline DVA result is demonstrated in Figure 4. Short- and long-term learning effects for horizontal motion tests were positively associated with baseline DVA (P < 0.05 for all the velocities). The short-term learning effect was positively associated with baseline DVA at 20 (P = 0.001), 40 (P < 0.001), and 80 dps (P < 0.001) DVA test of vertical motion. For the long-term learning effect in the vertical motion, it is positively related to baseline DVA at 20 (P = 0.003) and 80 dps (P < 0.001) but not 40 dps (P = 0.307). The results of diagonal motion were similar to the vertical motion DVA test, which demonstrated a positive correlation between short-term learning effect and baseline DVA at 20 (P < 0.001), 40 (P = 0.010), and 80 dps (P = 0.002), and between long-term learning effect and baseline DVA at 20 (P < 0.001) and 80 dps (P = 0.001). 

Figure 4.

 

Linear regression showing the short- and long-term learning effect among different dynamic visual acuity. (A) Horizontal motion of 20 dps. (B) Horizontal motion of 40 dps. (C) Horizontal motion of 80 dps. (D) Vertical motion of 20 dps. (E) Vertical motion of 40 dps. (F) Vertical motion of 80 dps. (G) Diagonal motion of 20 dps. (H) Diagonal motion of 40 dps. (I) Diagonal motion of 80 dps. dps, degrees per second.

The subgroup analysis of the learning effect by gender and sports experience is demonstrated in Table 2. The short (P = 0.031) and long-term (P = 0.024) learning effect of male participants was significantly greater than that in female participants in the 80 dps horizontal motion DVA test. The short- or long-term learning effects did not differ between genders for other motion types (P > 0.05, respectively). The existence of sports experience did not affect the short- or long-term learning effect in horizontal, vertical, and diagonal motion DVA test of different velocities (P > 0.05, respectively). 

Table 2.

 

Short- and Long-Term Learning Effects of DVA Test in Subgroups by Gender and Sports Experience

 

Discussion

DVA is a promising indicator for evaluating real-life functional vision in the clinical setting. The learning effect exists in almost all visual tasks, which might affect the reliability of the test result for subjective examination. The present research recruited participants with normal corrected static visual acuity. DVA of different motion types and velocities was measured repeatedly in short- and long-term intervals. Among different motion types and velocities, the study showed a significant short-term learning effect in the DVA test, but the long-term learning effect was rarely demonstrated. The learning effect in repeated measurements was associated with the initial DVA results. 

Previous research demonstrated that DVA was related to motion types and velocities.^23,26 The present study performed the DVA test of 3 velocities, including 20, 40, and 80 dps, and found that DVA of 20 dps was better than 40 and 80 dps DVA. The results were consistent with previous research that DVA became worse as the velocity of the optotype increased.^23,26 As the target velocity increases, it is difficult for the human eye to pursue the movement accurately, especially when the velocity exceeds 50 dps.^27,28 As for different motion types, the present research found that horizontal motion DVA was better than vertical and diagonal motion DVA at low speed, and diagonal motion DVA was worse than the other two motions at high speed. The result was consistent with previous research that humans performed better at identifying targets with horizontal trajectory than other motion paths and outperformed in predictable motion compared with unpredictable motion.^29–31 The optotype exposure time is longer during horizontal than vertical motion. The longer exposure time means more time to pursue and identify the optotype. Unlike horizontal and vertical motion, the path of diagonal motion is unpredictable. Participants might require more time to initiate the pursuit and less time to perform the pursuit during the diagonal motion DVA test, especially in the high-speed DVA test. 

Perceptual learning exists in almost all visual tasks, including determining motion direction, orientation, and spatial frequency.^16–18 The learning effect could improve the performance in visual tasks but might affect the accuracy of subjective examination involving visual perceptual tasks. The present research demonstrated significant short-term learning effects in horizontal motion tests of 20, 40, and 80 dps and vertical and diagonal motion tests of 40 dps. The result was consistent with previous research that repeat training in short duration would increase the performance of the horizontal motion DVA test.^21,22 After short interval repeated tests, the performance improvement reflects an improved signal-to-noise ratio in perceptual processing by representation enhancement and information reweighting.¹⁵ In contrast, the long-term learning effect was only observed in the diagonal motion test of 40 dps but not in other combinations of different motion types and velocities. Consistently, previous research demonstrated that horizontal, vertical, and diagonal motion DVA tests had good repeatability when measured repeatedly in 2-week intervals.²³ The short- and long-term processes might contribute to perceptual learning, including between-session gain and forgetting.³² The between-session forgetting could diminish the performance improvement after training cessation. Interestingly, the short-term learning effect varied among different motion and velocity DVA tests, which was detected in horizontal motion regardless of velocity and 40 dps regardless of motion. The motion-specific learning effect might be attributed to the difference in the sensitivity and performance in motion detection to different motions. The difference in the velocity-related learning effect might be due to task difficulty, which is a crucial influential factor for learning effect.¹⁵ Future research is required to demonstrate the mechanism difference in the learning effect among the combination of different motion types and velocities. 

Perceptual learning is influenced by many factors, including stimuli, tasks, feedback, and so on.¹⁵ The present research demonstrated that the learning effect was associated with the initial DVA result. The results indicate that the participants with worse DVA tend to acquire greater improvement in repeated measurements, regardless of motion types and velocities. The outcome was consistent with previous research on the horizontal motion DVA test, which found that the training effect was more obvious for observers with poorer initial performance.²¹ The research also found that male participants had greater short- and long-term learning effects in the horizontal motion DVA test of 80 dps. The difference might be related to gender-specific disparity in visual perceptual processing, which requires further investigation. 

The present research has significant implications for the DVA test. Averagely, the present found obvious short-term learning effects on repeated DVA tests, especially in horizontal motion tests and high-speed tests. The results imply that a sufficient explanation and pretraining might be required before the DVA test to obtain an accurate outcome. A repeated test might be required for participants with inferior DVA in an initial test. In the current test, the pretraining was simple, which presented optotypes of fixed size and velocity consecutively. The protocol for pretraining, including the optotype and passing standard, should be further investigated to avoid the learning effect in formal tests. The long-term learning effect was not marked in the present study, which indicates the reliability of the DVA test. Additionally, the learning effect of the DVA test found in the present research indicates the potentiality of plasticity and trainability in identifying details of moving objects. For subjects with worse than normal DVA, repeated training might improve the DVA in a temporal or sustained way, which could improve the participants’ quality of life. More research is required to explore the training protocol that could optimize the training effect. 

Certain limitation exists in the present research. First, only young participants with corrected to normal static vision were included. Age significantly affects DVA²⁵ and might also influence the learning effect of the DVA test. Future research on perceptual learning should include participants of different ages and ocular health. Second, the sequence of DVA test motion types and velocities was fixed in repeated measurements. As for visual tasks, there is usually some transfer of learning to correlated stimuli and tasks, which depends on task difficulty, training process, adaptation, and so on.¹⁵ The transfer learning might affect the result of subsequent tests after the initial horizontal motion DVA test, although an interval was applied between tests. Third, the mechanism of short- and long-term learning has not been investigated in the present research, and more studies are required to manifest the contribution of between-session gain, forgetting, and adaptation in perceptual learning. 

The present research enrolled normal participants and repeatedly performed DVA tests of different motion types and velocities in short- and long-term duration. The study found significant differences in DVA among different motion types and velocities. There was a significant short-term learning effect in the DVA test of horizontal motion and 40 dps vertical and diagonal motion, but the long-term learning effect was not observed, except for the diagonal motion of 40 dps. The learning effect in repeated measurements was greater in participants with worse initial DVA. The present research provides the basis for optimizing the DVA test protocol to provide more reliable results in clinical settings and implies the potentiality of improving DVA through repeated training. More research is required to guarantee the proper avoidance and application of perceptual learning during DVA tests in clinical and training settings. 

Acknowledgments

Supported by the National Natural Science Foundation of China (82201243) and Beijing Science and Technology Plan Project (Z221100005222031). 

Data Availability Statement: The data that support the findings of this study are available from the corresponding author upon reasonable request. 

Disclosure: X. Wang, None; M. Yan, None; J. Li, None; Y. Wang, None