Video-game play induces plasticity in the visual system of adults with amblyopia

Abstract

Abnormal visual experience during a sensitive period of development disrupts neuronal circuitry in the visual cortex and results in abnormal spatial vision or amblyopia. Here we examined whether playing video games can induce plasticity in the visual system of adults with amblyopia. Specifically 20 adults with amblyopia (age 15-61 y; visual acuity: 20/25-20/480, with no manifest ocular disease or nystagmus) were recruited and allocated into three intervention groups: action videogame group (n = 10), non-action videogame group (n = 3), and crossover control group (n = 7). Our experiments show that playing video games (both action and non-action games) for a short period of time (40-80 h, 2 h/d) using the amblyopic eye results in a substantial improvement in a wide range of fundamental visual functions, from low-level to high-level, including visual acuity (33%), positional acuity (16%), spatial attention (37%), and stereopsis (54%). Using a cross-over experimental design (first 20 h: occlusion therapy, and the next 40 h: videogame therapy), we can conclude that the improvement cannot be explained simply by eye patching alone. We quantified the limits and the time course of visual plasticity induced by video-game experience. The recovery in visual acuity that we observed is at least 5-fold faster than would be expected from occlusion therapy in childhood amblyopia. We used positional noise and modelling to reveal the neural mechanisms underlying the visual improvements in terms of decreased spatial distortion (7%) and increased processing efficiency (33%). Our study had several limitations: small sample size, lack of randomization, and differences in numbers between groups. A large-scale randomized clinical study is needed to confirm the therapeutic value of video-game treatment in clinical situations. Nonetheless, taken as a pilot study, this work suggests that video-game play may provide important principles for treating amblyopia, and perhaps other cortical dysfunctions.

Trial registration: ClinicalTrials.gov NCT01223716.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Figure 1. Consort flow diagrams.

This research project was commenced in late 2004 and completed in early 2009. The first author (RWL) was responsible for conducting clinical procedures in screening patients and assigning participants to interventions. Participants were pseudo-randomly allocated into three intervention groups. The first 10 enrolled patients participated in the action videogame group (MOH), the subsequently enrolled three patients participated in the non-action videogame group (SIM), and then another seven patients were recruited in the crossover intervention group (phase 1: occlusion therapy; phase 2: video game therapy, “joypad” symbol  =  MOH or SIM). Note that the subject allocation was not based on the clinical characteristics of participants.

Figure 2. Improved visual acuity (VA) with video-game experience.

(A) Action video game. Color coding is used throughout the figures to represent the type of amblyopia. Red, strabismic; green, anisometropic; Blue, mixed (strabismic & anisometropic); dark purple, mixed (strabismic & deprivation). Error bars: one s.e.m. (here and in all subsequent figures). (B) Non-action video game. In this experiment, participants were required to play a non-action video game (“chess” symbol: SIM) in the first 40 h and an action video game (“gun” symbol: MOH) in the second 40 h. Note that given the small sample size, the fitting curve is provided here for reference. (C) Control experiment. Another group of participants was required to first undertake occlusion therapy (OT, “patch” symbol) for 20 h, and then continue to the video-game phase (“joypad” symbol: MOH or SIM). Note that SB3 was not available to finish the complete course of video-game training. (D) Summary of acuity data. (Top left) A schematic logMAR letter chart. Each 0.1 logMAR represents 1 letter-line. Parentheses: Snellen acuity. (Top right) The visual acuity data from panels a–c are pooled together to calculate the mean data. (Bottom left) Percent improvement is replotted as a function of baseline visual acuity. Solid symbols: crowded acuity. Open symbols: uncrowded acuity. (Bottom right) Effect of video-game experience on visual crowding. Shaded area: decreased visual crowding.

Figure 3. Improved positional acuity with video-game experience.

(a) Position discrimination. The visual task was to pick the misaligned pair of Gabor patch groupings out of three choices (top, middle, or bottom) . Each grouping consisted of 8 Gabor patches. Positional noise to the Gabor patches was introduced by varying their vertical positions according to a Gaussian distribution function. (b) Percent improvement in positional acuity as a function of baseline positional acuity (zero noise). Each data point represents the mean improvement across different noise levels. (c) Effect of video-game experience on sampling efficiency. (d) Effect of video-game experience on internal noise. (e) Threshold versus noise (TvN) function. Three different neural mechanism signature profiles are illustrated. SB2: TvN function shifts downward (increase in efficiency). SA5: The knee point of TvN function shifts downward and to the left (decrease in internal noise). SS3: combination of both.

Figure 4. Improved spatial attention with video-game experience.

(a) Visual counting. A number (N = 1–10 dots) of black circular dots was presented for 200 ms against a gray background. The target stimulus was then followed by a checkerboard pattern for another 100 ms. Observers were asked to enumerate the number of dots as quickly and accurately as they could. No feedback was given. Note that the dot size was scaled with visual acuity, and therefore the dots displayed on the screen were very visible. (b) Counting threshold. Non-amblyopic eye (NAE) versus amblyopic eye (AE). (c) Percent improvement of counting threshold in the amblyopic eye after video-game intervention. SIM: n = 4 (dotted circles). MOH: n = 10. (d–e) Subgroup analysis—Undercounting. (d) Number of dots reported as a function of number of dots displayed. (e) Counting threshold calculation. An arrow indicates an increase in counting threshold. (f) Response latency as a function of number of dots.

Figure 5. Improved stereoacuity in anisometropic amblyopia with video-game experience.

(a) Stereoacuity as a function of video-game hours. The normal stereoacuity range is 20–40 arcsec. Dotted line: the lower measuring limit of the stereo test plates. Note: JS failed the test in the baseline session; her initial data point is thus arbitrarily set to 400 arcsec (the upper measuring limit of the test plates). (b) Stereoacuity data were replotted in terms of percent improvement.