Amblyopia and Stereoptic Games for Vision (AMBER)

Stereoptic Serious Games as a Visual Rehabilitation Tool for Individuals With Residual Amblyopia: A Crossover Randomized Controlled Trial - AMBlyopia and stEReoptic Games for Vision

Amblyopia is the most common developmental vision disorder in children, caused by abnormal visual experience in early life, especially a difference in refraction between the eyes, a misalignment of the eye axes, a combination of both. Besides a significant reduced visual acuity, the patients show deficits in 3D vision and functional vision impairment like reduced reading speed, selective attention or motor control skills. The gold standard treatment for amblyopia, occlusion therapy, can lead to relapses or residual amblyopia (i.e., amblyopia that persist into adulthood). The current study aims to test the efficacy of a novel home-based child-friendly rehabilitation program for amblyopia, Vivid Vision Home, involving playing serious videogames on a virtual reality (VR) headset at home. The VR training will be compared to standard care (wearing a spectacle correction) and to age-matched typically developing individuals. A cross-over design will be used so that each participant will receive both treatments (games, spectacles). There is a growing number of studies showing benefits of binocular stimulation for visual acuity and stereovision, but provided mainly by level III studies, with a need for rigorous level I or II studies, using more engaging therapies, to confirm or refute the efficacy of this approach as an adjunct or replacement for current amblyopia treatments. Embedding binocular stimulation in engaging, immersive serious games delivered on VR headsets at home, as implemented by Vivid Vision Home, can address this question.

Amblyopia is the most common developmental vision disorder in children, affecting 1-5% of the population in developed countries. It mostly results from a difference in refraction between the eyes (anisometropia), a misalignment of the eye axes (strabismus), a combination of both, or visual deprivation (due to congenital cataract, e.g.). Besides a significant reduced visual acuity, the patients show binocular dysfunction leading to functional vision impairment like reduced reading speed, selective attention or motor control skills. Whilst the disorder is commonly diagnosed around the age of 3-5 years, up to 50% of children will be left with residual amblyopia mainly due to a late diagnosis and start of treatment, poor compliance with treatment, not diagnosed or considered eccentric fixation. The gold standard treatment consists of the occlusion therapy, patching the dominant/healthy eye for 2-6 hours/day every day for several months up to years. However, an eye patch and the demand to use the weaker eye for visual tasks is challenging and so poorly complied with by pediatric patients. These problems in compliance lead to relapses (14-25%) or residual amblyopia (i.e., poor vision that persist into adulthood) creating multiple medical and social problems for the patients (including cognitive and emotional processing), their families and the society. Serious videogames delivered on tablets have been developed with a more engaging strategy to effectively treat amblyopia. These games focus on binocular stimulation, which targets visual acuity of the amblyopic eye and three-dimensional vision (stereovision), through the presentation of dichoptic images. Such dichoptic image presentation consists of showing a different but complementary image in each eye so that the task can only be performed if information across eyes is combined. Binocular stimulation signifies that both eyes are challenged but the devices are programmed in a way so that the weaker eye is primarily performing a given visual task. The objective of this project is to improve visual function of the amblyopic eye using image-fusion and stereopsis through playing videogames in a Virtual Reality (VR) environment. Also, besides the improvements of visual acuity and potentially stereovision, such VR-based serious games interventions can positively impact other aspects like cognitive and motor functions. The visual deficits seen in amblyopia are thought to be related to problems in attending to visual task-relevant objects, suggesting that such deficits in selective-attentional skills can hamper vision recovery. Similarly, poor stereovision will be reflected in deficits in motor control skills as seen in individuals with amblyopia. Our VR-based serious games training will have similar ingredients to those that are supposed to improve attentional skills and may have a positive impact on motor control skills. The VR-based technology developed by Vivid Vision to be used in the present study is FDA approved and CE (Conformité Européene) certified. They are offering virtual reality videogames with headsets, which may increase the beneficial treatment effect by a reduction of the misperception of 3-dimensional movement. The present project aims to test the described benefit in a home-based, child-friendly rehabilitation program involving serious video games in a virtual reality environment in children, adolescents, and young adults with residual amblyopia, compared to standard care (refractive correction) and to age-matched healthy individuals. The study has been designed as a blinded randomized cross-over trial to increase the available evidence allowing for a qualified evaluation of the benefit of binocular stimulation and for demonstrating a beneficial effect also in residual amblyopia in childhood and in older patients.

The optometrists evaluating visual function will be blinded to the current treatment and the treatment order

Participants with amblyopia receiving the experimental intervention consisting of serious videogames focusing on binocular function (with image modification) in a virtual reality environment.

Participants with amblyopia receiving the standard care intervention consisting of wearing spectacles with individualized refractive correction.

Participants without amblyopia or other conditions receiving the experimental intervention consisting of serious videogames focusing on binocular function (without any image modification) in a virtual reality environment.

The study intervention consists of playing serious games (Vivid Vision, San Francisco, USA) embedded in a virtual reality headset in a home environment 5 days a week for 30 minutes over 8 weeks (20 h of total gaming).

The control intervention will be refractive error correction that consists of wearing the lenses with the prescribed correction for 2 months.

Best-corrected visual acuity refers to the measurement of the best vision correction that can be achieved using glasses or contact lenses.

Best-corrected visual acuity refers to the measurement of the best vision correction that can be achieved using glasses or contact lenses.

Best-corrected visual acuity refers to the measurement of the best vision correction that can be achieved using glasses or contact lenses.

Best-corrected visual acuity refers to the measurement of the best vision correction that can be achieved using glasses or contact lenses.

Best-corrected visual acuity refers to the measurement of the best vision correction that can be achieved using glasses or contact lenses.

Stereoacuity refers to the the smallest detectable depth difference that can be seen in binocular vision. When binocular vision is present, the binocular function is the best stereoscopic acuity, measured in arc seconds, achieved for any of the below mentioned tests: Lang I and Lang II-Stereotest Titmus test Bagolini striated glasses test TNO-Test (TNO: Nederlandse Organisatie voor Toegepast Natuurwetenschappelijk Onderzoek)

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Stereoacuity refers to the the smallest detectable depth difference that can be seen in binocular vision. Stereoacuity, expressed in arc sec, will be investigated with a novel, 3D tablet-based test called ASTEROID.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Interocular misalignment refers to the degree to which two eyes's axes are not parallel. It can be measured with the Cover test. Result of Cover test of one eye turning upon covering the other indicates eye misalignment. Red filter involves asking patient to fixate on a white circle at the end of the room and placing a red filter on the patient's fellow eye. If the patient reports a red pinkish light, their eyes are aligned, and they do not have strabismus. The location of the red circle in relation to the white circle will tell the examiner about also the type of the strabismus.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Five built-in tests in the Vivid Vision Home software will be performed to delineate the patient's treatment (in the amblyopia cohort) and to adjust the treatment to the potential improvements, by setting parameters to increase or decrease the amount of blur or occlusion needed for the patient to play the games. In the healthy cohort, the same tests will be done, but they rather aim to accommodate the prisms built in the VR headset to the participants' view. A Composite Depth Score estimate is measured (1-30) where 0 indicates no stereovision and 30 - 20 arc sec. Patients need to choose which of 4 circular stimuli are floating off the surface, where with each correct response the stimuli become smaller and the disparity decreases.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Five built-in tests in the Vivid Vision Home software will be performed to delineate the patient's treatment (in the amblyopia cohort) and to adjust the treatment to the potential improvements, by setting parameters to increase or decrease the amount of blur or occlusion needed for the patient to play the games. In the healthy cohort, the same tests will be done, but they rather aim to accommodate the prisms built in the VR headset to the participants' view. Degree of binocular vision is estimated with a virtual Worth 4 Dot test, revealing normal vision, double vision, or suppression of left or right eye.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Five built-in tests in the Vivid Vision Home software will be performed to delineate the patient's treatment (in the amblyopia cohort) and to adjust the treatment to the potential improvements, by setting parameters to increase or decrease the amount of blur or occlusion needed for the patient to play the games. In the healthy cohort, the same tests will be done, but they rather aim to accommodate the prisms built in the VR headset to the participants' view. The minimal correction needed for patient's ocular posture in a horizontal, vertical and rotational prism is estimated in prism diopters through a Maddox rod like test, where patient aligns vertical and horizontal lines with a spot or a pair of horizontal lines

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Five built-in tests in the Vivid Vision Home software will be performed to delineate the patient's treatment (in the amblyopia cohort) and to adjust the treatment to the potential improvements, by setting parameters to increase or decrease the amount of blur or occlusion needed for the patient to play the games. In the healthy cohort, the same tests will be done, but they rather aim to accommodate the prisms built in the VR headset to the participants' view. The speed of patient's ability to switch between difference vergence demands is estimated in seconds as the participant is aligning a series of shapes or symbols until they make up a single line.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

Five built-in tests in the Vivid Vision Home software will be performed to delineate the patient's treatment (in the amblyopia cohort) and to adjust the treatment to the potential improvements, by setting parameters to increase or decrease the amount of blur or occlusion needed for the patient to play the games. In the healthy cohort, the same tests will be done, but they rather aim to accommodate the prisms built in the VR headset to the participants' view. Participant's maximal horizontal and/or vertical vergence ability is estimated in prism diopters as the the participant is aligning a series of shapes or symbols until they make up a single line.

Baseline, every week during experimental treatment, after experimental treatment (2 months) and 2 months follow-up

For this purpose, we will use the MNRead test. The test is administered through an app on an iPad©, electronically recorded (connected to a computer), and is designed to assess reading skills in people with low vision (MNRead, French electronic version, 2016). The MNRead test measures the smallest print readable by the person without making significant errors, as well as the smallest print that the person can read with maximum speed and the maximum reading speed.

Spatial cuing task designed by Folk et al. (1992), adapted to the current project's aims, will be administered separately for the amblyopic/fellow and the dominant eye. Participants search for a target diamond of a predefined color (e.g. blue) in an array of differently colored bars and need to report the bar's orientation (horizontal or vertical) by pressing keyboard keys. On every trial, this search array is preceded by an array where a task-irrelevant visual distractor is present that can have the same or a different color than the target. On 50% of the trials the cues are accompanied by trials. Selective attention is measured behaviorally by cuing effects, i.e., the difference in speed of responding when the cue and target are in the same versus different locations. Strength of visual selective attention is measured by the the difference in cuing effects elicited by distractors that matched the color of the target compared to distractors that matched a different color.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Traditional EEG processes of visual selective attention is measured by the N2pc event-related potential (ERP)component, a traditional marker of visual selective attention. The N2pc is a negative-going voltage deflection observed app. 200-300ms after presentation of the stimulus of interest, larger over electrodes contralateral than ipsilateral to the side of the stimulus. Strength of visual selective attention here is measured by the difference in the mean amplitude in the N2pc time-window elicited by distractors that matched the colour of the target compared to distractors that matched a different colour.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Topographic analyses of the EEG/ERPs focus on the reference-independent, multivariate characteristics of the electrical field across the whole scalp. Here, the topographic analyses involve using clustering of the group-averaged EEG/ERP activity over the time-window of the N2pc (see Outcome 11) to identify periods of stable topographic activity (topographic maps). After identification of an optimal number of the topographic maps, they are fit back into single-subject data; parameters like map duration (map onset, map offset) and global explained variance will be analysed. Strength of visual selective attention here is measured by the difference in the duration of maps present over the N2pc time-window elicited by distractors that matched the colour of the target compared to distractors that matched a different colour.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Global Field Power (GFP) of the EEG/ERPs is a standard deviation of the moment-by-moment voltage of the electrical field across the whole scalp. Here, the GFP analyses are conducted over the time-window of the N2pc (see Outcome 11) to identify if EEG responses across different conditions were modulated by differences in strength of the response the same, statistically indistinguishable brain network. Strength of visual selective attention here is measured by the difference in the GFP present over the N2pc time-window elicited by distractors that matched the colour of the target compared to distractors that matched a different colour.

Audiovisual selective attention will be measured, like the visual attention processes, with the Folk et al.'s spatial cuing task (see Outcome 10). Selective attention is measured by cuing effects, i.e., the difference in speed of responding when the cue and target are in the same versus different locations. Strength of audiovisual selective attention here is measured by the the difference in cuing effects elicited by colour distractors accompanied by sounds compared to colour distractors presented without sounds.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Traditional EEG processes of visual selective attention is measured by the N2pc event-related potential (ERP)component, a traditional marker of visual selective attention. The N2pc is a negative-going voltage deflection observed app. 200-300ms after presentation of the stimulus of interest, larger over electrodes contralateral than ipsilateral to the side of the stimulus. The N2pc has been observed during studies of attention to audiovisual stimuli. Strength of audiovisual selective attention here is measured by the difference in the mean amplitude in the N2pc time-window elicited by colour distractors accompanied by sounds compared to colour distractors presented without sounds.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Topographic analyses of the EEG/ERPs focus on the reference-independent, multivariate characteristics of the electrical field across the whole scalp. Here, the topographic analyses involve using clustering of the group-averaged EEG/ERP activity over the time-window of the N2pc (see Outcome 11) to identify periods of stable topographic activity (topographic maps). After identification of an optimal number of the topographic maps, they are fit back into single-subject data; parameters like map duration (map onset, map offset) and global explained variance will be analysed. Strength of audiovisual selective attention here is measured by the difference in the duration of maps present over the N2pc time-window elicited by colour distractors accompanied by sounds compared to colour distractors presented without sounds.

EEG will be recorded while participants are doing the adapted Folk et al.'s spatial cuing task (see Outcome 10). Global Field Power (GFP) of the EEG/ERP processes is a standard deviation of the moment-by-moment voltage of the electrical field across the whole scalp. Here, the GFP analyses are conducted over the N2pc time-window (see Outcome 11) to identify if EEG responses across different conditions were modulated by differences in strength of the response the same, statistically indistinguishable brain network. Strength of audiovisual selective attention here is measured by the difference in the GFP present over the N2pc time-window elicited by colour distractors accompanied by sounds compared to colour distractors presented without sounds.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. To evaluate the movement duration, we will calculate the time passed from the start to the end of the movement. We will split it into: reaching phase, manipulating phase and withdrawal phase. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. To evaluate the reaction time, we will calculate the time passed from the start of the trial (indicated with a custom script) to the start of the movement. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. To evaluate the smoothness, we will extract the information from the marker placed on the hand and we will calculate its trajectory straightness. The straighther, the smoother the data is, which indicates better motor control. We will split it into: reaching phase, manipulating phase and withdrawal phase. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. To evaluate the maximum grip aperture, we will extract the difference between the positions of the marker on the index and the marker on the thumb. The maximum difference will be used. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. To evaluate the time to maximum grip aperture, we will extract the difference between the positions of the marker on the index and the marker on the thumb. The time at which the maximum difference occured will be used. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. We will use a 128-channel EEG system to record brain activity during this task. We will perform frequency analysis to investigate the power at a specific frequency band between the start signal (custom script) till the start of the movement, time in which the participant should have planned the movement. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Motor planning and execution deficits will be evaluated during a visually guided reach-to-grasp task at different reaching depths using reflective markers attached to the upper limb. We will use a 128-channel EEG system to record brain activity during this task. We will perform frequency analysis to investigate the power at a specific frequency band after the start of the movement, time in which the participant is executing the movement. This will be done for 3 conditions: monocular dominant, monocular non-dominant, and binocular.

Visually evoked potentials (VEPs) will also be measured as an electrophysiological (EEG) index of the integrity (strength) of the visual cortical pathway from the retina to the occipital cortex. VEPs originate from the occipital cortex that receives and interprets visual signals. They consist of a sequence of voltage peaks measured over the occipital electrodes: negative peak (N1), positive peak (P1), negative peak (N2). VEPs will be recorded, for each eye separately, from the responses to the color targets in the selective-attention tasks.

Rasch scores for each questionnaire item will be obtained from published look-up tables available at www.pedig.net, and used to calculate a score for each amblyopic participant (Parent-PedEyeQ for <18-year-olds; adapted Child-PedEyeQ for >18-year-olds) and each treatment arm and group at each visit. Scores will also be converted to a 0-100 scale to aid in interpretation. Healthy individuals or their parents will not complete this questionnaire. Child PedEyeQ: Functional Vision, Bothered by Eyes and Vision, Social, Frustration / Worry . Parent PedEyeQ: Impact on Parent and Family, Worry about Child's Eye Condition, Worry about Self-perception and Interactions, Worry about Functional Vision

Adherence (total amount of training hours, regularity of training) throughout the treatment will be investigated and the visual outcome will be correlated to the adherence that is automatically recorded by the Vivid Vision Home system. By including these measures of adherence to the serious games treatment, we will obtain important information, as the traditional treatment is known to have low adherence.

At the end of the study, the frequency and type of adverse events will be analyzed and compared across groups by the researcher blinded to group allocation.

AMBLYOPIC COHORT Inclusion Criteria: Aged between 6 and 35 years old, Residual amblyopia defined as Best-Correct Visual Acuity (BCVA) of ≤ 20/20 in the amblyopic eye, an impairment of ≥ 2 lines in the amblyopic eye persisting even after refraction correction, Stable BCVA for at least 2 consecutive measurements over 6 months, Signed Informed Consent from the child's legal representative, by both the child and his/her legal representative for participants older than 14 years, or by full-aged participants themselves. Exclusion Criteria: Untreated or newly diagnosed anisometropic, strabismic or combined amblyopia, that is a child with BCVA interocular difference of ≥ 2 lines, Current atropine treatment or atropine treatment 3 months prior to enrolment in the study, Auditory deficits or loss, Eye surgery except those to correct strabismus, Strabismus over 20 dioptres (D) or with large eccentric fixation, Coexistence of ocular or neurological disease (e.g., seizure or epilepsy, incomitant strabismus, nerve palsy, horror fusionis), Developmental delay or disorder (e.g., dyslexia, dyspraxia, attention deficity hyperactivity disorder, autism spectrum disorders), Inability to follow and complete the procedures of the study (e.g., psychological or sensorimotor disorders). HEALTHY INDIVIDUALS COHORT Inclusion Criteria: Aged between 6 and 35 years old, Signed Informed Consent from the child's legal representative, by both the child and his/her legal representative for participants older than 14 years, or by full-aged participants themselves. Exclusion Criteria: Auditory deficits or loss, Eye surgery except those to correct strabismus, Strabismus over 20D or with large eccentric fixation, Coexistence of ocular or neurological disease (e.g., seizure or epilepsy, incomitant strabismus, nerve palsy, horror fusionis), Developmental delay or disorder (e.g., dyslexia, dyspraxia, ADHD, ASD), Inability to follow and complete the procedures of the study (e.g., psychological or sensorimotor disorders).

Individual participant data will be shared upon publication of results in a pseudonymized manner, according to local and european regulations of data protection and open science.

The study protocol and statistical analysis plan will become available after approval of the protocol by the Commission Cantonale d'Ethique de la Recherche sur l'être humain (CCER) ethics board, as a study protocol peer-reviewed article. The data and code will become available upon publication of study results.

Vedamurthy I, Nahum M, Huang SJ, Zheng F, Bayliss J, Bavelier D, Levi DM. A dichoptic custom-made action video game as a treatment for adult amblyopia. Vision Res. 2015 Sep;114:173-87. doi: 10.1016/j.visres.2015.04.008. Epub 2015 Apr 24.

Birch EE, Li SL, Jost RM, Morale SE, De La Cruz A, Stager D Jr, Dao L, Stager DR Sr. Binocular iPad treatment for amblyopia in preschool children. J AAPOS. 2015 Feb;19(1):6-11. doi: 10.1016/j.jaapos.2014.09.009.

Gambacorta C, Nahum M, Vedamurthy I, Bayliss J, Jordan J, Bavelier D, Levi DM. An action video game for the treatment of amblyopia in children: A feasibility study. Vision Res. 2018 Jul;148:1-14. doi: 10.1016/j.visres.2018.04.005. Epub 2018 May 12.

Ziak P, Holm A, Halicka J, Mojzis P, Pinero DP. Amblyopia treatment of adults with dichoptic training using the virtual reality oculus rift head mounted display: preliminary results. BMC Ophthalmol. 2017 Jun 28;17(1):105. doi: 10.1186/s12886-017-0501-8.

Verghese P, McKee SP, Levi DM. Attention deficits in Amblyopia. Curr Opin Psychol. 2019 Oct;29:199-204. doi: 10.1016/j.copsyc.2019.03.011. Epub 2019 Mar 22.

Grant S, Suttle C, Melmoth DR, Conway ML, Sloper JJ. Age- and stereovision-dependent eye-hand coordination deficits in children with amblyopia and abnormal binocularity. Invest Ophthalmol Vis Sci. 2014 Aug 5;55(9):5687-57015. doi: 10.1167/iovs.14-14745.