Virtual reality (VR) has been described as an “interactive computer-generated environment.”1 A VR set-up usually consists of a computer system to generate virtual imagery, a visual display and a variety of interactive devices. Visual displays, however, can be used on their own to view non-computer generated, non-interactive imagery. A popular device for viewing virtual imagery today is the head-mounted display (HMD). Since their introduction in the 1960s,2 HMDs have advanced technologically, improving in display quality and ergonomics. Their applications include medical training,3–5 engineering,6 military aviation,7,8 and an expanding entertainment market.9 However, there have been reports of adverse side effects on humans after viewing virtual imagery with HMDs. These examples include ocular and non-ocular symptoms,10–13 physiological changes,14 and oculomotor changes.11,15–17 Inconsistencies in the findings of previous studies investigating the effects of HMD use can be attributed in large part to differences in the optical and ergonomic characteristics of the individual HMDs assessed, the test batteries administered, the viewing times, and the presence or absence of an interactive component to the imagery viewed.
HMDs can be categorized as being either monocular, biocular (non-stereoscopic), or binocular (stereoscopic). Monocular HMDs present imagery to only one eye and are commonly used in military aviation.7,8 Biocular HMDs present identical imagery to both eyes. The lack of disparity in the imagery results in constant accommodative and vergence demands, but these are not always matched. In binocular HMDs, different images corresponding to the left and right eye views, are relayed to the appropriate eye through separate channels.9 The images are focused at a fixed plane providing a constant accommodative demand. However, vergence demands change as disparity cues in the imagery change. The majority of commercially available HMDs are either binocular or biocular with most studies having used the binocular type. Few studies have investigated the effects of viewing non-stereoscopic imagery with HMDs and the results from these studies have been inconsistent.11,17–24 Although the conflicting accommodative and vergence demands present in stereoscopic displays have been proposed as a cause of adverse oculomotor or symptom responses,16 several attributes of biocular HMDs may also affect user comfort. Such attributes include the prismatic effect of high powered lenses used to focus the imagery,25 poor display quality,8 screen proximity,26 and poor ergonomics.14 In addition, the presence of monocular depth cues in the non-stereoscopic imagery of biocular HMDs may lead to accommodation or vergence-related changes following viewing.
There is also a lack of knowledge on the effects of viewing virtual imagery with HMDs on the visual and oculomotor systems of children. It is known that before the ages of 6 to 8 years, the visual system undergoes a “critical period” during which time it is plastic and vulnerable to change imposed by abnormal visual input.27–29 The oculomotor system is also amenable to change during childhood, however, it is unclear if plasticity is greater in children than in adults.30–32 This is of particular concern as children are a growing market for HMD use. The reducing costs of production will make HMDs more ubiquitous and accessible to the general public, including children, and improvements in the display quality and ergonomics of HMDs will also increase their appeal. In addition, a growing application of HMD use is for entertainment, where children are a likely target. The HMD investigated in this study, the Binocular Viewer (MicroOptical, Westwood MA), is a new generation HMD that incorporates many improvements on the features of earlier generation HMDs, including improved ergonomics and display quality. Despite the presence of these upgrades, the lack of knowledge on the effect of viewing with biocular HMDs on the human oculomotor system, especially the developing visual systems of children, warrants investigation.
The primary aim of this investigation was to evaluate symptoms and oculomotor changes after viewing with a new generation bi-ocular HMD for 30 min, in children aged 5 to 16 years, and comparing these changes with those that occur after viewing with a high definition television (HDTV) display. It is envisaged that viewing with the Binocular Viewer will not lead to any more discomfort or oculomotor effects than viewing with the HDTV. The secondary aim was to determine if the effect of an extended viewing time of 80 min led to greater symptom or oculomotor changes than those measured after 30 min of viewing.
The Binocular Viewer is a new generation biocular HMD. It represents a significant improvement over past generations of HMDs in ergonomics and display quality with a light weight of 58 g and resolution of Snellen equivalent 6/11. Subjects viewed imagery with the Binocular Viewer for 30 min. In the control condition, subjects viewed 30 min of different imagery projected onto a wall in non-stereoscopic mode to simulate a HDTV. Subjects, in the 13 to 16 years age group, returned for a third viewing session to determine the effects of viewing imagery with the HMD for an extended period of 80 min.
The subjects consisted of 60 children, aged from 5 to 16 years. There were 29 males and 31 females with a mean age of 11.0 years. To ensure an even distribution of ages, the 60 children were divided into three age ranges (5 to 8 years, 9 to 12 years, and 13 to 16 years) with 20 subjects recruited into each of these age groups. All subjects had at least 6/9.5 unaided visual acuity (VA) in each eye, stereopsis >60 sec arc, normal binocular vision, no ocular pathology and were functionally emmetropic according to the following criteria: −0.25 to +0.75 dioptres (D) mean spherical equivalent, astigmatism <0.75 D, and anisometropia <0.75 D mean spherical equivalent. None of the subjects suffered from migraines or epilepsy. This investigation was approved by the Institutional Human Ethics Committee.
A modified cross over experimental design was used and involved two variables: viewing condition and imagery presented. The subjects were first divided into two groups according to the viewing condition; one group viewed 30 min of imagery with the HMD at the first session and with the HDTV at the second session. The order was reversed for the other group. The two groups were balanced for age, gender, and the preliminary oculomotor measures as shown in Table 1. Thirty minutes of viewing virtual imagery was in keeping with previous studies where viewing times of 20 min,10,12,22 25 min,33,34 and 30 min21 have been used. In addition, each viewing condition group was further divided into two subgroups according to the imagery presented, with fifteen in each subgroup. The first subgroup in each group viewed the two sets of imagery at the two viewing sessions in a particular order. This order was reversed for the second subgroup to reduce the effect of treatment-period interactions. The sets of imagery used were animated and were similar in length and differed only in the story (see “Imagery” section). The two sessions were administered at least 1 week apart.
The 13- to 16-year-old subjects participated in a third viewing session of 80 min duration using the HMD to determine if an extended viewing session resulted in any greater symptom or oculomotor changes. The duration of many animated movies is approximately 80 min and this represents a typical length of time which children may use the HMD. Subjects younger than 13 years of age were not used in the extended viewing session as it was envisaged that fewer subject withdrawals would occur with older children. Thus, a total of two sessions were completed by subjects aged 5 to 12 years, and three sessions for subjects aged 13 to 16 years. The third session was administered at least 1 week after the second session.
An oculomotor test battery and symptom questionnaire were administered to each subject at three time points in each session: immediately before viewing, immediately after viewing, and 10 min after viewing, as previous studies have shown a rapid dissipation of oculomotor measures and symptoms after viewing virtual imagery with a HMD.11,35,36 The oculomotor test battery and symptom questionnaire were designed to take no longer than 10 min to administer.
Experimental Condition: Binocular Viewer
Subjects viewed non-stereoscopic virtual imagery with a Binocular Viewer which had a field of view of 12° (diagonal), resolution of 320 × 240 pixels (Snellen equivalent of 6/11), refresh rate of 60 Hz and weight of 58 g. The inter-ocular distance (IOD) of the lenses used to focus the virtual imagery had three fixed settings and could not be adjusted. The smallest size HMD (IOD 53 mm) was used for subjects with a inter-pupillary distance (IPD) of less then 58 mm, the medium size (IOD 58 mm) was used for subjects with an IPD from 58 to 62 mm, and the largest size (IOD 63 mm) was used for subjects with a distance IPD >62 mm. The image focus plane of the HMD (all sizes) was 1 m and could not be adjusted. The mean luminance level at the subjects’ eyes was 61.0 ± 48.0 cd/m2 and ranged from 5 to 208 cd/m2 according to the imagery viewed. Unlike previous VR studies10,37 head-tracking was not used in the current investigation, thus the Binocular Viewer was likened to a “personal viewing system.”18
Control Condition: HDTV
Subjects viewed imagery projected onto a featureless white wall using an Eiki LC-7000 multimedia projector (Eiki Industrial, Osaka, Japan). The projected imagery was non-stereoscopic, with a refresh rate of 60 Hz, resolution of 860 lines and a diagonal size of 68 cm which simulated commercially available HDTVs at the time of the study. The projected imagery was viewed at 3.2 m (diagonal field of view of 12° and Snellen equivalent of 6/4.2). The luminance level at the subjects’ eyes was 30.0 ± 29.8 cd/m2 and ranged from 1 to 105 cd/m2 according to the imagery viewed.
The imagery viewed in both viewing conditions was animated, dynamic, and age-appropriate (rated “G” for general exhibition). To ensure the imagery was interesting for subjects to view, different imagery was shown to the three age groups (5 to 8 years, 9 to 12 years, and 13 to 16 years) for each viewing condition. The imagery existed on digital versatile disc (DVD) format and was commercially available at the time of the study. The DVDs ranged in length from 29.1 to 31.9 min. The DVD viewed by the 13- to 16- year-old subjects in the third, extended viewing session was “Monsters Inc.” This DVD was a full movie length. The DVDs were played on a Pioneer DV-355-S DVD player.
Oculomotor Test Battery
The test battery consisted of eight oculomotor tests that were administered in the following order: distance unaided VA, near unaided VA, stereo-acuity, near point of convergence, distance heterophoria, near heterophoria, response accommodative convergence/accommodation (AC/A) ratio, and the convergence accommodation/convergence (CA/C) ratio.
Lea symbols were used to assess the vision of subjects aged from 5 to 8 years. Lea symbols consist of four picture optotypes (circle, square, apple, and house) that are readily recognized by preschool and young children and are well standardized.38–40 Sloan letters were used for subjects aged from 9 to 16 years and were of a 5 × 5 non-serif format.41 Both Lea symbols and Sloan letters were in logMAR format, i.e., symbol/letter sizes progressed with a constant ratio of 0.1 log units, and both distance and near unaided VA were assessed binocularly. The Lea and Sloan distance charts were internally illuminated, high contrast (>85%) and positioned at 3 and 4 m, respectively. Both Lea and Sloan near vision charts were held at 40 cm. The 9- to 16-year-old subjects responded verbally whereas the 5- to 8-year-old subjects had a choice of responding verbally or by pointing to one of the four optotypes presented on a nearby card. Vision was recorded in logMAR format.
Stereo-acuity was assessed with the Randot Stereotest (Stereo Optical, Chicago, IL). The Randot Stereotest (test distance 40 cm) consisted of 10 groups of three circles with one circle in each group having crossed disparity. The amount of crossed disparity differed between groups and ranged from 400 to 20 sec of arc. Each subject viewed the targets while wearing cross-polarized filters and was asked to identify the circle with crossed disparity. A cardboard sheet with a rectangular hole allowed each group of circles to be presented individually while occluding all other groups of circles on the test page. The groups of circles were presented randomly to prevent recall of correct answers based on order of presentation.
Near Point of Convergence
The near point of convergence was assessed using an objective push-up convergence test. The subject was asked to maintain fixation on a small target while it was moved along the midline of their head toward their ocular plane. The target was a picture located on an apparatus known as a “budgie stick” (a double-sided paddle containing letter optotypes, text and pictures including that of a budgie). To maintain the interest and fixation of the younger subjects as the budgie stick was moved closer, a question was asked about the target such as “what color is the budgie’s head?” As the target was moved closer, an objective assessment of the break in bifixation was made. The distance at which a break in convergence was detected was measured using a tape measure held to the bridge of the nose and attached to the budgie stick. This was recorded as the near point of convergence.
Distance and near heterophoria were assessed using the Modified Thorington technique. A red Maddox rod, which removed the stimulus for binocular fusion, was placed in front of one eye. The other eye viewed a tangent scale, with a scale resolution of a half prism dioptre. A red light emitting diode was positioned at the center of the tangent scale and appeared through the Maddox rod as a vertical line orthogonal to the tangent scale. For distance heterophoria, the Maddox rod was placed in front of the right eye and the tangent scale placed at 6 m (and internally illuminated). For near heterophoria, the Maddox rod was placed in front of the left eye with the scale at 50 cm. The eye with the Maddox rod was occluded and subjects were asked to focus on the numbers on the tangent scale. Once the numbers appeared clear, the occlusion was briefly removed and the subject asked to immediately report the position of the red line along the tangent scale.
Response Accommodative Convergence/ Accommodation Ratio
The AC/A ratio quantifies the gain of the accommodation-vergence cross-link and is equal to the change in convergence per unit change in accommodation.42 The vergence feedback loop was opened by placing a Maddox rod in front of the left eye so that only stimulation from the accommodative system, via the accommodative-convergence cross-link, could induce convergence. The Shin-Nippon SRW-5000 (Shin-Nippon Commerce, Tokyo, Japan) infrared autorefractor was used to measure the accommodation responses to different accommodative stimuli and was interfaced with a computer to record measurements online. The SRW-5000 has been shown to be an accurate and reliable autorefractor when tested on adults43 and children aged between 4 and 8 years.44 The accommodative stimulus was altered using five lenses ranging from −1.50 to +1.50 D in increments of 0.75 D, and three measurements of the accommodative response were taken for each stimulus level. The order of presentation of the lenses was randomized. The Modified Thorington technique (tangent scale at 50 cm) was used to measure the change in vergence for each lens.
Convergence Accommodation/Convergence Ratio
The CA/C ratio quantifies the gain of the convergence- accommodation cross-link and is equal to the change in accommodation per unit change in convergence.45 The accommodative feedback loop was opened to allow it to respond solely to input from the vergence system through the convergence-accommodation cross-link. The accommodative feedback loop was opened by having subjects fixate on a spotlight that was created by mounting a 0.2 mm diameter laser-drilled precision pinhole (Ealing Optical, Watford, Herts: Catalogue no. 43-5297) in front of a white light emitting diode. At 50 cm, a 0.2 mm diameter spotlight is a non-accommodative target.24,46 The pinhole was surrounded by a black background and placed at 50 cm in front of the left eye. The vergence stimulus was altered using five prisms that were introduced randomly (in front of the left eye) with powers that were evenly distributed across the subject’s predetermined fusional vergence range. The fusional vergence ranges varied between 11 and 26 prism dioptres (Δ). Once the pinhole of light had been fused, three measurements of the accommodative response of the right eye were taken for each vergence stimulus level using the Shin-Nippon SRW-5000 infrared autorefractor. The same range of prism powers was used in both sessions for each subject.
The symptom questionnaire was based on the VR Symptom Questionnaire (VRSQ) developed by Ames et al.36 to evaluate the change in symptoms in adult subjects after viewing with a binocular HMD. The VRSQ was formulated from questionnaires used in other investigations of the effects of viewing virtual imagery, and contained symptoms that were most frequently reported after viewing. The questionnaire consisted of 13 symptoms, 8 nonocular, and 5 ocular, with an average administration time of 1 min.
The age range (and therefore the range in the level of comprehension) of the subjects used in the current investigation was relatively large, and warranted the generation of questionnaires containing age-appropriate vocabulary. Three variations of the VRSQ were developed and used for the three different age ranges: 5 to 8 years, 9 to 12 years, and 13 to 16 years. The symptom questions for all three age ranges are shown in Table 2. Word lists, rating the familiarity of words to children of different ages, were used to help develop the age-appropriate questions. In addition, three descriptors were used to grade the responses of the 5- to 8-year-old subjects (“no,” “a bit,” and “yes”), whereas four descriptors were used to grade the responses of the subjects in the two oldest age ranges (“no,” “a little bit,” “a fair bit,” and “a lot”). The questionnaire was administered verbally. The alternative response descriptors were presented to the 9- to 16-year-old subjects on a card, but were presented verbally (and in random order) to the 5- to 8-year-old subjects.
One additional question to the VRSQ was included in the current investigation. The question assessed the subjects’ enjoyment of the imagery viewed and five descriptors were used to grade the response. The average time to administer both the postviewing oculomotor test batteries and symptom questionnaires was 7.0 ± 1.5 min.
Prior to subject participation, a parent or guardian was asked to read the information sheet and sign the consent form after an explanation of all experimental procedures and possible consequences. At the first session, a preliminary eye examination was administered by one of the authors (PK). If eligible, each subject was allocated to a viewing condition in their age group to best balance that group’s age, gender and oculomotor measures.
The oculomotor test battery and symptom questionnaire were administered before viewing. The subjects were then seated on a fixed chair facing a wall at 3.2 m. The HMD was fitted to the subject’s head and the room lights switched off. The duration of the viewing session was 30 min, during which the subject could withdraw at any time. Immediately after viewing and at 10-min postviewing, the oculomotor test battery and symptom questionnaire were again administered. The average time between the first and second sessions was 14 ± 2 days. However, because of the unavailability of one subject, this subject’s sessions were separated by 5 days.
Sixteen of the 13- to 16-year-old subjects returned for an additional viewing session using the HMD at least 6 days following the second session. The viewing time was 80 min and the oculomotor test battery and questionnaire were administered at the same time points as the previous two sessions. The average time between the second and third sessions was 55 ± 11 d. There were no subject withdrawals from either the 30- or 80-min viewing sessions.
The change in symptom rating and oculomotor measures from pre- to postviewing were analyzed using the Wilcoxon matched-pairs signed-rank test. Non-parametric statistical techniques were used due to the categorical nature of the symptom data and the non-normal distribution of the changes for several of the oculomotor measures. All normality tests were carried out using GraphPad Prism (Version 3.02; GraphPad Software, La Jolla, CA), which uses the method of Kolmogorov and Smirnov to test for deviations from normality and calculates a p value using the Dallal and Wilkinson approximation to Lilliefors method.47 The four-point symptom rating scale used for the 9- to 16-year-old subjects allowed for a maximum change in symptom rating of three, whereas the maximum change in symptom rating for the 5- to 8-year-old subjects was two because they were given a three-point rating scale. The symptom data for the 5- to 8-year-old subjects were, therefore, analyzed separately from the symptom data of the oldest two age groups.
The changes in symptom rating and oculomotor measures between the viewing conditions and between the 30- and 80-min viewing sessions were also analyzed using the Wilcoxon matched-pairs signed-rank test. The non-parametric statistical tests were performed using Minitab Statistical Software (Release 13.1, Minitab Inc., State College, PA). The response AC/A ratios and the CA/C ratios were calculated using linear regression analysis (GraphPad Prism, Version 3.02; GraphPad Software). Changes in the slopes of the response AC/A and the CA/C ratios were then analyzed in the same manner as the other oculomotor changes, using non-parametric tests. Cross-over effects (treatment-period interactions and period effects) were measured using the Mann-Whitney test.
Effects of Viewing with the Binocular Viewer for 30 min
Experimental results revealed very few differences in the oculomotor or symptom changes between the three age ranges of subjects and any differences were inconsistent. Thus, the oculomotor data for all subjects was pooled. Symptom data for subjects aged from 9 to 12 years and 13 to 16 years were pooled together. However, the symptom data for the 5- to 8-year-old subjects, which used a different symptom rating scale, remained separate. There were no consistent significant differences between the symptom data of the three age groups.
There were no significant differences between the symptom increases following viewing with the HMD and the HDTV at either of the postviewing time points in both age ranges. In addition, there were no early withdrawals from viewing with the HMD; therefore symptoms experienced were not enough to induce any subjects to withdraw from viewing.
Immediately after viewing with the HMD, four symptoms increased significantly above the previewing level in the 9- to 16-year-old subjects: “feeling tired” (p < 0.01), “feeling sleepy” (p < 0.01), “difficulty concentrating” (p < 0.05), and “sore/aching eyes” (p = 0.05). The increases in the symptoms “feeling tired” and “feeling sleepy” were the most commonly reported symptom increases and were reported by 43 and 40% of subjects, respectively. Increases in the symptoms “sore/aching eyes” and “difficulty concentrating” were reported by fewer subjects (18 and 15% of subjects, respectively). At 10-min postviewing, no symptoms were significantly increased. In the 5- to 8-year-old subjects, there were no symptoms that were increased significantly immediately after viewing and only one symptom was increased significantly above the previewing level at 10-min postviewing: “difficulty concentrating” (p < 0.05).
After viewing with the HDTV, two symptoms were increased significantly above the previewing level in the 9- to 16-year-old subjects: “feeling tired” (p < 0.01) and “feeling sleepy” (p < 0.01). Increases in these symptoms were reported by 40 and 35% of subjects, respectively. At-10 min postviewing, “feeling tired” and “feeling sleepy” remained significantly increased and increases were reported by 38 and 25% of subjects, respectively. Two other symptoms also increased significantly above the previewing level: “sore/aching eyes” (p < 0.05) and “eyestrain” (p < 0.05). Twenty-three percent of subjects reported an increase in both of these symptoms. There were no symptoms that increased significantly after viewing with the HDTV in the 5- to 8-year-old subjects.
None of the significant symptom increases were caused by cross-over effects (treatment-period interactions or period effects). The significant symptom increases following viewing with the HMD and the HDTV are summarized in Table 3. For symptoms that increased significantly in the HMD viewing condition, the percentages of 9- to 16-year-old subjects and 5- to 8-year-old subjects that rated either no change or an increase in the symptoms are presented as pie charts in Figs. 1 and 2, respectively.
The most commonly observed oculomotor change during viewing with the HMD was a significant reduction in near unaided VA, which occurred immediately after viewing (p < 0.01) and at 10-min postviewing (p < 0.05). The reduction in near unaided VA immediately after viewing occurred in 67% of subjects and the change ranged from 0.14 logMAR increase to 0.18 logMAR decrease. At 10-min postviewing, near unaided VA was reduced in 47% of subjects. Near unaided VA was also significantly reduced (p < 0.01) immediately after viewing with the HDTV in 45% of subjects. The change ranged from 0.06 logMAR increase to 0.2 logMAR decrease. Box-plots, showing the range of near unaided VA changes for both viewing conditions, are shown in Fig. 3. A significant reduction in distance unaided VA (p < 0.05) also occurred immediately after viewing with the HMD. Distance unaided VA was reduced in 42% of subjects, and the change ranged from 0.08 logMAR increase to 0.2 logMAR decrease. Unlike near unaided VA, the change in distance unaided VA did not remain significantly reduced at 10 min after viewing and was not significantly reduced after viewing with the HDTV. The reductions in near and distance unaided VA did not significantly differ between the two viewing conditions.
The CA/C ratio was significantly increased (p < 0.01) immediately after viewing with the HMD, and remained significantly increased at 10 min after viewing (p < 0.05). The change in the CA/C ratio ranged from 0.028 D/Δ decrease to 0.050 D/Δ increase immediately after viewing and from 0.043 D/Δ decrease to 0.069 D/Δ increase at 10 min after viewing. The repeatability coefficient in the measurement of the CA/C ratio was 0.038 D/Δ. After viewing with the HDTV, the CA/C ratio was not significantly different to the previewing measure. Box-plots, showing the range of changes in the CA/C ratio for both viewing conditions, are shown in Fig. 3. Statistical outliers of the change in CA/C data were removed from analysis. The changes in CA/C ratio did not significantly differ between the two viewing conditions.
Distance heterophoria was significantly shifted in the exophoric direction immediately after viewing with the HDTV (p < 0.05). This exophoric shift occurred in 38% of subjects and was significantly greater than the shift immediately after viewing with the HMD (p < 0.01). The changes, however, ranged from 4Δ exophoric shift to 6Δ esophoric shift. After viewing with the HMD, distance heterophoria was not significantly different to that measured at previewing. Box-plots, showing the range of changes in distance heterophoria for both viewing conditions, are shown in Fig. 3.
Stereo-acuity was significantly reduced immediately after viewing with the HMD (p < 0.01) and with the HDTV (p < 0.01). In the HMD viewing condition, stereo-acuity was reduced in 30% of subjects and the change ranged from 10 sec arc increase to 50 sec arc decrease. In the HDTV viewing condition, stereo-acuity was reduced in 33% of subjects with the change ranging from a 10 sec arc increase to a 100 sec arc decrease. Stereo-acuity was unchanged in 62 and 58% of subjects after viewing with the HMD and HDTV, respectively. Stereo-acuity did not remain significantly reduced at 10 min after viewing in either viewing condition. The changes in stereo-acuity did not significantly differ between the two viewing conditions.
Two oculomotor measures were significantly different from the previewing level, only at 10 min after viewing with the HMD: near point of convergence was significantly reduced (p < 0.05) and the AC/A ratio was significantly increased (p = 0.05). The change in the AC/A ratio ranged from 1.122 D/Δ decrease to 2.209 Δ/D increase at 10 min after viewing. The repeatability coefficient in the measurement of the AC/A ratio was 1.715 Δ/D. Outliers of the change in AC/A data were removed from analysis. The changes in near point of convergence and the AC/A ratio did not significantly differ between the two viewing conditions. The p-values for the oculomotor changes in both viewing conditions are shown in Table 4.
Effects of Viewing for 80 min with the Head-Mounted Display
Immediately after viewing with the HMD, three non-ocular symptoms increased significantly above the previewing level: “feeling tired” (p < 0.01), “feeling bored” (p < 0.05), and “feeling sleepy” (p < 0.01). The increases in “feeling tired,” “feeling bored,” and “feeling sleepy” were reported by 69, 37, and 69% of subjects, respectively. One ocular symptom, “tired eyes,” also increased significantly immediately after viewing and was reported by 37% of subjects. At 10-min postviewing, the symptoms “feeling tired” and “feeling bored” remained significantly increased (p < 0.05) and was reported by 56 and 37% of subjects, respectively.
There were no early withdrawals from viewing with the HMD for 80 min. Thus, viewing for a period of time similar to the duration of many animated movies, did not lead to symptoms that were sufficient to induce any subjects to withdraw from viewing. The significant symptom increases after viewing with the HMD for 80 min are summarized in Table 5. For the symptoms that increased significantly, the percentages of subjects that rated either no change or an increase in the symptoms are presented as pie charts in Fig. 4.
Similar to the 30-min viewing session, the most commonly observed oculomotor change measured immediately after viewing with the HMD for 80 min was a significant reduction in near unaided VA (p < 0.01). The reduction in near unaided VA occurred in 75% of subjects and the change ranged from 0.02 logMAR increase to 0.22 logMAR decrease. Near unaided VA did not remain significantly reduced at 10 min postviewing. A significant reduction in near point of convergence (p < 0.05) also occurred immediately after viewing, however, the change ranged from a decrease in length by 1 cm to an increase in length by 1.5 cm.
Distance heterophoria was significantly shifted in the esophoric direction (p < 0.01) immediately after viewing. This esophoric shift occurred in 56% of subjects and ranged from 0 to 1.5Δ esophoric. At 10 min after viewing, distance heterophoria remained significantly shifted in the esophoric direction (p < 0.05) in 50% of subjects, and ranged from 0.5Δ exophoric to 1Δ esophoric shift.
The significant oculomotor changes measured after viewing with the HMD for 80 min are summarized in Table 5. Box-plots, showing the range of changes in near unaided VA, near point of convergence, and distance heterophoria are shown in Fig. 5.
Comparison of Symptom and Oculomotor Changes Between the 30 min and 80-min Viewing Sessions with the Binocular Viewer
Only one symptom, “feeling sleepy,” increased significantly more (p < 0.05) immediately after viewing with the HMD for 80 min than for 30 min of viewing. There were no significantly different symptom increases between the two viewing durations at 10 min after viewing. In addition, there were no significant differences in the oculomotor changes after viewing with the HMD for the two viewing durations.
The Binocular Viewer was found to be a safe HMD for children, and as comfortable to view with as the simulated HDTV. This HMD’s light weight is likely to account for the few non-ocular symptom increases during viewing. Only a few ocular symptoms increased during viewing. This is most likely due to the high image quality presented by the HMD. The symptoms reported in the current investigation are an improvement on the severity of symptoms reported in previous studies that used older generation binocular10,35 or biocular18 HMDs. In addition, the most commonly reported symptom increases were not significantly different from those in the control HDTV condition.
Despite symptoms being experienced during the 30-min viewing sessions, these symptoms were mild and reported by <50% of subjects. In addition, the symptom increases were transient in nature, dissipating by 10-min postviewing. In a study by Ames,24 subjects aged from 10 to 15 years were able to view 30 min of stereoscopic imagery with a VR6 HMD (Virtual Research, Santa Clara), but when subjected to an extended viewing session, 50% of the subjects could not view for more than 1 h. Ames attributed the withdrawals to the heavy weight of the HMD (821 g). In the current investigation, a similar small number of significant symptom increases occurred in the 80-min viewing session as in the 30-min viewing session, and there were no withdrawals from either session. Given that the average time between the 30 and 80 min sessions was 55 ± 11 d, and different imagery was viewed between the sessions, it is unlikely that an increased familiarity of the use of the HMD was responsible for this observation. Thus, the experience of adverse symptoms may not be cumulative with time. Overall, use of the HMD as a personal display by children for entertainment purposes was not hindered by user discomfort.
Except for the reduction in near unaided VA, which was consistent across both viewing durations the profile of oculomotor changes after 30 and 80 min of viewing were different. In addition, the magnitude of the oculomotor changes was small and in many cases close to the resolution limit of the particular oculo-motor measures.
Given that near unaided VA was the most commonly observed oculomotor change during viewing with the HMD, one can speculate that it may have been caused by an inappropriate accommodative response (either over- or under-accommodation), relative to the virtual imagery focal plane. However, current evidence does not support this proposition. It has been previously reported that a myopic shift (increase) in accommodation can occur when using optical instruments such as microscopes, a phenomenon known as “instrument myopia.”48 In addition, Kotulak and Morse8 observed that accommodation and vergence increased in response to the “perceived nearness” of symbology presented with an aviator HMD. An increase in accommodation great enough to reduce near unaided VA would certainly reduce distance unaided VA and possibly increase the CA/C ratio. The esophoric shift in distance heterophoria, during 80 min of viewing, could also be attributed to an increase in accommodation, since a positive vergence adaptation is unlikely to occur with a reduced near point of convergence. However, the observed significant reduction in near unaided VA and increase in CA/C ratio (10 min following the 30-min viewing session), was not accompanied by a reduction in distance unaided VA. A increased reduction of distance rather than near unaided VA would be expected if accommodation had increased.
In contrast, a reduction in accommodation (in particular a reduction in the accommodative response gradient), during viewing with a biocular HMD has been reported.19 It was speculated that this reduction was caused by sympathetic over-stimulation, because the reduction only occurred in subjects that exhibited symptoms associated with sympathetic stimulation.22 In the current investigation, however, the accompanying oculomotor changes do not support the notion of a reduction in accommodation. A reduction in distance unaided VA would be unlikely, and the directions of the changes in the CA/C ratio in the 30-min viewing session and distance heterophoria in the 80-min viewing session are contradictory to those expected with a decreased accommodation.
It is unlikely that the accommodative response changed during viewing with the Binocular Viewer for several reasons. First, the imagery had a high resolution (Snellen equivalent of 6/11) and, therefore, provided a good accommodative stimulus. Second, the focal plane of the virtual imagery was approximately equal to the level of tonic accommodation in children.30,32,49 Third, the HMD was used as a personal viewing display and was not linked to an interactive VR system. Given that cognitive demand can influence accommodation,50 the lack of any interactivity between the user and the VR system in the current investigation is likely to have contributed to a stable accommodative response during viewing. Previous studies on biocular HMDs17,22 involved a level of interactivity, subjects not only viewed imagery but also performed a task such as playing a game or navigating through a virtual environment. The cognitive processing required to perform these tasks may have contributed to the oculomotor changes observed in those studies. In addition, the lack of interactivity removes the need for head tracking during use of the HMD. This removes the possibility of visual-inertial sensory conflict that may occur with display update lag during head movement.51
It is also unlikely that any adaptation of the vergence system occurred during viewing with the HMD. There were no changes in distance or near heterophoria during viewing for 30 min. Although a distance esophoric shift occurred during the 80-min viewing session, it was not related to any possible prismatic effect induced by the discrepancy between the subject’s IPD and the IOD of the lenses of the HMD. Consistent with previous reports,17,52 no correlation was found between the IPD-IOD mismatch and the change in distance heterophoria. Although the IOD of the HMD was fixed and could not be adjusted, the three HMD sizes were sufficient to cater for children between 5 and 16 years of age. Because heterophoria is partly determined by an accommodative input, changes in accommodation may also affect heterophoria. However, a change in the accommodative response was unlikely to have occurred in the 80-min session given that no changes in near heterophoria, the AC/A ratio or the CA/C ratio were observed. Also, no correlation was found between the changes in distance and near heterophoria, which suggests that accommodative adaptation is unlikely. Thus, the cause of the distance esophoric shift is not clear and may simply be associated with the noise in the measurement of the distance heterophoria. This is a feasible proposition given that noise is the likely reason for other oculomotor changes such as the changes in AC/A ratio and the near point of convergence which were observed only at 10 min after viewing in the 30-min viewing session.
In summary, the Binocular Viewer has been demonstrated to be a good HMD display for use by children. Viewing for 30 min was relatively asymptomatic, resulting in few symptom increases, and these were transient in nature and reported by <50% of subjects. The HMD was able to be used comfortably for up to 80 min and is, therefore, ideal for such applications as watching movies and playing video games. The high image quality and ergonomics of the HMD (particularly its light weight), resulted in few oculomotor changes, most of which were no greater than that observed after viewing with a HDTV. The reduction in near unaided VA was the only oculomotor measure to have consistently changed in both the 30- and 80-min viewing sessions and was the most commonly observed oculomotor change in both these sessions. Further investigation into the cause of the reduction in near unaided VA is warranted.
We thank the support of Essilor International PLC.
Neville A. McBrien
Department of Optometry and Vision Sciences
The University of Melbourne
Melbourne, Victoria 3010, Australia
1. Wilson PN, Foreman N, Tlauka M. Transfer of spatial information from a virtual to a real environment in physically disabled children. Disabil Rehabil 1996;18:633–7.
3. Nichols S, Ramsey A, Cobb S, Neale H, D’Cruz M, Wilson JR. Incidence of Virtual Reality Induced Symptoms and Effects (VRISE) in Desktop And Projection Screen Display Systems. HSE Contract Research Report 274/2000. Nottingham: University of Nottingham; 2000.
4. Ferlitsch A, Glauninger P, Gupper A, Schillinger M, Haefner M, Gangl A, Schoefl R. Evaluation of a virtual endoscopy simulator for training in gastrointestinal endoscopy. Endoscopy 2002;34:698–702.
5. Letterie GS. How virtual reality may enhance training in obstetrics and gynecology. Am J Obstet Gynecol 2002;187(Suppl):37–40.
6. Bao JS, Jin Y, Gu MQ, Yan JQ, Ma DZ. Immersive virtual product development. J Mater Process Technol 2002;129:592–6.
7. Behar I, Wiley RW, Levine RR, Rash CE, Walsh DJ. Visual Survey of Apache Aviators (VISAA). Report no. 90-15. Fort Rucker, AL: United States Army Aeromedical Research Laboratory; 1990.
8. Kotulak JC, Morse SE. Oculomotor responses with aviator helmet-mounted displays and their relation to in-flight symptoms. Hum Factors 1995;37:699–710.
9. Pastoor S, Wopking M. 3-D displays: a review of current technologies. Displays 1997;17:100–10.
10. Regan EC, Price KR. The frequency of occurrence and severity of side-effects of immersion virtual reality. Aviat Space Environ Med 1994;65:527–30.
11. Rushton S, Mon-Williams M, Wann JP. Binocular vision in a bi-ocular world: new-generation head-mounted displays avoid causing visual deficit. Displays 1994;15:255–60.
12. Regan EC. Some evidence of adaptation to immersion in virtual reality. Displays 1995;16:135–9.
13. Howarth PA, Costello PJ. The occurrence of virtual simulation sickness symptoms when an HMD was used as a personal viewing system. Displays 1997;18:107–16.
14. Cobb SV, Nichols S, Ramsey A, Wilson JR. Virtual reality-induced symptoms and effects (VRISE). Presence: Teleop Virt Environ 1999;8:169–86.
15. Mon-Williams MA, Wann JP, Rushton S. Virtual reality displays may cause change in the AC/A ratio. Invest Ophthalmol Vis Sci 1995;36(Suppl):459.
16. Mon-Williams M, Wann JP. Binocular virtual reality displays: when problems do and don’t occur. Hum Factors 1998;40:42–9.
17. Howarth PA. Oculomotor changes within virtual environments. Appl Ergon 1999;30:59–67.
18. Howarth PA, Costello PJ. The nauseogenicity of using a head mounted display, configured as a personal viewing system, for an hour. In: Proceedings of the Second FIVE International Conference, Palazzo dei Congressi, Pisa, Italy, December 19–20, 1996. London, UK: Queen Mary & Westfield College: 1996:146–53.
19. Bullard HL, Ball WL, Morse SE, Jiang BC. Accommodation stimulus response functions of symptomatic and asymptomatic subjects differ after virtual reality immersion. Optom Vis Sci 1998;75(Suppl):252.
20. Neveu C, Blackmon T, Stark L. Evaluation of the effects of a head-mounted display on ocular accommodation. Presence: Teleop Virt Environ 1998;7:278–89.
21. Peli E. The visual effects of head-mounted display (HMD) are not distinguishable from those of desk-top computer display. Vision Res 1998;38:2053–66.
22. Morse SE, Jiang BC. Oculomotor function after virtual reality use differentiates symptomatic from asymptomatic individuals. Optom Vis Sci 1999;76:637–42.
23. Geelhoed E, Falahee M, Latham K. Safety and comfort of eyeglass displays. Lect Notes Comput Sci 2000;1927:236–47.
24. Ames SL. The effects of viewing stereoscopic virtual imagery with a head mounted display on the human oculomotor system. PhD Thesis. University of Melbourne; 2003.
25. Wann JP, Rushton S, Mon-Williams M. Natural problems for stereoscopic depth perception in virtual environments. Vision Res 1995;35:2731–6.
26. Mon-Williams M, Pascal E. Virtual reality displays: implications for optometrists. Optom Today 1995;35:30–3.
27. von Noorden GK. New clinical aspects of stimulus deprivation amblyopia. Am J Ophthalmol 1981;92:416–21.
28. Oliver M, Neumann R, Chaimovitch Y, Gotesman N, Shimshoni M. Compliance and results of treatment for amblyopia in children more than 8 years old. Am J Ophthalmol 1986;102:340–5.
29. Keech RV, Kutschke PJ. Upper age limit for the development of amblyopia. J Pediatr Ophthalmol Strabismus 1995;32:89–93.
30. Rosenfield M, Chiu NN, Ciuffreda KJ, Duckman RH. Accommodative adaptation in children. Optom Vis Sci 1994;71:246–9.
31. Wong L, Rosenfield M, Chiu NN. Vergence adaptation in children. Optom Vis Sci 1994;71(Suppl):71.
32. Gwiazda J, Bauer J, Thorn F, Held R. Shifts in tonic accommodation after near work are related to refractive errors in children. Ophthalmic Physiol Opt 1995;15:93–7.
33. Hasebe H, Oyamada H, Ukai K, Toda H, Bando T. Changes in oculomotor functions before and after loading of a 3-D visually-guided task by using a head-mounted display. Ergonomics 1996;39:1330–43.
34. Krebs WK, Kaiser JP, Darken RP, Sinai MJ, McCarley JS. Visual deficits following prolonged exposure in a virtual environment. Invest Ophthalmol Vis Sci 1999;40(Suppl):433.
35. Mon-Williams M, Wann JP, Rushton S. Binocular vision in a virtual world: visual deficits following the wearing of a head-mounted display. Ophthalmic Physiol Opt 1993;13:387–91.
36. Ames SL, Wolffsohn JS, McBrien NA. The development of a symptom questionnaire for assessing virtual reality viewing using a head-mounted display. Optom Vis Sci 2005;82:168–76.
37. Singer MJ, Ehrlich JA, Allen RC. Virtual environment sickness: adaptation to and recovery from a search task. In: Proceedings of the Human Factors and Ergonomics Society 42nd Annual Meeting, Chicago, IL, October 5-9, 1998. Santa Monica: Human Factors and Ergonomics Society;1998:1506–10.
38. Hyvarinen L, Nasanen R, Laurinen P. New visual acuity test for pre-school children. Acta Ophthalmol (Copenh) 1980;58:507–11.
39. Becker R, Hubsch S, Graf MH, Kaufmann H. Examination of young children with Lea symbols. Br J Ophthalmol 2002;86:513–6.
40. Repka MX. Use of Lea symbols in young children. Br J Ophthalmol 2002;86:489–90.
41. Sloan LL. Measurement of visual acuity: a critical review. Arch Ophthalmol (Chicago) 1951;45:704–25.
42. Manas L. The inconsistency of the AC/A ratio. Am J Optom Arch Am Acad Optom 1955;32:304–15.
43. Mallen EA, Wolffsohn JS, Gilmartin B, Tsujimura S. Clinical evaluation of the Shin-Nippon SRW-5000 autorefractor in adults. Ophthalmic Physiol Opt 2001;21:101–7.
44. Chat SW, Edwards MH. Clinical evaluation of the Shin-Nippon SRW-5000 autorefractor in children. Ophthalmic Physiol Opt 2001;21:87–100.
45. Schor CM, Tsuetaki TK. Fatigue of accommodation and vergence modifies their mutual interactions. Invest Ophthalmol Vis Sci 1987;28:1250–9.
46. Miller RJ. Ocular vergence-induced accommodation and its relation to dark focus. Percept Psychophys 1980;28:125–32.
47. Dallal GE, Wilkinson L. An analytical approximation to the distribution of Lilliefors’s test statistic for normality. Am Statistician 1986;40:294–6.
48. Leibowitz HW, Owens DA. Anomalous myopias and the intermediate dark focus of accommodation. Science 1975;189:646–8.
49. Rosner J. Relation between clinically measured tonic accommodation and refractive status in 6- to 14-year-old children. Optom Vis Sci 1989;66:436–9.
50. Kruger PB. The effect of cognitive demand on accommodation. Am J Optom Physiol Opt 1980;57:440–5.
51. Nichols S, Patel H. Health and safety implications of virtual reality: a review of empirical evidence. Appl Ergon 2002;33:251–71.
52. Best S. Perceptual and oculomotor implications of interpupillary distance settings on a head-mounted virtual display. In: Proceedings of the IEEE 1996 National Aerospace and Electronics Conference (NAECON), Dayton, OH, May 20-23, 1996. Piscataway, NJ: IEEE Service Center;1996:429–34.