Using Digital Game as Clinical Screening Test to Detect
Color Deficiency in Young Children
Linh-Chi Nguyen
Weiquan Lu
Ellen Yi-Luen Do
Keio-NUS CUTE Center
Interactive & Digital Media Institute
National University of Singapore
Keio-NUS CUTE Center
Interactive & Digital Media Institute
National University of Singapore
Keio-NUS CUTE Center
Interactive & Digital Media Institute
National University of Singapore
[email protected]
[email protected]
[email protected]
Audrey Chia
Yuan Wang
Singapore National Eye Center
11 Third Hospital Avenue
Singapore 168751
Keio-NUS CUTE Center
Interactive & Digital Media Institute
National University of Singapore
[email protected]
[email protected]
of two, a child has already has the capability to play many
computer games [22]. Children appreciate computer games
for their novelty, similar to other forms of entertainment
such as comic books and cartoons, as such activities
provide extreme experiences [10]. With their ability to
engage young children, it seems that computer games have
the potential to be used for situations for children in the
healthcare settings, such as in eye health screening.
ABSTRACT
Digital games as education tools for children have been
studied in the past. However, the use of digital games in
clinical environments such as for children’s healthcare is
still rare in the research community. This paper reports on
the development of a digital tablet game called “Dodo’s
catching adventure” which examines the use of games in
visual color-deficiency screening for young children. A user
study was conducted at a National Eye Centre. Results of
the study show that the digital game demonstrates
sensitivity and specificity on Red-Green color deficiency
detection, and is comparable to the two gold standards in
color deficiency tests, namely the Ishihara and Farnsworth
D15. Furthermore, children found the game to be more
enjoyable than the Ishihara test. This provides evidence for
the feasibility of using such games as diagnosis tools for
early childhood health conditions.
However, it is interesting to note that (to the best of our
knowledge), none of the validated color visions test for
children are in the form of a digital game. The conduct of
color screening tests for young children have traditionally
been wrought with difficulty since it often requires
advanced verbal or cognitive skills [9]. Such tests may also
reinforce the social stigma regarding such color
deficiencies, since children may become discouraged if they
compare their performance with their normal vision peers.
By using digital game technology via a screen-based
activity, our game (“Dodo's catching adventure,” or Dodo
game for short) is designed to detect color vision defects
effectively in children from two and a half to six years of
age, through a fun and relaxed manner.
Author Keywords
Digital game, color deficiency test, children game
ACM Classification Keywords
H.5.2 User Interfaces, K.8.0 Games
General Terms
THE NEED OF COLOR VISION DEFICIENCY TESTS FOR
YOUNG CHILDREN AND ITS CURRENT LIMITATIONS
Colorblind, children game
INTRODUCTION
Computer and videogames can be very engaging for
children [10]. Such games play a significant role in
children’s development. A recent study showed that 82% of
American children are avid gamer players, and by the age
Congenital Dyschromatopsia affects 8% to 10% of males
and 0.4% to 0.5% of females [5,7,17,18]. In Red-Green
color deficiency (Deuteranopia or Protanopia) are relatively
common, while Blue-Yellow color blindness (Tritanopia),
are extremely rare [5,17].
Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full
citation on the first page. Copyrights for components of this work owned by others
than ACM must be honored. Abstracting with credit is permitted. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee. Request permissions from [email protected].
IDC’14, June 17-20, 2014, Aarhus, Denmark.
Copyright 2014 ACM 978-1-4503-2272-0/14/06$15.00.
http://dx.doi.org/10.1145/2593968.2610486
Isochromatic plates, arrangement tests, anomaloscopes, and
lantern tests are commonly used in clinical practice today
[2,20]. The Ishihara test and the Farnsworth D15 test are
currently the most widely used screening tests. Ishihara is
used to detect Red-Green color deficiency and total color
deficiency. The test contains plates with circular dots of
different colors forming a number character that would be
invisible to a person with color deficiency but visible to a
normal vision person, or vice versa. The D15 test is used
337
Figure 1: Dodo game’s starting screen on a
tablet computer
Figure 2: Sub-game 3 layout
bubble next to the Dodo character with the “Pick This” text
and audio. Then, four groups of objects appear on the
screen for the player to pick the most similar one. The
player will pick either normal vision selectable objects or
color-blind selectable objects or neutral objects, depending
on his/her color vision.
worldwide to select applicants for employment in
occupations that require good color vision. In the test,
participants are asked to arrange sixteen color caps in a
logical rainbow order [4, 11].
Color vision tests for young children are important [1,8,20].
However, approximately 40% of color vision deficient
people are unaware of their condition [8]. Color vision
testing for young children remains a challenge for clinicians
because many screening tests require a relatively high
cognitive demand on children, especially on those younger
than 5 years old [6,9,19]. In fact, both the Ishihara and D15
tests are not suitable of children under the age of 5 [6,20].
Computer-based tests do not solve this cognitive limitation
[1,17,23]. While there exist screening tests specially
designed for young children, they are not as effective as
Ishihara and D15, and they require special communication
from the examiner, therefore the result may not be objective
[9,15,17,18,21].
The game’s win-lose condition is independent on whether
the players have color deficiencies. The player would be
prompted to play the game again if his/her choices were the
neutral (control) objects indicating that the player does not
fully understand the rules of the game, or that the player is
being uncooperative. Otherwise, the player will always win
the game. At the end of the game the screen always displays
the message “You Win!” This is designed intentionally to
not discourage color deficient children. The color
deficiency test results can be accessed by clicking on the
information icon located on the top right of the screen.
Design method
Based on Gestalt principles, such as the law of similarity in
visual perception, humans naturally tend to group similar
objects by shape, color, size or brightness together [25].
The design method of Dodo game is based on the
assumption that differences in hue and saturation of objects
would direct players with color deficiency to group objects
differently than those with normal vision.
DODO GAME DESIGN
Dodo game’s main target users are young children. Hence,
the game adopts a color matching game type with audio
support, which is suitable for young children. The game is
designed with bright color tones against a white background
to ensure minimal interference with game object. It contains
colorful graphics and interesting characters in order to
engage young children. Dodo game is currently available
on iPads and PC platforms (Figure1).
When a color deficient individual cannot distinguish the
hue of objects (since they look alike), they can only notice
the difference of object’s saturation and brightness (Figure
4). The key feature of the game design is to compensate for
the low resolution in the recognition of hues, as colorblind
individuals tend to be more sensitive to the differences in
saturation [16]. An individual with normal vision can
hardly notice the difference of saturation between the
sample object and the normal vision object, and will
naturally choose this object when he/she is asked to pick an
object similar to the sample object. Figure 4 shows a color
In this game, players are asked to tap the screen to pick an
object with similar color profiles. The Dodo game includes
4 sub-games: Sub-game 1 (Ladybug game) is designed to
test total color deficiency. However, both Red-Green and
Blue- Yellow color deficiency sufferers can also fail this
game; Sub-game 2 (Fish game) is used for testing BlueYellow color deficiency, Sub-game 3 (Butterfly game) is
used for testing Red-Green color deficiency and the Subgame 4 (Owl game) is a classification test for two types of
Red-Green color deficiency: deuteranopia and protanopia.
Each sub-game consists of six quick stages where each
player is asked to pick one object out of four objects that
looks the most similar in color to the a sample object.
Objects in each game can be grouped into four groups: a
sample object, normal vision objects, color-blind objects
and neutral objects. At the start of each sub-game, the
blinking sample object (e.g., a butterfly) appears in a speech
Figure 4: Dodo game design concept
338
profile of the sample object (2), the color-blind selectable
object (3) and the normal vision selectable object (1) in subgame 3. When being asked to pick the object similar to a
sample object (2), a Red-Green color deficient person sees
the hues of all objects are nearly alike, hence they will
choose (3) with similar saturation. A normal vision person
will choose (1) with similar hue. All the sub-games are
designed based on the same principle for different types of
color deficiencies.
test. Dodo game was successful in detecting color deficient
subjects with no false positives.
Observation of Dodo game’s usability on toddlers and
pre-school children
In the game development process, we asked a small group
of young children to play Dodo game. They were all
detected as having normal color vision as indicated by the
Dodo game. All of the children were able to play and win
the game easily in the first trial with minimal instruction.
None of the children realized that the Dodo game was a
screening test.
The color design was adopted from the Ishihara PC based
version with isocromatic data for Red–Green color
deficiency established by Lakowski [15,17]. Vischeck color
deficiency vision simulation was used to simulate the color
rendering of objects to color deficiency color vision for
color-blind selectable objects [13]. Figure 5 shows how
Sub-game 3 appears when viewed by a normal vision
person and the Red-Green color deficient person
respectively, using the Vischeck computer simulation.
USER STUDY
An ideal color vision test should reliably detect, categorize
and grade the severity of color vision deficiencies.
Currently, the two gold standard tests, Ishihara and the D15,
are the most used color vision test in optometry clinics
[7,11]. Dodo game’s target users were children from two to
six years old. However, such young color deficient subjects
were not available for recruitment because the current
clinical practice usually does not ask children at that age to
do colorblind tests at hospitals. Thus, as the two gold
standard tests are not recommended for children under six
years old, there did not exist a ground-truth to evaluate the
game’s reliability with the target group. Optometrists at
Singapore National Eye Centre suggested using Ishihara
test and the Farnsworth D15 test (D15) to compare Dodo
game’s effectiveness on Red-Green color deficiency
detection. During the six-month user study, we were not
able to recruit Blue-Yellow (Tritanopia) color deficiency
participants due to the rarity of such sufferers.
Figure 5: Sub-game 3 in normal vision (left) and in RedGreen color deficient vision simulated by Vischeck (right)
Dodo game’s criteria
Both Ishihara and D15 do not provide a clear grading and
give little indications of severity based on the number of
mistakes. Therefore, Dodo game follows this grading
convention. In terms of the number of errors that a player is
allowed to make, the percentage to distinguish between
mild and strong color deficiency is set to be more than 50%.
Therefore in Red-Green color deficient detection, if the
ratio is above 10/18 errors, the player will be a strong color
deficient. Players that make between 4-9/18 errors would be
classified as having slight/mild color deficiency.
A within-subjects repeated measures user study was
supervised by optometrists at the National Eye Centre.
Thirty-two subjects (N=32, Mean Age = 11.42, 1 Female)
were recruited. All the tests were conducted under good
indoor lighting in a clinical environment with both eyes
tested simultaneously. The iPad version of Dodo game was
used in this study. The screen brightness was kept balanced
at the default setting (auto brightness). Sixteen of subjects
were Red-Green color deficient, as detected by Ishihara
test. The control group consisted of sixteen normal color
vision participants. The average playtime of Dodo game
was 3.14 minutes to finish four sub-games.
Pre-test on adults to verify design method
A pre-test was conducted on six male adults, including
three Red-Green color deficient individuals, and three
normal vision individuals, all tested using the Ishihara test.
The reason for this pre-test, was that adult players could
comprehend and communicate with researchers directly,
thus preventing any confusion or negative results caused by
a lack of understanding. A pre-test result would be useful to
determine if our current prototype was functional, and the
pre-test would help researchers fine-tune the different
elements. The tests were conducted under good room
lightening, and both eyes were tested simultaneously. A PC
version of Dodo game on a 13 inch Macbook laptop was
used in this pilot study. Color calibration was performed
before each participant was asked to play the game. All the
participants were able to play the game from the start of the
Sensitivity and specificity
Using the Ishihara test results as reference (which detected
all sixteen colorblind subjects), D15 mis-classified four
subjects as normal (25%), while Dodo game mis-classified
three subjects as normal (18%). All tests classified the
normal subjects correctly.
Comparing enjoyment level of Dodo game to the two
gold standard tests
For enjoyment measuring, we modified a related sub-scale
of Intrinsic Motivation Inventory (IMI) with 7 questions,
using a seven-point Likert scale [12]. Three tests were
counterbalanced to avoid order effects, with the same
questionnaires administered after participants completed
339
each tests. Using repeated measures ANOVA, we analyzed
the results. In terms of Enjoyment, the results showed a
significant main effect (F[2,62] = 4.01, p < .05).
(NZHTA) NZHTA Report 7
nzhta.chmeds.ac.nz/publications/nzhta7.pdf
9. Cotter S.A., Lee, French A.L. Evaluation of a new color
vision test: ‘Color Vision Testing Made Easy. Optom
Vis Sci 76 (1999), 631–6.
10. Crawford, C. The Art of Computer Game Design
McGraw-Hill (1984)
www.stanford.edu/class/sts145/Library/Crawford%20on
%20Game%20Design.pdf
11. Dain, S.J.Clinical colour vision tests, Clin Exp Optom
87 4-5 (2004), 276-293
12. Deci, E.L., Ryan, R.M. Intrinsic Motivation Inventory.
http://www.selfdeterminationtheory.org/questionnaires/
10-questionnaires/50
13. Dougherty, B., Wade, A. Vischeck simulates colourblind vision. www.vischeck.com
14. Grammenos, D., Savidis, A., Stephanidis, C. Designing
universally accessible games. Comput Entertain 7, 1
(2009), 1-29.
15. Hoffmann, A., Menozzi, M. Applying the Ishihara test
to a PC-based screening system, Displays 20 1 (1998),
39-47
16. Kohei, M. Color Universal Design (CUD) How to make
figures and presentations that are friendly to Colorblind
people. http://jfly.iam.u-tokyo.ac.jp/color/#see
17. Lakowski, R. Theory and practice of color vision
testing. A review. Part 1 Br J Industr Med 26 (1969),
173–288.
18. Ling, B.Y., Dain, S.J. Color vision in children and the
Lanthony New Color Test. Visual Neuroscience 25
(2008), 441-444.
19. Mäntyjärvi, M. Color vision testing in pre-school-aged
children, Ophthalmologica 202 3 (1991), 147-151.
20. Melamud, A., Hagstrom, S., Traboulsi, E. I., Color
vision testing, Ophthalmic Genetics 25 3 (2004), 159–
187.
21. Neitz, M., Neitz, J. A new mass screening test for colorvision deficiencies in children, Color Research &
Application, Special Issue: In Proc of the International
Color Vision Society 26 S1 (2001), 239-249.
22. NPD Group. Among American kids ages 2-17, 82
percent report that are gamers. Press Release (Dec 2)
Port Washington, New York, 2009.
23. Pardo, P.J. PeÂrez, A.L., Suero, A.L. A new color
vision test in a PC-based screening system, Displays 21
5 (2002), 203-6
24. Prensky, M. Digital Game-Based Learning, chapter 5:
Fun, Play and Games: What Makes Games Engaging,
2001.
25. Soegaard, M. Gestalt principles of form perception.
www.interactiondesign.org/encyclopedia/gestalt_principles_of_form_per
ception.html
Using paired samples t-tests between Ishihara and D15
(t[31] = -1.24, p=.23), D15 and Dodo game (t[31] = -1.55, p
= .13), Ishihara and Dodo game (t[31] = -2.925, p<.01),
Dodo game was significantly more enjoyable than Ishihara.
DISCUSSION, CONCLUSION AND FUTURE WORK
The results suggested that Dodo game had slightly better
sensitivity than D15, but less than Ishihara. In terms of
enjoyment, Dodo game was significantly more enjoyable
than Ishihara, while being on par with D15.
The results give support for the potential use of Dodo game
as a new approach for the use of digital games to screen
children at pediatric clinics or at home. As a new clinical
tool for color vision screening, it is compatible with the two
gold standard Ishihara and D15 tests in sensitivity and
specificity. Portable, easy and fun - the game may be a
good choice for color vision screening for children below
the ages of five. Future work will involve the fine-tuning of
Dodo game, and testing it with a larger population of young
children.
ACKNOWLEDGMENTS
This research is supported by the Singapore National
Research Foundation under its International Research
Center Keio-NUS CUTE Center @ Singapore Funding
Initiative and administered by the IDM Program Office.
REFERENCES
1. Arden, G., Gundoz, K., Perry, S. Color vision testing
with a computer graphics system: Preliminary results,
Documenta Ophthalmologica 69 (1998), 167-174.
2. Birch, J. A practical guide for color-vision examination:
Report of the Standardization Committee of the
International Research Group on Color-Vision
deficiencies. Ophthalmic Physiol Opt 5 (1985) 265–285.
3. Birch, J. Efficiency of the Ishihara test for identifying
red-green color vision deficiency. Ophthalmic Physiol
Opt 17, 5 (1997), 403–8.
4. Birch, J. Use of the Farnsworth—Munsell 100‐Hue test
in the examination of congenital color vision defects,
Ophthalmic and Physiological Optics, Wiley Online
Library, 1998.
5. Chia, A., Gazzard, G., Tong, L., Zhang, X., Sim, EL.,
Fong, A., Saw, S.M. Red-Green color blindness in
Singaporean children. Clinical & Experimental
Ophthalmology 36 5 (2008) , 464–7.
6. Choi, S.Y., Hwang, J.M. Ishihara test in 3- to 6-year-old
children, Jpn J Ophthalmol 53 5 (2009), 455-7
7. Cole, B.L. The new Richmond HRR
pseudoisochromatic test for colour vision is better than
Ishihara test, Clin Exp Optom 89 2 (2006), 73-80
8. Color vision screening: a critical appraisal of the
literature, New Zealand Health Technology Assessment
340