Clinical Therapeutics/Volume 38, Number 3, 2016
RGB Measurements From Endoscopic Photography as a
New and Objective Diagnostic Method for Chronic Tonsillitis
Alper Yenigun, MD; Remzi Dogan, MD; Orhan Ozturan; and Aysenur Meric Hafiz
Bezmialem Vakif University, Faculty of Medicine, Department of Otorhinolaryngology, Adnan Menderes
Bulvar Vatan Caddesi 34093, Fatih, Istanbul, Turkey
ABSTRACT
Purpose: This study assessed the diagnostic role of
a digital photographic assessment of chronic tonsilli-
tis. This study was performed at a tertiary referral
center.
Methods: Three cohorts of children were studied
based on their diagnoses: current chronic tonsillitis
(before tonsillectomy), previous chronic tonsillitis
(3 months after tonsillectomy), and healthy children
(control group). Oropharyngeal regions of all subjects
were examined and photographed endoscopically.
Red/green/blue (RGB) values of particular oropharyn-
geal points were then measured.
Findings: The current chronic tonsillitis group differed
signicantly from the other 2 groups. The RGB values of
the current chronic tonsillitis group were signicantly
higher than those of the previous chronic tonsillitis group
(3 months after tonsillectomy) and the control group
(P 0.018 (red), P 0.024 (green), and P 0.004
(blue)). The RGB values did not differ signicantly
between the previous chronic tonsillitis group and the
control group (P 0.684 [red], P 0.890 [green], and
P 0.672 [blue]). The RGB values of the previous chronic
tonsillitis group decreased signicantly 3 months after
tonsillectomy (P 0.015 [red], P 0.033 [green], and
P 0.001 [blue]). Sensitivity was 80%, specicity was
90%, the positive predictive value was 59%, the negative
predictive value was 96%, and the general power of the
test was 88.6% regarding the power and competence of
the RGB test in the diagnosis of chronic tonsillitis.
Implications: Measuring RGB values may be an
inexpensive and simple way to quickly provide objective
and corroborative information on hyperemia of the
anterior pillars in diagnosing chronic tonsillitis in con-
junction with subjective methods. (Clin Ther.
2016;38:625630) & 2016 Elsevier HS Journals, Inc.
All rights reserved.
Key words: chronic disease, diagnosis, photography,
tonsillectomy, tonsillitis.
March 2016
INTRODUCTION
Chronic tonsillitis is the most common upper respira-
tory tract infection observed in childhood.1,2 The
frequent recurrence and the chronic nature of tonsil
infections involve major costs.35 Those who experi-
ence acute tonsillitis attacks Z7 times in 1 year, Z5
times in 2 years, or Z3 times in 3 years, and those
who experience symptoms such as continuous pain in
the throat and fatigue for at least 1 year have adequate
indications of a chronic tonsillitis diagnosis and
indications for a tonsillectomy.6
Anterior plica hyperemia, tonsillar hypertrophy
according to the Brodsky scale, tonsillar asymmetry,
debris coming out of the crypts when the tonsil is
pressed, the existence of intratonsillar abscesses, and
cervical lymph node positivity are important exami-
nation ndings in the diagnosis of chronic tonsillitis.7
A chronic tonsillitis diagnosis is still based on medical
history and physical examination ndings.
Red, green, and blue (RGB) lights are color compo-
nents of white light. The color most reected by an
object that we look at is perceived by our brain as the
color of that object.8 In this study, we attempted to
determine the numerical differences by comparing RGB
values in digital photographs obtained as endoscopic
oropharyngeal images of patients having chronic
tonsillitis and of healthy individuals and to provide
an objective way to diagnose chronic tonsillitis.
MATERIALS AND METHODS
This was a prospective study and approved by the local
clinical research ethics committee. Chronic tonsillitis
patients, healthy volunteers, and/or their parents were
Accepted for publication January 26, 2016.
http://dx.doi.org/10.1016/j.clinthera.2016.01.019
0149-2918/$ - see front matter
& 2016 Elsevier HS Journals, Inc. All rights reserved.
625
 Clinical Therapeutics

informed about the research project, and their consent

was obtained before participation in the study.
Uvula
The study included 226 patients (127 females 3
3 3
[56.2%] and 99 males [43.8%]) who came to the Right Left
anterior 2 3 anterior
Ear, Nose, and Throat Outpatient Clinic between 2 1
2 plica
plica 3
January and June 2013. Patients were divided into 3 3 1
2
groups based on their diagnoses: current chronic 1 2 1
2
tonsillitis (before tonsillectomy), previous chronic 1
Posterior pharynx
1

tonsillitis (3 months after tonsillectomy), and healthy

children (control group).
Chronic tonsillitis was dened as recurrent throat
infection with a frequency of at least 7 episodes in the
past year, at least 5 episodes per year for 2 years, or at
least 3 episodes per year for 3 years. Patients were
included if their medical records documented each episode
of sore throat with Z1 of the following: temperature
438.31C, cervical adenopathy, tonsillar exudate, or
positive test result for group A -hemolytic streptococcus.6
Patients were excluded if they were older than 18 years
of age; did not agree to undergo the photography process
or had chronic lung disease, cardiac disease, hypothyr-
oidism, chronic kidney disease, obstructive sleep apnea,
craniofacial anomalies, cerebral palsy, malnutrition, neu-
romuscular diseases, chronic allergic rhinitis, mental and/
or physical retardation due to various illnesses such as
Down syndrome, or various other illnesses that might
cause chronic inammation. Group 1 consisted of 76
patients with current chronic tonsillitis, with an age range
of 3 to 18 years. They came to the Ear, Nose, and Throat
Outpatient Clinic because of symptoms of poor appetite
and frequent and recurrent tonsillitis. All patients were
evaluated for tonsillar hypertrophy using the scale devel-
oped by Brodsky et al.9 Tonsil size alone was not
accepted as a criterion for exclusion. Patients who had
experienced an acute attack of tonsillitis within the
preceding 2 weeks were not included. All patients in
this group later underwent tonsillectomies.
Group 2 consisted of 72 patients who had been
included in Group 1 at 3 months after their tonsillec-
tomy operation. Their age range was also 3 to 18 years.
The control group consisted of 78 healthy children
from the outpatient clinic of the Pediatrics Department
who were found to be clinically healthy on their
follow-up visits. Their age range was 4 to 18 years.
None of the patients underwent surgical treatments.
RGB Measurements
During otorhinolaryngology examinations, the an-
terior plicas, tonsils, uvula, and posterior oropharynx
Right Left
tonsil tonsil
Figure 1. Oropharyngeal points for the RGB
measurements. 1 Group 1; 2
Group 2; 3 Group 3.
were photographed using a zero-degree, 4-mm endo-
scope. In the oropharyngeal region, RGB measure-
ments were taken at a total of 18 different points: 3
points each in 6 particular regions (right anterior
plica, left anterior plica, right tonsil, left tonsil, uvula,
and posterior pharynx) (Figure 1). The distance
between the tissue and endoscope was  2 cm in all
patients. All measurements were performed by the
same surgeon at the same distance with the same
endoscope. When a circular structure with a 1-mm
diameter is matched on these standard points, the
Adobe Photoshop Elements 7.0 program (Adobe
Systems, San Jose, California) automatically provides
their RGB values.
Statistical Analyses
Statistical analyses were carried out using the SPSS
software (Version 13.0 for Windows, SPSS Inc., Chi-
cago, Illinois). All quantitative variables were estimated
using measures of central location and of dispersion. In
addition to descriptive statistics (mean and SD) in
evaluating the data, 1-way ANOVA and Tukeys
honestly signicant difference (HSD) test were used to
compare the quantitative data and the normally dis-
tributed parameters between groups. The signicance
levels were set at P o 0.001 and P o 0.05.
RESULTS
Demographic Information
For the 3 groups, the total number of study
participants, the number of females/males, and the

626 Volume 38 Number 3

 A. Yenigun et al.

Table I. Demographic characteristics of the groups.

Variables Group 1 Group 2 Control Group P

Sex
Male 33 32 34 0.674
Female 43 40 44 0.241
Age, median (SD) 17.2 (12.4) 16.1 (9.6) 15.8 (9.6) 0.199
Age, y, range 318 318 418

A 2-sample t test was administered as a post hoc test to identify within-group differences (P o 0.05 and P o 0.001 were
accepted as statistically signicant).

average (SD) ages were as follows: Group 1 (n 76,
43/33, 17.2 [12.4]), Group 2 (n 72, 40/32, 16.1
[9.6]), and Group 3 (n 78, 44/3 4, 15.8 [9.6]). The
groups were statistically comparable in terms of age
and sex. Tukeys HSD test results comparing the mean
ages of the groups did not reveal a statistically
signicant difference (P 4 0.05) (Table I).
RGB Values in the Oropharynx
RGB values of the 6 regions (18 points) in the
oropharynx were compared in the 3 groups. One-way
ANOVA revealed that the RGB values of Group 1
patients were signicantly higher than those of Group
2 and the control group (P 0.018 [red], P 0.024
[green], and P 0.004 [blue]) (Table II). In contrast,
the RGB values in Group 2 and the control group did
not differ signicantly (P 0.684 [red], P 0.890
[green], and P 0.672 [blue]) (Table II).
When the RGB values were compared between
Groups 1 and 2, signicant decreases were observed
in Group 2 (P 0.015 [red], P 0.033 [green], and
P 0.001 [blue]) (Table II).
Regional RGB Values in the Oropharynx
When the RGB values of 6 regions (18 dif-
ferent points) in the oropharynx were assessed,
Group 1 had the highest values in the anterior pillar,
followed by the tonsillar region; the lowest RGB
values were in the posterior pharyngeal region
(Figure 2).
Sensitivity was 80% and specicity as 90%, the
positive predictive value was 59%, the negative
predictive value was 96%, and the general power of
the test was 88.6% with regard to the power and
competence of the RGB test in the diagnosis of chronic
tonsillitis.

Table II. Mean oropharyngeal RGB values of the groups and statistics.

Groups Red, Mean (SD) Green, Mean (SD) Blue, Mean (SD)

Group 1 (n 76) 195 (33.4) 133.3 (18.5) 132 (20.5)

Group 2 (n 72) 160 (27.7) 101.6 (21.2) 104.1 (19.9)
Group 3 (n 78) 152.5 (26.0) 93.5 (8.6) 95 (19.8)
1-way ANOVA test Group 1 vs Groups 2/3 0.018* 0.024* 0.004*
1-way ANOVA test Group 2 vs Group 3 0.684 0.890 0.672
1-way ANOVA test Group 1 vs Group 2 0.015* 0.033* 0.001**

RGB red, green, and blue.

One-way ANOVA test was administered as a post hoc test to identify within group differences.
*
P o 0.05 and **P o 0.001 were accepted as statically signicant.

March 2016 627

 Clinical Therapeutics

RED GREEN
250
180
160
200
140
150 120
100
Group 1 Group 1
100 80
Group 2 60 Group 2
50
Group 3 40 Group 3
20
0
0
Right Right tonsil Uvula Posterior Left tonsil Left anterior
anterior pharynx plica Right Right Uvula Posterior Left tonsil Left
plica anterior tonsil pharynx anterior
plica plica

BLUE
180
160 Group 1 (n = 76) (before tonsillectomy)
140 Group 2 (n = 72) (3 months after tonsillectomy)
120
Group 3 (n = 78) (control group)
100
Group 1
80
Group 2
60
Group 3
40
20
0
Right Right Uvula Posterior Left tonsil Left
anterior tonsil pharynx anterior
plica plica

Figure 2. Mean red, green, and blue values of the oropharyngeal regions.

DISCUSSION tonsillectomy. The most important reasons for sur-

Diseases of the tonsils are one of the main reasons that gical treatment are the recurrence of acute tonsillitis
children visit an otolaryngology clinic. Chronic ton- attacks despite antibiotic therapy, the continuation
sillitis, meaning that the inammation symptoms last of a patients pharyngeal symptoms even between
at least 3 months and are accompanied by a sore attacks, and adverse effects on respiration and
throat, is an important infectious tonsil condition.10 swallowing functions, particularly in children.
Suppurative complications can appear in patients who Although a chronic tonsillitis diagnosis and tonsil-
are not treated appropriately or in a timely manner, lectomy indications are determined according to
such as peritonsillar abscess, parapharyngeal abscess, medical history and physical examination ndings,
and retropharyngeal abscess. Nonsuppurative compli- the frequency of such operations varies according to
cations can also occur, such as scarlet fever, acute age, geographic regions, the quality of the health
rheumatic fever, and post-streptococcal glomerulo- system and how widespread it is in the country,
nephritis. Infections in this region are typically poly- the views of otorhinolaryngology and pediatric
microbial and often reect oropharyngeal ora.11 To physicians regarding the operation, and education
identify the agents responsible for tonsil infections, levels and socioeconomic status of patients and their
swab cultures are usually taken from the tonsil parents.13,14
surface. Because oral secretions contaminate the As antibiotic use became widespread and more
tonsil surface, normal ora reproduction is generally effective antibiotics were found, the frequency of
followed up using cultures.12 infections leading to tonsillectomies was dramatically
Chronic tonsillitis is a clinical term, dened based reduced.15 Although the diagnosis and treatment of
on reliable medical history information and physical chronic tonsillitis can be successful, they are not
examination ndings. Treatment usually involves always. Follow-up often cannot be achieved,

628 Volume 38 Number 3


particularly among patients whose medical history is
not adequately obtained and a focal infection may
need to be researched. The decision to perform a
tonsillectomy may be disruptive, and satisfactory
results may not be obtained. Although some patients
prefer medical treatment, some may believe that
tonsillectomy will heal other unconnected diseases.
As a result, trust problems may occur regarding
medical history information: the patient or family
may exaggerate some symptoms while concealing
others. It may be impossible to observe and control
acute tonsillitis attacks in a patient in whom follow-up
is difcult. We tried to provide an effective, inexpen-
sive, noninvasive, and objective auxiliary method for
diagnosing chronic tonsillitis infection by assessing
and comparing RGB values from endoscopic digital
photographs of the oropharyngeal area.
Light is electromagnetic radiation emitted from a
point in every direction. Our eyes perceive color as
a result of the reection of rays that are emitted from a
light source and then hit an object. The human eye is
most sensitive to 3 main colors: RGB. According to the
Optical Society of America, each color is formed by a
combination of 3 different properties: type, saturation,
and brightness.8 Every color in nature is formed by
combinations of these 3 main colors and can be coded
(between 0 and 255): white light is caused by a 100%
(255) mixture of the colors, whereas black is caused by
a 0% (0) mixture. Because RGB shades are not exact
and their codes may change, the colors will vary as
well. When photographs are transferred to a computer,
the Adobe Photoshop Elements 7.0 program can
measure the RGB values of a desired point or area on
the photograph.
Today, RGB measurements are used in various
endoscopic procedures, and pathology can be inter-
preted with these values. Some relationships have
already been identied between the changes in RGB
and the severity of disease, particularly with mucosal
involvement.8 Tugrul et al.16 identied a signicant
correlation between RGB values of tympanic membrane
images and the presence of effusion in the middle ear.
Because color changes in the mucosa can be meas-
ured objectively in the computer program based on
RGB codes, without being seen by the naked eye,
diagnosis and treatment may be improved during the
early stages of the disease, reducing the likelihood of
morbidities developing in the future. Additionally,
because hyperemia cannot be observed by the naked
March 2016
A. Yenigun et al.
eye, and different physicians may interpret hyperemia
images differently, RGB values may provide an objec-
tive and practical method for diagnosing this condition.
The key nding of this study was that when RGB
values were compared between Groups 1 and 2,
Group 2 values were signicantly decreased
(P 0.015 [red], P 0.033 [green], and P 0.001
[blue]), whereas the RGB values of Group 2 did not
differ signicantly from those of the control group
(P 0.684 [red], P 0.890 [green], and P 0.672
[blue]). RGB values may be used in chronic tonsillitis
cases to determine the timing of surgical intervention
and for postoperative follow-up and may help to
prevent local and/or systemic complications.
In our study, the highest RGB values in measure-
ments of the oropharyngeal region were observed in
the anterior plica region, followed by the tonsillar
region. The lowest RGB values were observed in the
posterior pharyngeal region (Table II). Thus, the
anterior plica and tonsillar region are apparently
most affected in chronic tonsillitis. Therefore, we
demonstrated an easy, inexpensive, and objective
auxiliary method for use in the diagnosis of chronic
tonsillitis: the anterior plica and tonsillar region
hyperemia, represented numerically by RGB values.
RGB values can be effective in both diagnosis and
follow-up for patients with suspicious symptoms.
More research is needed, including more patients to
conrm our ndings.
In our study, the idea that all red is not chronic
tonsillitis was taken as a fundamental. To be con-
sidered chronic tonsillitis, the intensity of red and its
location within the oropharynx was studied. For this
purpose, the intensity of point regions determined as
RGB at several different points was measured. Aver-
age RGB values of oropharyngeal sites should be
assessed instead of considering only 1 side when
diagnosing chronic tonsillitis. We took threshold
RGB values as R 4155, G 4105, and B 4100 at
the oropharyngeal site. The signicance of scientic
average RGB values was pointed out according to
these results. The increases in these values are impor-
tant for the diagnosis. In this way, the (chronic
tonsillitis /) R value that can be perceived as
similar to the naked eye is expressed as a unit with
the computer program evaluation method, and the
values that are above of certain units are accepted as
signicant. We think that a false diagnosis would be
prevented by this approach.
629
 Clinical Therapeutics
In our study, the positive predictive value was
calculated as 59% and the negative predictive value
as 96% in the RGB test for the diagnosis of chronic
tonsillitis. We think that hyperemia such as reux that
is not caused by infection may be the reason of high
SDs and the lower positive predictive value.
Our ndings demonstrated that the RGB values are
positively correlated with disease severity in patients
with chronic tonsillitis. They also suggest that RGB
values, which increase as a consequence of the
systemic inammatory response in chronic tonsillitis,
may be valuable in the clinical follow-up of patients
with chronic tonsillitis. This study is the rst to report
the signicant value of RGB measurements in chronic
tonsillitis.
CONCLUSIONS
Because measuring RGB values objectively distin-
guishes the plica and tonsil hyperemia in chronic
tonsillitis patients, it is a cost-effective, fast, and easy
method. Oropharyngeal RGB measurements may be a
useful auxiliary diagnostic method, along with sub-
jective methods, in diagnosing chronic tonsillitis in
clinical practice.
ACKNOWLEDGMENTS
All authors contributed equally.
CONFLICTS OF INTEREST
The authors have indicated that they have no conicts
of interest regarding the content of this article.
REFERENCES
1. Kvaerner KJ, Nafstad P, Jaakkola JJ. Otolaryngological
surgery and upper respiratory tract infections in children:
an epidemiological study. Ann Otol Rhinol Laryngol. 2002;
111:10341039.
2. Kvaerner KJ, Nafstad P, Jaakkola JJ. Upper respiratory
morbidity in preschool children: a cross-sectional study.
Arch Otolaryngol Head Neck Surg. 2000;126:12011206.
Address correspondence to: Alper Yenigun, MD, Department of Otorhi-
nolaryngology, Faculty of Medicine, Bezmialem Vakif University, Adnan
Menderes Bulvar Vatan Caddesi, 34093 Fatih/_Istanbul, Turkey. E-mail:
alperyenigun@gmail.com
630
3. Karevold G, Kvestad E, Nafstad P, Kvaerner KJ. Respira-
tory infections in schoolchildren: co-morbidity and risk
factors. Arch Dis Child. 2006;91:391395.
4. Aydogan M, Toprak D, Hatun S, et al. The effect of
recurrent tonsillitis and adenotonsillectomy on growth
in childhood. Int J Pediatr Otorhinolaryngol. 2007;71:1737
1742.
5. Birnbaum HG, Morley M, Greenberg PE, Colice GL.
Economic burden of respiratory infections in an employed
population. Chest. 2002;122:603611.
6. Silva S, Ouda M, Mathanakumara S, et al. Tonsillectomy
under threat: auditing the indications for performing
tonsillectomy. J Laryngol Otol. 2012;126:609611.
7. Webb CJ, Osman E, Ghosh SK, Hone S. Tonsillar size is an
important indicator of recurrent acute tonsillitis. Clin
Otolaryngol Allied Sci. 2004;29:369371.
8. Osada T, Arakawa A, Sakamoto N, et al. Auto uores-
cence imaging endoscopy for identication and assessment
of inammatory ulcerative colitis. World J Gastroenterol.
2011;17:51105116.
9. Brodsky L, Moore L, Stanievich J. The role of Haemophilus
inuenzae in the pathogenesis of tonsillar hypertrophy in
children. Laryngoscope. 1988;98:10551060.
10. Discolo CM, Darrow DH, Koltai PJ. Infectious indica-
tions for tonsillectomy. Pediatr Clin North Am. 2003;50:
445458.
11. Vieira F, Allen SM, Stocks RM, Thompson JW. Deep neck
infection. Otolaryngol Clin North Am. 2008;41:459483.
12. Gul M, Okur E, Ciragil P, et al. The comparison of
tonsillar surface and core cultures in recurrent tonsillitis.
Am J Otolaryngol. 2007;28:173176.
13. Faramarzi A, Kadivar MR, Heydari ST, Tavasoli M.
Assessment of the consensus about tonsillectomy and/or
adenoidectomy among pediatricians and otolaryngologists.
Int J Pediatr Otorhinolaryngol. 2010;74:133136.
14. Erickson BK, Larson DR, St Sauver JL, et al. Changes in
incidence and indications of tonsillectomy and adenoton-
sillectomy, 19702005. Otolaryngol Head Neck Surg. 2009;
140:894901.
15. Motta G, Esposito E, Motta S, et al. Acute recurrent
pharyngotonsillitis and otitis media. Acta Otorhinolaryngol
Ital. 2006;26:3055.
16. Tugrul S, Degirmenci N, Eren SB, et al. RGB measure-
ments as a novel objective diagnostic test for otitis media
with effusion. Acta Otolaryngol. 2015;135:342347.
Volume 38 Number 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without
permission.