ORIGINAL ARTICLE
Paranasal sinus opacification-to-pneumatization ratio applied as a rapid
and validated clinician assessment
Michael J. Marino, MD1 , Charles A. Riley, MD1 , Amit S. Patel, MD1 , Jason D. Pou, MD1 ,
Raymond H. Kessler, BS1 and Edward D. McCoul, MD, MPH1,2,3
Background: The utility of clinician-applied instruments,
particularly the Lund-Mackay score, in the assessment of
paranasal sinus computed tomography (CT) in chronic rhi-
nosinusitis (CRS) remains incompletely defined. The pur-
pose of this study was to determine if a new approach to the
evaluation of sinus CT could accurately predict the extent
of opacification while remaining simple for clinician use.
Methods: Twenty-four sinus CT scans were measured for
the percent of sinus opacification using three-dimensional
(3D) volumetric analyses. The same scans were also evalu-
ated using the Lund-Mackay score to measure opacification
and the Assessment of Pneumatization of the Paranasal
Sinuses (APPS) score to measure total sinus volume (TSV).
Correlation analysis was performed for the Lund-Mackay
to APPS score ratio as a predictor of percent opacification.
Validation analysis was also performed to determine the
optimal orientation for Lund-Mackay scoring, which has not
previously been described.
Results: The Lund-Mackay to APPS score ratio was very
strongly correlated with the percentage of sinus opaci-
fication measured by 3D volumetric analysis (r = 0.862,
T he utility of paranasal sinus computed tomogra-
phy (CT) in the evaluation of chronic rhinosinusi-
tis (CRS) remains incompletely defined. Previous attempts
1 Department of Otolaryngology-Head and Neck Surgery, Tulane
University School of Medicine, New Orleans, LA; 2 Department of
Otorhinolaryngology, Ochsner Clinic Foundation, New Orleans, LA;
3 Ochsner Clinical School, University of Queensland School of
Medicine, New Orleans, LA
Correspondence to: Edward D. McCoul, MD, MPH, 1514 Jefferson Highway,
CT-4, New Orleans, LA 70121; e-mail: emccoul@gmail.com
Additional Supporting Information may be found in the online version of this
article.
Potential conflict of interest: None provided.
Abstract accepted as an oral presentation at the American Academy of
Otolaryngic Allergy (AAOA) Annual Scientific Meeting, September 17, 2016,
in San Diego, CA, and as a poster presentation at the American Rhinologic
Society Annual Meeting, September 17, 2016 in San Diego, CA
Received: 27 May 2016; Revised: 1 July 2016; Accepted: 13 July 2016
DOI: 10.1002/alr.21833
View this article online at wileyonlinelibrary.com.
1 International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016
r2 = 0.743, p < 0.001). Lund-Mackay scoring was not statisti-
cally different between axial-only, coronal-only, or triplanar
groups for interrater (p = 0.379) and intrarater reliability
(p = 0.312).
Conclusion: The Lund-Mackay score is validated for rater
reliability in multiple orientations. Using the APPS score
as a measure of TSV, the Lund-Mackay-to-APPS ratio very
strongly correlates with the percentage of sinus opacifica-
tion by 3D volumetric analysis. Further study will be re-
quired to determine if this ratio is predictive of symptom
severity. C 2016 ARS-AAOA, LLC.
Key Words:
paranasal sinuses; computed tomography; chronic rhinosi-
nusitis; Lund-Mackay score; sinus opacification
How to Cite this Article:
Marino MJ, Riley CA, Patel AS, Pou JD, Kessler RH,
McCoul ED. Paranasal sinus opacification-to-pneuma-
tization ratio applied as a rapid and validated clinician
assessment. Int Forum Allergy Rhinol. 2016;XX:1–6.
to quantify disease severity based on CT imaging using
clinician-applied instruments such as the Lund-Mackay
scoring system have rarely demonstrated a correlation with
clinical symptoms.1–4 More recently, computer-assisted
three-dimensional (3D) volumetric or two-dimensional
(2D) area analyses of the actual percentage of sinus opaci-
fication have been shown to better correlate with symp-
tom severity.5–7 Detailed computer-assisted CT analysis,
however, has not been fully automated and remains time
consuming.5, 7–9 Furthermore, the necessary software is not
routinely used in clinical practice. Nevertheless, in the most
recent update to the clinical practice guideline for the di-
agnosis and management of CRS, sinus CT was identified
as an objective modality with “excellent” sensitivity for
evaluating mucosal inflammation.10
The Lund-Mackay score is the most frequently used clin-
ical metric for evaluating the radiographic extent of disease
in CRS. This scoring system has been shown to have sub-
stantial interrater and intrarater reliability,11, 12 and that the
Marino et al.
FIGURE 1. The same axial slice is shown at multiple points in the 3D volumetric analysis. (A) The opacified portions of the sinuses are manually segmented
using the 2D ROI tool. (B) All of the bony and soft tissue is included in a single ROI using the 3D ROI tool in order to isolate the aerated portions of the sinuses.
(C) The 3D ROI tool is used to automatically calculate the volume of the aerated spaces, after all the bony and soft tissue has been switched to 3024 HU.
3D = three-dimensional, 2D = two-dimensional, ROI = region of interest, HU = Hounsfield units.
rater reliability exceeds other proposed systems.13, 14 The
Lund-Mackay score can also be performed by clinicians
with varying degrees of experience in several minutes.12
These features make the Lund-Mackay system more ac-
cessible for routine clinical use than 3D volumetric anal-
ysis, despite concerns of poor correlation with symptom
severity.
The Assessment of Pneumatization of the Paranasal Si-
nuses (APPS) instrument has recently been introduced as
a validated metric for clinician evaluation of the extent
of sinus pneumatization.11 This score has been shown to
correlate with total sinus volume (TSV) as calculated by a
3D volumetric method.8 The APPS score can also be de-
termined in under 1 minute without the use of specialized
imaging software.8
In the current study we hypothesize that the Lund-
Mackay score may more accurately reflect the percentage
of sinus opacification when contextualized by the APPS
score as a measure of TSV. Therefore a ratio of Lund-
Mackay to APPS score might correlate with the percent-
age of sinus opacification calculated by a 3D volumetric
method. To our knowledge, rater reliability according to
CT orientation has not been studied previously for the
Lund-Mackay score. Therefore, a secondary aim of this
study was to investigate and validate the optimal orienta-
tion for Lund-Mackay scoring, and if axial-only or coronal-
only scoring was comparable to scoring using triplanar
reconstructions.
Patients and methods
Twenty-four CT scans with varying degrees of sinus opaci-
fication were selected for 3D volumetric analysis. CT scans
were acquired as part of the evaluation of CRS. Only
CT scans acquired in the axial plane with a slice thick-
ness of 0.625 mm as part of an image guidance proto-
col were considered for study inclusion. Scans that met
these criteria were consecutively selected from a library
of 70 complete Digital Imaging and Communications in
Medicine (DICOM) image sets, which had previously been
transferred to a Macbook Pro computer (Apple Inc., Cu-
pertino, CA) for use with OsiriX image viewer (version
7.0; Pixmeo, Geneva, Switzerland). Osirix runs exclusively
in a Mac operating system environment. The volume of
soft tissue opacification of the sinuses was calculated by
manual segmentation with the 2D region of interest (ROI)
tool in OsiriX. The interval threshold setting was used and
varied from 50 to 200 Hounsfield units (HU) for each in-
dividual scan (Fig. 1A). After calculating the volume of
opacification, the aerated spaces of the CT scan were iso-
lated by including all bony and soft tissue densities in a
single ROI using the automated 3D ROI tool. This was
performed with the bounded threshold setting at −350 to
3024 HU. The voxels within this ROI were then switched to
3024 HU (similar to dense bone), so that all aerated spaces
were isolated (Fig. 1B). The volume of the aerated space of
the sinuses was then calculated using the 3D ROI tool with
an interval threshold setting of 200 HU, and automated
errors were corrected using the 2D ROI tool (Fig. 1C). Per-
cent opacification was calculated from the volume of sinus
opacification of the entire paranasal sinus complex divided
by TSV.
Sinus opacification and TSV were then evaluated by clin-
ical metrics. The Lund-Mackay score was used to evaluate
sinus opacification, whereas the APPS score was employed
for measuring TSV. Lund-Mackay scores were assigned in
the standard fashion, with a score of 0 for “no abnor-
mality,” 1 for “partial opacification,” and 2 for “com-
plete opacification.” APPS scores were assigned for each of
9 pneumatization variants bilaterally, with a score of 0 for
“absent” and 1 for “present.” The scores were summed
bilaterally for a total possible range of 0 to 18. The Lund-
Mackay score was divided by the APPS as a ratio intended
to express the degree of sinus opacification. Simple linear
regression was performed with the calculated ratio as a pre-
dictor of percent opacification by 3D volumetric analysis.
A validation analysis of the Lund-Mackay scoring sys-
tem in axial compared to coronal orientation was then per-
formed. Fifty patients who underwent sinus CT for the
workup of CRS from September 2012 to July 2015 at
the Ochsner Clinic Foundation were included. The axial
group consisted of 25 patients, whereas the remaining 25
patients were included in the coronal group. All scans were
obtained as part of the diagnostic workup of CRS. Five
International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016 2
Assessment of paranasal sinus opacification

FIGURE 2. Schematic diagram of the assignment of sinus CT scans, with relevant inclusion criteria, to the volumetric and validation arms of the study.
CT = computed tomography, DICOM = Digital Imaging and Communications in Medicine.

FIGURE 3. Simple linear regression model of the Lund-Mackay-to-APPS score ratio as a predictor of the percentage of sinus opacification calculated by
3D volume analysis. LMS = Lund-Mackay Score, APPS = Assessment of Pneumatization of the Paranasal Sinuses, 3D = 3-dimensional.

reviewers then scored the degree of sinus opacification us-
ing the Lund-Mackay score. In the axial group, only axial
sections were available to the reviewer, and in the compar-
ison group only coronal images were available. Two sepa-
rate trials were conducted 2 weeks apart. Fleiss kappa was
used to calculate interrater reliability for both trial 1 and
2 separately, and intrarater reliability between trials 1 and
2. The reviewers included a staff physician with fellowship
training in rhinology (E.D.M.), 2 senior otolaryngology res-
idents (M.J.M., C.A.R.), and 2 junior residents (A.S.P.,
J.D.P.). Deidentified CT scans were distributed to the
reviewers using OsiriX, and all scans were viewed us-
ing the bone window algorithm available in OsiriX. A
schematic diagram of the process by which CT scans were
assigned to the volumetric and validation analysis arms is
shown in Figure 2. Statistical analysis was performed using

3 International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016

Marino et al.
TABLE 1. Demographic and radiographic characteristics of validation analysis
Total study
Patient characteristics group (n = 50)
Age (years), mean ± SD 54.8 ± 16.7
Sex
Males, n (%) 18 (36)
Females, n (%) 32 (64)
Trial 1 Lund-Mackay score, mean ± SD 5.78 ± 4.27
Trial 2 Lund-Mackay score, mean ± SD 5.44 ± 4.11
Resolution
High resolution (0.625 mm), n (%) 38 (76)
Low resolution (2.5 mm), n (%) 12 (24)
SD = standard deviation.
SAS software (version 9.3; SAS Institute Inc, Cary, NC).
The Institutional Research Board of the Ochsner Clinic
Foundation approved the study.
Results
The mean ± standard deviation age of the 24 patients in-
cluded in the volumetric analysis was 55.8 ± 17.8 years,
and there was an equal number of males and females. The
mean Lund-Mackay score was 6.52 +/− 4.22), while the
mean APPS score was 9.92 (SD = 2.57). This resulted in a
mean Lund-Mackay-to-APPS ratio of 0.72 +/− 0.55. The
Lund-Mackay-to-APPS ratio was very strongly correlated
with 3D volumetric percentage of sinus opacification by
Pearson correlation coefficient (r = 0.862, p < 0.001). The
95% confidence interval for this correlation was 0.702 to
0.939, and is consistent with strong to very strong cor-
relation. Simple linear regression was performed for this
correlation (Fig. 3), with goodness of fit evaluated by r2 (r2
= 0.743, adjusted r2 = 0.731). The regression model shown
in Figure 3 was used to calculate a “predicted percent opaci-
fication” from the Lund-Mackay-to-APPS ratio (Table S1).
The mean predicted percent opacification was 23.7% +/−
18.1%, while the mean actual percent opacification from
3D volumetric analysis was 23.4% +/− 20.9%. The Lund-
Mackay score was strongly correlated with the percent of
sinus opacification (r = 0.722, p ࣘ 0.001), but less so than
the Lund-Mackay-to-APPS ratio.
The demographic and radiographic characteristics of the
validation portion of the study are shown in Table 1. There
were no statistical differences in age, gender, CT resolution,
or disease extent between the axial and coronal groups
(Table 1). There was moderate to substantial interrater
and intrarater agreement for Lund-Mackay scoring in both
axial-only and coronal-only groups (Table 2). All kappa
values were statistically significant (p < 0.05), and the ex-
tent of rater agreement was assessed using previously pub-
lished guidelines.15 When compared to our previous data of
Axial group Coronal group
(n = 25) (n = 25) p
55.2 ± 17.5 54.3 ± 16.3 0.850
9 (36) 9 (36) 1
16 (64) 16 (64)
6.00 ± 4.17 5.55 ± 4.38 0.407
5.78 ± 3.99 5.10 ± 4.22 0.188
18 (72) 20 (80) 0.509
7 (28) 5 (20)
Lund-Mackay scoring using triplanar images,11 there were
no statistical differences between the 3 groups by corre-
lated 1-way analysis of variance for interrater (p = 0.379)
or intrarater reliability (p = 0.312).
Discussion
There has been understandable interest in developing an
objective CT scoring system that correlates with the ac-
tual percentage of sinus opacification and patient symp-
tom severity in CRS.5, 6, 10 The Lund-Mackay scoring sys-
tem was introduced in 1993,14 and has the virtue of
substantial rater reliability and simplicity for clinician
use.11–13 There have been, however, persistent concerns
that this scoring system does not correlate well with pa-
tient symptomatology.1–6 Newer imaging technologies have
been employed to calculate the exact percentage of sinus
opacification.5–7 These methods have shown improved cor-
relation with patient reported symptoms using quality of
life (QOL) instruments, but sacrifice the ease of use of
clinical metrics such as the Lund-Mackay score. A scor-
ing system that correlates with the exact percent of sinus
opacification, but remains efficient and validated for rater
reliability, has the potential to offer an objective standard
for evaluating disease extent in CRS.
The current study demonstrates that contextualizing the
Lund-Mackay score with TSV as measured by the APPS
score results in a ratio that is very strongly correlated with
the actual percentage of sinus opacification calculated by
3D volume analysis. The Lund-Mackay-to-APPS ratio has
reasonable goodness of fit in simple linear regression. This
demonstrates the utility of this ratio as predictor of actual
percent opacification (Table S1), and may be particularly
useful as an objective measure of sinus opacification on
CT in large outcome studies. Furthermore, the ratio can
be calculated without requirements for additional technol-
ogy or clinician time commitment. The APPS instrument
also tracks 9 anatomic variants, many of which have been
International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016 4
Assessment of paranasal sinus opacification

TABLE 2. Interrater and intrarater reliability for Lund-Mackay items

Interrater reliability Interrater reliability

Intrarater Intrarater
Trial 1 Trial 2 reliability Trial 1 Trial 2 reliability
Sinus (axial) (axial) (axial) (coronal) (coronal) (coronal)

Maxillary 0.74 0.78 0.86 0.76 0.76 0.86

Anterior ethmoid 0.55 0.43 0.62 0.48 0.55 0.65
Posterior ethmoid 0.53 0.61 0.69 0.75 0.69 0.79
Frontal 0.66 0.61 0.77 0.69 0.68 0.68
Sphenoid 0.67 0.59 0.72 0.54 0.56 0.71
OMC 0.42 0.31 0.72 0.61 0.44 0.80
Mean 0.60 0.56 0.73 0.64 0.61 0.75

OMC = osteomeatal complex.

previously recognized as critical items in the evaluation
of sinus CT.16–18 Therefore, the Lund-Mackay and APPS
scores can enable clinicians to use validated metrics to
rapidly evaluate sinus CT for anatomic variation and the
extent of mucosal disease with very strong correlation to
the exact percentage of opacification.
Separately, sinonasal CT imaging has evolved consid-
erably since the introduction of the Lund-Mackay score.
High-resolution images with triplanar reconstructions are
now routinely available. In the study by Oluwole et al.,12
which validated Lund-Mackay scoring for interrater and
intrarater reliability, coronal images with a slice thickness
of 5 mm were used. Our own previous validation of Lund-
Mackay scoring allowed raters to use any of the triplanar
reconstructions available in higher resolution image sets
(0.625 mm or 2.5 mm slice thickness).11 In the present
study, raters were restricted to using either only coronal
or only axial images. Interestingly, there was no statis-
tical difference in rater reliability between coronal-only,
axial-only, or triplanar image sets. These rater reliabilities
were also consistent with those reported by Oluwole et al.12
The APPS instrument is scored only in the coronal orien-
tation, as rater reliability for various proposed anatomic
variants was superior in this orientation vs items scored in
axial or sagittal orientation. This may indicate that coronal
imaging is a reasonable starting point for clinician survey
of anatomic variation and disease extent on sinus CT in
CRS. Nonetheless, triplanar reconstructions are useful for
the examination of specific anatomic variations and disease
presentations.9, 19, 20
The principal weakness of this study is that patient symp-
tom severity is not investigated. While the Lund-Mackay-
to-APPS score ratio very strongly correlates with the actual
percentage of opacification, the correlation of this ratio
with patient-reported symptoms remains a question of in-
terest. That question, however, would be better evaluated
by a prospective study design of a larger number of pa-
tients. Because the Lund-Mackay and APPS scores can be
applied rapidly, several hundred patients could be enrolled
in a future prospective correlation analysis with symptom
severity. The current study is restricted to a smaller pop-
ulation because 3D volumetric analysis of opacified scans
requires up to 1 hour for the complete evaluation of a single
scan. Nevertheless, this data supports that the exact percent
of sinus opacification can be evaluated by radiographic met-
rics, and serves as the basis for the future investigation of
these instruments as objective measures of patient symptom
severity.
Conclusion
The Lund-Mackay-to-APPS score ratio is very strongly cor-
related with the actual percentage of sinus opacification,
and can be performed rapidly by clinicians without special-
ized imaging software. These metrics may offer a validated
method for rapid evaluation of the degree of sinus disease
in addition to anatomic variation. Further study will be re-
quired to investigate if this is predictive of patient symptom
severity.
Acknowledgments
We thank Mariella Gastanaduy BA, MPH, and Qingyang
Luo, PhD, for biostatistical consultation.

References
1. Basu S, Georgalas C, Kumar BN, Desai S. Corre-
lation between symptoms and radiological findings
in patients with chronic rhinosinusitis: an evaluation
study using the sinonasal assessment questionnaire
and Lund-Mackay grading system. Eur Arch Otorhi- and paranasal sinus computed tomographic findings.
nolaryngol. 2005;262:751–754. Arch Otolaryngol Head Neck Surg. 1997;123:1189–
2. Bhattacharyya T, Piccirillo J, Wippold FJ. Relation- 1192.
ship between patient-based descriptions of sinusitis

5 International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016

Marino et al.
3. Smith TL, Rhee JS, Loehrl TA, et al. Objective test-
ing and quality-of-life evaluation in surgical can-
didates with chronic rhinosinusitis. Am J Rhinol.
2003;17:351–356.
4. Stewart MG, Siccard MW, Piccirillo JF, Diaz-
Marchan PJ. Severity staging in chronic sinusitis: are
CT scan findings related to patient symptoms? Am J
Rhinol. 1999;13:161–167.
5. Pallanch JF, Lifeng Y, Delone D, et al. Three-
dimensional volumetric computed tomographic scor-
ing as an objective outcome measure for chronic
rhinosinusitis: Clinical correlations and comparison
to Lund-Mackay scoring. Int Forum Allergy Rhinol.
2013;3:963–972.
6. Likness MM, Pallanch JF, Sherris DA, et al. Computed
tomography scans as an objective measure of disease
severity in chronic rhinosinusitis. Otolaryngol Head
Neck Surg. 2014;150:305–311.
7. Garneau J, Ramirez M, Armato SG, et al. Computer-
assisted staging of chronic rhinosinusitis corre-
lates with symptoms. Int Forum Allergy Rhinol.
2015;5:637–642.
8. Marino MJ, Riley CA, Kessler RH, McCoul ED.
Clinician assessment of paranasal sinus pneumatiza-
tion is correlated with total sinus volume. Int Forum
Allergy Rhinol. (in press). Epub 2016 May 9. doi:
10.1002/alr.21779.
9. Pirner S, Tingelhoff K, Wagner I, et al. CT-based man-
ual segmentation and evaluation of paranasal sinuses.
Eur Arch Otorhinolaryngol. 2009;266:507–518.
10. Rosenfeld RM, Piccirillo JF, Chandrasekhar SS, et al.
Clinical practice guideline (update): adult sinusitis.
Otolaryngol Head Neck Surg. 2015;152:S1–S39.
11. Marino MJ, Weinstein JE, Riley CA, et al. Assessment
of pneumatization of the paranasal sinuses: a com-
prehensive and validated metric. Int Forum Allergy
Rhinol. 2016;6:429–436.
12. Oluwole M, Russell N, Tan L, et al. A comparison of
computerized tomographic staging systems in chronic
sinusitis. Clin Otolaryngol. 1996;21:91–95.
13. Okushi T, Nakayama T, Morimoto S, et al.
A modified Lund-Mackay system for radiological
evaluation of chronic rhinosinusitis. Auris Nasus
Larynx. 2013;40:548–553.
International Forum of Allergy & Rhinology, Vol. , No. , xxxx 2016
14. Lund VJ, MacKay IS. Staging in rhinosinusitis. Rhi-
nology. 1993;31:183–184.
15. Landis JR, Koch GG. The measure of observer agree-
ment for categorical data. Biometrics. 1977;33:159–
174.
16. Deutschmann MW, Yeung J, Bosch M, et al. Radio-
logic reporting for paranasal sinus computed tomog-
raphy: a multi-institutional review of content and con-
sistency. Laryngoscope. 2013;123:1100–1105.
17. Vaid S, Vaid N, Rawat S, Ahuja AT. An imaging
checklist for pre-FESS CT: framing a surgically rel-
evant report. Clin Radiol. 2011;66:459–470.
18. Becker SS. Preoperative computed tomography eval-
uation in sinus surgery: a template-driven approach.
Otolaryngol Clin North Am. 2010;43:731–751.
19. Javer AR, Alandejani T. Prevention and management
of complications in frontal sinus surgery. Otolaryngol
Clin North Am. 2010;43:827–838.
20. Kew J, Rees GL, Close D, et al. Multiplanar re-
constructed computed tomography images improves
depiction and understanding of the anatomy of the
frontal sinus and recess. Am J Rhinol. 2002;16:119–
123.
6