Accepted Manuscript

Title: Introducing a simple method of maxillary sinus volume

assessment based on linear dimensions

Authors: Przystańska Agnieszka, Kulczyk Tomasz, Rewekant

Artur, Sroka Alicja, Jończyk-Potoczna Katarzyna,
Lorkiewicz-Muszyńska Dorota, Gawriołek Krzysztof,
Czajka-Jakubowska Agata

PII: S0940-9602(17)30130-9
DOI: https://doi.org/10.1016/j.aanat.2017.09.010
Reference: AANAT 51191

To appear in:

Received date: 12-10-2016

Revised date: 6-9-2017
Accepted date: 7-9-2017

Please cite this article as: Agnieszka, Przystańska, Tomasz, Kulczyk, Artur,
Rewekant, Alicja, Sroka, Katarzyna, Jończyk-Potoczna, Dorota, Lorkiewicz-
Muszyńska, Krzysztof, Gawriołek, Agata, Czajka-Jakubowska, Introducing a simple
method of maxillary sinus volume assessment based on linear dimensions.Annals of
Anatomy https://doi.org/10.1016/j.aanat.2017.09.010

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Introducing a simple method of maxillary sinus volume assessment based on linear
dimensions

Przystańska Agnieszka 1*, Kulczyk Tomasz 2*, Rewekant Artur 3, Sroka Alicja 4, Jończyk-Potoczna
Katarzyna 5, Lorkiewicz-Muszyńska Dorota 6, Gawriołek Krzysztof 1, Czajka-Jakubowska Agata 1

*
both authors equally contributed to the study
1
Department of Oral Rehabilitation, Division of Prosthodontics, Poznań University of
Medical Sciences, Poznań, Poland
2
Section of Dental Radiology, Poznań University of Medical Sciences, Poznań, Poland
3
Faculty of Physical Education and Health Preservation, State University of Applied
Sciences, Konin, Poland
4
Department of Anatomy, Poznań University of Medical Sciences, Poznań, Poland
5
Department of Paediatric Radiology, Poznań University of Medical Sciences, Poznań,
Poland
6
Department of Forensic Medicine, Poznań University of Medical Sciences, Poznań, Poland

Corresponding author:
Agnieszka Przystańska, PhD, DDS, Department of Oral Rehabilitation, Poznań University of
Medical Sciences, ul. Bukowska 70, 60-812 Poznań, e-mail: aprzyst@ump.edu.pl

Abstract: Measuring sinus volume in a general practice clinic is a complex and time-
consuming procedure, requiring experience in the use of radiological methods. In the
presented research, the automatically estimated maxillary sinus volume was compared with
maxillary sinus volume assessed with mathematical formulas used to calculate the volume of
spheres and pyramids. The starting point for the statistical analysis were specific
measurements of the sinuses. We wanted to discover which geometric shape has the volume
that is nearest to the automatically estimated volume.

The study was performed using samples of CT scans of pediatric patients age 1 to 17. The
dimensions (maximal width, maximal height, maximal length) were used for manual
calculations. For the automatic volume calculation, the CT Image Segmentation algorithm
(Syngo Via for Oncology, Siemens) was used. Pearson's correlation coefficient was applied to
analyse the interrelationship between automatically and manually calculated volume of
maxillary sinus. It was statistically established that the "sphere", "pyramid" and "mean"
manually calculated maxillary sinus volume were accurate and strongly correlated with the
automatically estimated maxillary sinus volume. The volume of the sphere corresponds better
with the automatic measurements than the volume of the pyramid. The variations are
significant and they were made reliable with the application of a statistical test. It is quick and
easy to calculate the maxillary sinus volume based on its linear dimensions instead of
applying advanced software. The manual method for maxillary sinus volume calculation
requires three linear measurements of the sinus (length, width, and height) and can be
recommended if the automatically estimated volume cannot be obtained.

Keywords: maxillary sinus, anatomy, postnatal growth, maxilla, morphometry

1. Introduction
The pneumatization pattern of the human maxillary sinus is of great clinical importance
for head and neck surgeons, otorhinolaryngologists, pediatricians and radiologists. The
accurate assessment of sinus anatomy is significant in pathological diagnosis and treatment
planning. The size, shape and dimensions of the maxillary sinuses significantly changes
during postnatal growth, therefore all surgical procedures require awareness of age-related
changes in the volume. As the volume of the maxillary sinus is the strongest indicator of its
hypoplasia (Bolger et al., 1990), particular emphasis should be given to its role in the
diagnosis and treatment of patients with craniofacial abnormalities manifested by, for
example, undeveloped maxillary sinus (i.e., in patients with cleidocranial dysplasia, the
maxillary sinuses may be small or missing as the result of undeveloped maxilla, Farronato et
al., 2009).
The anatomical dimensions of many structures including the paranasal sinuses can be
measured from CT images (Sahlstrand Johnson et al. 2011). Previous studies have shown that
dimensions of maxillary sinus obtained by CT scans correspond with the measurements on
human skulls (Aryji et al. 1996) thus CT is considered a reliable method for morphometrical
evaluation of the maxillary sinus (Sahlstrand Johnson et al. 2011). The shape and size of the
maxillary sinus can be easily observed in the CT. According to the descriptive anatomy
standards, the shape of the maxillary sinus is referred to as a square pyramid or cube.
Anagnostopoulou et al. (1991) subsequently arranged the sinus into 4 classes, according to its
similarity to solids such as: semi-ellipsoid (class a: 15% of specimens); paraboloid (class b:
30%); hyperboloid (class c: 47%) and cone (class d: 8%). The anatomical characteristics
observed on the CT reconstructed images confirmed the variability in shape of the maxillary
sinus (Kim et al. 2002).
 Many attempts have been made to obtain volumetric information on the paranasal
sinuses. Kawarai et al. (1999) acknowledged the volume of the paranasal sinuses as the most
important index of paranasal sinus evaluation. Aryji et al. (1996) proposed a convenient index
of estimation correlated to sinus volume. The practical use of the maxillary sinus volume has
been described by Bolger et al. (1990), who classified maxillary sinus hypoplasia into types
based on the volume estimate ratio.
Due to its hidden location within the midfacial skeleton, measuring the sinus volume in a
general clinical practice is difficult. The procedure is complex and requires experience in the
use of radiological methods. Some researchers have measured the volume of paranasal
sinuses directly by taking anatomical measurements of dry skulls (Anagnostopoulou et al.
1991) or injecting various materials into the cadaveric sinuses (Schumacher et al.
1972). Modern imaging techniques such as CT scans (Emirzeoglu et al. 2007, Sahlstrand
Johnson et al. 2011, Shah et al. 2003) and MR images (Barghouth et al. 2002) have also been
used for morphometric analysis of the paranasal sinuses. Nonetheless, as the inner contour
of each sinus differs, making an accurate calculation of the volume using CT-based methods
depends on the manual proficiency and experience of the radiologist. Any imprecise mapping
of the sinus outline may influence the automatically calculated volume index.
Finding an accurate manual method of calculating the developing maxillary sinus volume
based on its correspondence to a geometric shape and comparing it with the automatically
calculated volume was the aim of the study.

2. Material and methods

2.1.CT analysis
The multi slice computed tomography (MSCT) scans of patients (aged 0-18 years) from
the database of the electronic system (PACS) of the Karol Jonscher Clinical Children’s
Hospital (Poznań University of Medical Sciences) were retrospectively studied. All patients
who had undergone CT imaging of the skull on suspicion of trauma or neurological disease
were examined on the 128- slice CT scanner Somatom Definition AS+ (Siemens Healthcare)
in the Department of Pediatric Radiology. Patients suffering from neurological diseases or
developmental abnormalities, pathologies in the skeletal system, midfacial injuries or
fractures within the skull and paranasal sinus disease were excluded from the study. Scans
showing unilateral pathologies within the maxillary sinuses were also not included in the
study. Only images described as being normal by radiologists were included in the study.
 The research sample consisted of the CT scans of 170 patients subdivided into 17 groups
based on age. Patients 0-2 years old (less than 24 months of age) were grouped as 1, those 2-3
years old (less than 36 months) as 2, those 3-4 years (less than 48 months) as 3, etc. Finally,
the last group, 17, was composed of the 17-18 year old patients (less than 18 years). Within
each group, the scans of 10 children (5 males and 5 females) were investigated. A total of 340
maxillary sinuses were examined. The study protocol was approved by the University
Bioethical Committee.
For the present analysis, one male and one female CT image, respectively,
were randomly chosen from each age group. The investigated group thus consisted
of 17 females and 17 males aged 0-17 (34 individuals). Assessment of the maxillary sinus in
each patient included bilateral measurements in maximum diameter in three planes:
a) maxillary sinus height (MSH) - maximal craniocaudal diameter of the maxillary sinus
(in vertical plane) defined as the longest distance from the lowest point of the inferior wall to
the highest point of the superior wall as presented on the sagittal image
b) maxillary sinus width (MSW) - maximal transverse diameter of the maxillary sinus
(in horizontal plane) defined as the longest distance perpendicular from the most prominent
point of the medial wall to the most prominent point of the lateral wall as presented on the
axial image.
c) maxillary sinus length (MSL) - maximal anteroposterior diameter of the maxillary
sinus (in sagittal plane) defined as the longest distance from the most anterior point of the
anterior wall to the most posterior point of the posterior wall on the axial image.
To establish which shape the maxillary sinus most resembled we analyzed the data from
anatomy and radiology handbooks and scientific papers. Having concluded that the shape can
be either a pyramid or sphere (ball), calculations for the volume of every sinus based on it
being each of the two shapes were performed. The volume called “manually calculated
Maxillary Sinus Volume” (mMSV) was calculated based on following formulae:
1. V= 4/3 πr3. It was called “sphere” mMSV.
2. V= 1/3 A x h. It was called “pyramid” mMSV.
The mean of 1. and 2. was calculated and called “mean” mMSV.
The results were then compared to, and analyzed with, the automatically calculated MSV
(aMSV). The automatic CT Image Segmentation algorithm (Syngo Via for Oncology,
Siemens) was used. The automatic Detection and Segmentation method uses the difference in
density between the bone and air filling the maxillary sinus. The volume of the sinus was
automatically calculated based on its diagonal. An experienced pediatric radiologist (K. J-P.)
was involved in this part of study. The estimated volume was called Automatically Estimated
Maxillary Sinus Volume (aMSV).

2.2.Statistical analysis
The statistics were produced by the Statistica 10.0 software (Statsoft Inc, USA). The
statistical analysis of the data was made by calculating the mean, standard deviation and
standard error, and the Shapiro-Wilk test was used to test the distribution of analyzed
variables. Because the analyzed variables did not significantly differ from the normal
distribution, the parametric Student t-test for dependent samples was applied to compare the
results obtained by both manually and automatically calculated MSV. The assumed
significance level was 0.05. The null hypothesis (H0) was rejected if p-value < 0.05. The
Pearson product-moment correlation coefficient (PPMCC) was used to analyze the correlation
of the measurements. We applied Pearson’s correlation coefficient to analyze the degree of
interrelationship between automatically and manually calculated volume of paranasal sinuses.
Since no statistically significant difference between left and right maxillary sinuses were
observed we have combined the data of the right and left sides.

3. Results
The comparison of the two samples of the measurements (34 manual and 34 automatic)
showed no statistically significant differences between the volume of maxillary sinus
manually calculated based on the volume of the sphere formula (“sphere” mMSV) and the
automatically estimated volume (aMSV, Table 1). The degree of agreement between
the aMSV and “sphere” mMSV was almost perfect (correlation coefficient r=0.99, coefficient
of determination r2 =0.98). The linear correlation was very strong for comparable
“spheres” mMSV and the aMSV (Table 1, Fig. 1).
The comparison of the two samples of the measurements (34 manual and 34 automatic)
showed no statistically significant differences between the volume of maxillary sinus
manually calculated based on volume of the pyramid formula (“pyramid” mMSV) and the
automatically estimated volume (aMSV). The degree of agreement between the aMSV and
“pyramid” mMSV was also very high (correlation coefficient r=0.99, coefficient of
determination r2 = 0.98). The linear correlation was very strong for comparable
“pyramid” mMSV and the aMSV (Table 1, Fig. 2).
The “sphere” and “pyramid” had normal distribution.
 The comparison of the two samples of the measurements (34 manual and 34 automatic)
showed no statistically significant differences between the manually calculated
“mean” mMSV and the estimated aMSV. The degree of agreement between the aMSV and
“mean” mMSV was almost perfect (correlation coefficient r=0.99, coefficient of
determination r2=0.98). The linear correlation was very strong for comparable “mean”
mMSV and the aMSV (Table 1, Fig. 3).
Although it has been statistically established that all the manually calculated volumes ( the
“sphere”, “pyramid” and “mean” mMSV) strongly correlate with the aMSV, the values of
mMSV were closest to values of aMSV.

Discussion
Our findings show that the volume of human maxillary sinus may easily be estimated by a
simple mathematical calculation based on three linear dimensions of the sinus (length, height
and width). Moreover, the presented method of the maxillary sinus volume estimation does
not require any expensive equipment, time-consuming procedures or sophisticated software.
The “sphere MMSV” and "pyramid mMSV “differ substantially. In all cases, the
"sphere mMSV" is larger than the “pyramid mMSV”, moreover, the "sphere mMSV" is larger
than the aMSV and the “pyramid mMSV” is always smaller. The mMSV of the sphere and
pyramid would seem to be a more accurate measurement than the aMSV. All things
considered, determining the shape of the maxillary sinus is difficult, but, if the volume is
analysed purely mathematically its shape would appear to be somewhere between a sphere
and a pyramid.
Numerous attempts to determine the shape and size of the paranasal sinuses with use of
modern imaging techniques can be found in scientific literature. Most of them (Ariji et al.
1996, Czechowski et al. 2001, Emirzeoglu et al. 2007, Jun et al. 2005, Kawarai et al. 1999,
Kim et al. 2002, Lorkiewicz-Muszyńska et al. 2015, Park et al. 2000, Sahlstrand Johnson et
al. 2011, Spaeth et al. 1997) used computed tomography for morphometrical analysis of the
sinuses, but there have also been reports presenting use of MRI (Barghouth et al. 2002, Lee et
al. 2000). Among the researchers who observed the sinuses on the CT scans, only a few
used the automatic measurement software for maxillary sinus volume evaluation (Sánchez
Fernández et al. 2000, Spaeth et al. 1997).
As derived from the literature, it is much more popular to estimate sinus volume
by calculating it from CT images. Jun et al. (2005) and Kawarai et al. (1999) used
computerized reconstruction to measure the volume of paranasal sinuses quantitatively.
Although they used a similar method, Kawarai et al. (1999) presented higher volume values
than reported in any other study.
Emirzeoglu et al. (2007) obtained the maxillary sinus volume using software combined
with an imaging tool. According to Behar and Todd (2000), the advantage of the manual
method is that it can be applied to any set of printed CT images regardless of access to
computer technology.
The use of dimensions obtained by an imaging technique to calculate the volume of the
sinus based on its resemblance to a geometric shape was presented by Barghouth et al. (2002).
In this study a sinus volume index was computed using the modified (simplified) ellipsoid
volume by analogy to methods used in estimation of prostate volume.
The report of Sahlstrand Johnson et al. (2011) aimed to compare the manually and
automatically estimated sinus volume and to define a formula that helps manual estimation of
the sinus volume. The authors suggested that the manual estimation is easier to perform and
less time consuming. We agree that this is an advantage of manual calculation.
Sahlstrand Johnson et al. (2011) estimated the volume of the maxillary sinuses using an
automatic volume application based on the gas attenuation in the maxillary sinus. Manually,
they calculated the volume using the formula axbxc x 0.5 (a,b,c being the dimensions of the
maxillary sinus in three planes). They obtained an almost perfect (II 0.90-0.93, random error
1.9-2.4 cm3) degree of agreement between the automated measurement of the volume of
maxillary sinus and the volume calculated manually. Although in 87% of cases, the
automatically estimated volume was on average 14-17% greater than that of the manually
calculated volume. We did not observe any such differences in our study. The possible
explanation may be the use of a very accurate method for automatic sinus volume estimation.
There are two methods of automatic sinus volume estimation most often used in CT analysis.
One is based on the determination of the solid block which best fits the shape of the sinus and
from that, the volume of the block is then automatically calculated. The second requires
tracing the inner margin of the sinus cavity from serial slices obtained by CT, and then the
software automatically calculates the volume of the outlined space. This method depends on
the examiner’s experience and is time-consuming. The advantage of Syngo Via for Oncology
software presented in our study is that it is a very simple and quick method of sinus volume
estimation. After the examiner indicates the longest diameter of the maxillary sinus, the
software calculates the volume of the area of the same density adjacent to the measurement.
The subject of further study could be finding a method of establishing the degree of exudation
from the maxillary sinus and comparing it with the amount of fluid in the aerated part of the
sinus. Of course, the limitations are the cost and accessibility to the software Syngo Via for
Oncology, which is dedicated to tumor analysis. As this is not standard software, it is not
accessible to all.
The alternative method of maxillary sinus calculation may be of interest to several
clinical and non-clinical disciplines, especially when access to CT is limited. The advanced
and specialized software that allows precise measurements and automatic calculations of the
volume to be made is not widely available for general use in anthropological research of
archaeological osseous materials. The presented manual method can be used universally in
anthropological studies of skeletal remains. Although such estimation is not suitable for
research purposes, we believe that this simplified manual method might be beneficial in
anthropological practice for estimating maxillary sinus volume, where volume measurement
applications are not available. Moreover, if the 1:1 values of three dimensions are
available, the calculations may be applied to dry skulls, as well as all imaging techniques.
5. Conclusions
The comparison study determined a manual method for maxillary sinus volume
calculation based on its linear dimensions. The method is accurate and comparable with
advanced software calculations. It requires three measurements of the sinus and can be
recommended if the automatically estimated volume cannot be obtained.

6. Conflict of interest:
Authors declare no conflict of interest.
7. Funding:
This research did not receive any specific grant from funding agencies in the public,
commercial, or not-for-profit sectors.
8. Acknowledgments:
Authors would like to thank Mrs. Ruth Hounam for her language support and Dr. Ireneusz
Wrociński for his mathematical help.
References
1. Anagnostopoulou, S., Venieratoss, D., Spyropoulos, N., 1991. Classification of human
maxillary sinuses according to their geometric features. Anat. Anz. 173, 121-130.
2. Ariji, Y., Ariji, E., Yoshiura, K., Kanda, S., 1996. Computed tomographic indices for
maxillary sinus size in comparison with the sinus volume. Dentomaxillofac. Radiol.
25, 19-24.
3. Barghouth, G., Prior, J.O., Lepori, D., Duvoisin, B., Schnyder, P., Gudinchet, F.,
2002. Paranasal sinuses in children: size evaluatuion of maxillary, sphenoid, and
frontal sinuses by magnetic resonance imaging and proposal of volume index
percentile curves. Eur. Radiol. 12, 1451-1458.
4. Behar, P.M., Todd, N.W., 2000. Paranasal sinus development and choanal atresia.
Arch. Otolaryngol. Head Neck Surg. 126, 155-157.
5. Bolger, W.E., Woodruff, W.W., Morehead, J., Parsons, D.S.,1990. Maxillary sinus
hypoplasia: classification and description of associated uncinate process hypoplasia.
Otolaryngol. Head Neck Surg. 103, 759-765.
6. Czechowski, J., Janeczek, J., Kelly, G., Johansen, J., 2001. Radiation dose to the lens
in sequential and spiral CT of the facial bones and sinuses. Eur. Radiol. 11, 711-713.
7. Emirzeoglu, M., Sahin, B., Bilgic, S., Celebi, M., Uzun, A., 2007. Volumetric
evaluation of the paranasal sinuses in normal subjects using computer tomography
images: A steorogical study. Auris Nasus Larynx. 34, 191-195.
8. Farronato, G., Maspero, C., Farronato, D., Gioventù, S., 2009. Orthodontic Treatment
in a Patient with Cleidocranial Dysostosis. Angle Orthod. 79, 178-185.
9. Jun, B.C., Song, S.W., Park, C.S., Lee, D.H., Cho, K.J., Cho, J.H.. 2005. The analysis
of maxillary sinus aeration according to aging process; volume assessment by 3-
dimensional reconstruction by high resolution CT scanning. Otolaryngol. Head Neck
Surg. 132, 429-434.
10. Kawarai, Y., Fukushima, K., Ogawa, T., Nishizaki, K., Gunduz, M., Fujimoto,
M., Masuda, Y., 1999. Volume quantification of healthy paranasal cavity by three-
dimensional CT imaging. Acta Otolaryngol. Suppl. 540, 45-49.
11. Kim, H.J., Yoon, H.R., Kim, K.D., Kang, M.K., Kwak, H.H., Park, H.D., Han, S.H.,
Park, C.S., 2002. Personal-computer-based three-dimensional reconstruction and
simulation of maxillary sinus. Surg. Radiol. Anat. 24, 393-399.
12. Lee CH, Rhee CS, Oh SJ, Jun YH, Min YG, Kim IO. Development of the paranasal
sinuses in children: a MRI study. Korean J. Otolaryngol. Head Neck Surg. 2000, 43,
507-513.
13. Lorkiewicz-Muszyńska, D., Kociemba, W., Rewekant, A., Sroka, A., Jończyk-
Potoczna, K., Patelska-Banaszewska, M., Przystańska, A., 2015. Development of the
maxillary sinus from birth to age 18. Postnatal growth pattern. Int. J. Pediatr.
Otorhinolaryngol. 79, 1393-1400.
14. Park CH, Kim KD, Park CS. Measurement of maxillary sinus volume using computed
tomography. Korean J Oral Maxillofac Radiol 2000, 30: 63-70.
15. Sahlstrand-Johnson, P., Jannert, M., Strombaeck, A., Abul-Kasim, K., 2011.
Computed tomography measurements of different dimensions of maxillary and frontal
sinuses. BMC Med. Imaging. 11, 8.
16. Sánchez Fernández, J.M., Anta Escuredo, J.A., Sánchez Del Rey, A., Santaolalla
Montoya, F., 2000. Morphometric study of the paranasal sinuses in normal and
pathological conditions. Acta Otolaryngol. 120, 273-278.
17. Schumacher, G.H., Heyne, H.J., Fanghänel, R., 1972. Anatomy of the human
paranasal sinuses. 2. Volumetric measurement. Anat. Anz. 130(1), 143-157.
18. Shah, R.K., Dhingra, J.K., Carter, B.L., Rebeeiz, E.E., 2003. Paranasal sinus
development: a radiographic study. Laryngoscope. 113, 205-210.
19. Spaeth, J., Krugelstein, U., Schlondorff, G., 1997. The paranasal sinuses in CT-
imaging: development from birth to age 25. Int. J. Pediatr. Otorhinolaryngol. 39, 25-
40.
Fig 1
Fig 2
Fig 3
Table 1. Comparison of means for manually
and automatically estimated sinus volume.
The data of the right and left sides has been
combined (N = 34).
Means in cm3 are given for each method in
brackets
Method t p-Value
Males
Sphere (11.62) 0.77 < 0.18
Automatic
(9.51)
Pyramid (6.92) 1.56 < 0.08
Mean (9.27) 0.67 < 0.48
Females
Automatic (9.02) Sphere (10.15) 0.89 < 0.28
Pyramid (6.00) 1.52 < 0.07
Mean (8.07) 0.96 < 0.29