Open Jou nal o Radiology, 2016, 6, 39-48
Published Online Ma ch 2016 in SciRes. h p://www.sci p.o g/jou nal/oj ad
h p://dx.doi.o g/10.4236/oj ad.2016.61006
How o ci e his pape : Ka jalainen, M., e al. (2016) Rep oducibili y o 3 mm-Slice-Thick Recons uc ion o Pa anasal Sinus
Compu ed Tomog aphy Scans. Open Jou nal o Radiology, 6, 39-48. h p://dx.doi.o g/10.4236/oj ad.2016.61006
Rep oducibili y o 3 mm-Slice-Thick
Recons uc ion o Pa anasal
Sinus Compu ed Tomog aphy Scans
Ma i Ka jalainen1*, Anna Julkunen2,3*, An i Ma kkola4, P asun Das ida 5, Heini Huh ala6,
Mikko Su inen1,7, Anna-Maija Kuukka1, Ma kus Rau iainen1,8, Ju a Numminen1#,
Sanna Toppila-Salmi2,9#
1Depa men o O o hinola yngology, Uni e si y o Tampe e, Tampe e, Finland
2T ansplan a ion Labo a o y, Haa man Ins i u e, Uni e si y o Helsinki, Helsinki, Finland
3Depa men o O al and Maxillo acial Diseases, Uni e si y o Helsinki and Helsinki Uni e si y Hospi al, Helsinki,
Finland
4Uni e si y o Helsinki and HUS Imaging, Helsinki, Finland
5Medical Imaging Cen e, Depa men o Radiology, Tampe e Uni e si y Hospi al, Finland
6School o Public Heal h, Uni e si y o Tampe e, Tampe e, Finland
7Depa men o O o hinola yngology, Tampe e Ci y Hospi al, Tampe e, Finland
8Depa men o Ea and O al diseases, Tampe e Uni e si y Hospi al, Tampe e, Finland
9Depa men o Alle gy, Uni e si y o Helsinki and Helsinki Uni e si y Hospi al, Helsinki, Finland
Recei ed 19 Janua y 2016; accep ed 23 Ma ch 2016; published 28 Ma ch 2016
Copy igh © 2016 by au ho s and Scien i ic Resea ch Publishing Inc.
This wo k is licensed unde he C ea i e Commons A ibu ion In e na ional License (CC BY).
h p://c ea i ecommons.o g/licenses/by/4.0/
Abs ac
Backg ound: A e he ailu e o medical ea men , he su ge y o ch onic hinosinusi is (CRS) is
planned acco ding o endoscopic and pa anasal sinus compu ed omog aphy (CT) indings. Objec-
i e: The aim o his p ospec i e s udy was o e alua e whe he his s udy me hod migh be eligi-
ble in s udies aiming a adia ion dose educ ion. Sinus CT scans we e chosen as a model because
o he high a ia ion o he adiological ana omy o su gically impo an sinonasal s uc u es. We
hypo hesized ha 3 mm-slice- hick econs uc ion CT had poo ep oducibili y. Me hods: 59 CRS
pa ien s unde wen ou ine mul i-de ec o sinus CT (CTMD). CT3mm was econs uc ed om CTMD
da a-se s. Lund-Mackay (LM) sco es and 43 o he s uc u al pa ame e s we e analyzed blinded.
Ag eemen was s udied be ween CTMD and CT3mm (in a-obse e ep oducibili y), and be ween
h ee obse e s (in e -obse e ep oducibili y) by using Cohen’s kappa. Resul s: The in e -ob-
se e ag eemen was mode a e (kappa 0.4 - 0.6, p < 0.01) in he majo i y o s uc u es o CT3mm
scans. The in a-obse e ep oducibili y o CT3mm scans was e y good in mos s uc u es, how-
e e , i was poo in impo an s uc u es such as on al and spheno-e hmoid ecess, lamina pa-
*Ma i Ka jalainen and Anna Julkunen ha e equally con ibu ed o his wo k.
#
Ju a Numminen and Sanna Toppila-Salmi a e sha ed las au ho s.
M. Ka jalainen e al.
40
py acae, and loca ion o op ic ne e o an e io e hmoidal a e y. The g ade o su geon’s con i-
dence o CT3mm in compa ison o CTMD was lowe (kappa 0.2 - 0.4, P < 0.05). Conclusion: This me-
hodology migh ha e some use in s udies aiming a adia ion dose educ ion. As was expec ed, 3
mm-slice- hick econs uc ion CT had poo ep oducibili y and su geon’s con idence. Mo e ecen
me hods such as cone beam compu ed omog aphy scans ha e nowadays mo e ele an dose e-
duc ion po en ial.
Keywo ds
Ch onic Rhinosinusi is, Compu ed Tomog aphy, Image Recons uc ion, In e -Obse e Ag eemen ,
In a-Obse e Ag eemen , Kappa, Radia ion-Dose Reduc ion, Rep oducibili y, Pa anasal Sinus,
Sinus Imaging
1. In oduc ion
Ch onic hinosinusi is (CRS) is a common, mul i ac o ial and a iable disease wi h a p e alence o a ound 10%
- 16% [1]. The diagnos ics o CRS is based on ypical symp oms and clinical indings [1]. Compu ed omog a-
phy (CT) scans a e he imaging modali y o choice con i ming he ex en o pa hology and he need o su ge y.
The main indings in CRS a e mucosal changes wi hin he os eomea al complex and/o sinuses. O he cha ac e-
is ic indings a e ai - luid le els, mucosal hickening and opaci ica ion o he no mally ae a ed sinus lumen.
The only CRS indica ing change can be scle o ic, hickened bone o he sinus wall [1]. The numbe o ana omi-
cal a ian s is e y high in pa anasal sinuses [2] [3]. Se e al o hem a e loca ed close o sinonasal su gical a ea
and hus he in aope a i e lesions o hem may lead o se e e illness o be li e h ea ening. The ope a i ely im-
po an s uc u es a e inse ion o middle u bina e and he uncina ed p ocess, he loca ion o an e io e hmoidal
a e y, and Ke os class, In ao bi al cells, posi ion o he Agge nasi cell, and he ana omical a ian s loca ed in
he ope a i e a ea [4]-[6]. S udies demons a e ha he e is dose- educ ion po en ial o CT scans [7]-[10].
The aim o his p ospec i e s udy was o e alua e whe he ou me hod migh be eligible in s udies aiming a
adia ion dose educ ion. Sinus CT scans we e chosen as a model because o he high a ia ion o he adiologi-
cal ana omy o su gically impo an sinonasal s uc u es. We hypo hesized ha 3 mm-slice- hick econs uc ion
o sinus CT scans had poo dose- educ ion po en ial and in e -obse e ag eemen .
2. Ma e ials and Me hods
2.1. Pa ien s
This p ospec i e coho s udy was ca ied ou in he Depa men o O o hinola yngology a Tampe e Uni e si y
Hospi al, Finland om 2006 o 2015. The s udy was app o ed by he e hical commi ee o he Pi kanmaa Hos-
pi al Dis ic (no 96032). W i en in o med consen was ob ained o each pa icipan . A andom sample o
59CRS pa ien s, who we e e alua ed o bene i om sinus CT scans du ing 2006-2007, was en olled o his
s udy. Pa ien da a was collec ed om medical eco ds and by a ques ionnai e a he ime he sinus CT was pe -
o med. The ollow-up da a o sinonasal ope a ions and ime o su ge y was collec ed o m pa ien eco ds o
Tampe e Uni e si y Hospi al o Tampe e Ci y Hospi al in 2015 in a e age 9 yea s a e he ime o pe o ming
sinus CT scans. None o he subjec s had used Aspi in desensi iza ion, alle gen immuno he apy, o an i IgE
he apy p io o o du ing he sinus CT scans o du ing he ollow-up.
2.2. CT Scans
The pa ien s unde wen ou ine sinus mul iple de ec o compu ed omog aphy (CTMD) examina ions o clinical
pu poses. Two di e en CTMD machines we e used: GE Ligh Speed 16 (GE Heal hca e, Milwaukee, Wisconsin)
and Philips B illiance 64 (Philips, Bes , he Ne he lands). The pa ien s we e imaged in supine posi ion wi h a
kilo ol age o 120 kV and a milliampe e second o 100 mAs. In he GE machine, he slice hickness was 0.625
mm wi h co onal econs uc ions a 1.5 mm. In he Philips machine, he slice hickness was 0.9 mm wi h co onal
econs uc ions a 0.9 mm. Bo h we e h ee dimensional (3D) in na u e wi hou any gap. In all cases, he imag-
M. Ka jalainen e al.
41
ing was pe o med using a bone il e echnique. The imaging co e ed he en i e sinonasal a ea bo h in he axial
and co onal di ec ions. The co onal e o ma ions spanned h ough he en i e a ea, s a ing om he an e io wall
o he on al sinuses and ending o he le el o he pos e io wall o he sphenoid sinuses. We pe o med 3
mm-slice- hick econs uc ions (CT3mm) om co onal, sagi al and axial da a se s o CTMD. The co onal e o ma-
ions in he CTMD and CT3mm spanned h ough he en i e a ea, s a ing om he an e io wall o he on al si-
nuses and ending a he le el o he pos e io wall o he sphenoid sinuses.
2.3. E alua ion o CT Scans
CTMD and CT3mm scans we e obse ed by h ee independen obse e s blinded o each o he and o he pa ien
his o y da a. The obse a ion o he same pa ien ’s scans occu ed a leas one week apa . The ocus was o
compa e ag eemen be ween CTMD and CT3mm scans and be ween h ee obse e s in e alua ing CRS- ela ed
changes and adiological ana omy o he s uc u es. The obse e s we e an expe ienced head and neck adiolo-
gis (AM), an expe ienced ENT- and hinosu geon (JN), and a i h yea ENT esiden (ST-S). They illed in a
49 -i em o m o sinonasal s uc u es om bo h scans and om bo h sides o each pa ien (Table 3). Simila
o ms we e illed om CTMD and CT3mm scans. E alua ion o he same pa ien ’s scans occu ed a leas 7 days
apa . All asked s uc u es had 2 - 5 di e en choices. Be o e s a ing he e alua ion o he CT scans, all choices
we e ca e ully discussed by he obse e s. Obse e s also made a pilo o 15 CT scans in o de o make su e ha
all obse e s unde s ood how o ill in he o ms. The Radiologis did no espond o he ques ions: “Need o
sep oplas y” and “G ade o su geon’s con idence based on images”.
2.4. Da a Analysis
S a is ical analysis was ca ied ou by he SPSS Base 15.0 S a is ical So wa e Package (SPSS Inc., Chicago, IL,
USA). Cohen’s kappa was used o compa e he deg ee o ag eemen o CTMD and CT3mm scans (e.g. in aob-
se e ag eemen ); and he in e -obse e ag eemen o CT3mm scans. The calcula ion is based on he di e ence
be ween how much ag eemen is ac ually p esen compa ed o how much ag eemen would be expec ed o be
p esen by chance alone. The es ablished in e p e a ion o Kappa- alue is classi ied in o 6 subg oups: Poo ≤ 0.2,
Fai 0.21 - 0.4, Mode a e 0.41 - 0.6, Good 0.61 - 0.8, and Ve y Good 0.81 - 1.0. A alue unde ze o means ha
he ag eemen is wo se han by chance, and he alue ange is om −1 o +1. Two- ailed P- alues o <0.05 we e
conside ed s a is ically signi ican .
3. Resul s
3.1. Pa ien Cha ac e is ics
The cha ac e is ics o pa ien s a e shown in Table 1. The mean (min-max) age was 43 (13 - 77) yea s by he
ime o aking he CT scans. 54.2% o pa ien s unde wen sinonasal ope a ion wi hin a yea a e he CT scans
we e pe o med (Table 1). O hese pa ien s ha we e ope a ed a he ime o CT scans, 7 (21.9%) unde wen
e ision sinonasal su ge y in a e age 3 yea s la e . 47.5% o pa ien s epo ed su e ing om diseases (o he
han CRS) wi h egula need o medica ions. The mos equen diseases we e (numbe o pa ien s): hea and
ascula diseases [11], hypo hy eosis [8], mig aine [3] and a h osis/a h i is [4].
3.2. In e -Obse e Ag eemen o CT3mm Scans
Lund-MacKay sco es (Table 2) and 43 o he s uc u es we e e alua ed by a adiologis , an ENT su geon and an
ENT esiden om bo h CTMD and econs uc ed CT3mm scans (Table 2, Table 3). The in e -obse e ag eemen
o CT3mm scans was mode a e (kappa 0.4 - 0.6) in he majo i y o s uc u es (Table 3). In CT3mm scans, he su -
gically impo an s uc u es wi h g ea es disag eemen we e ound in Ke os classi ica ion ( adiologis s. ENT
esiden ), he loca ion o an e io e hmoidal a e y and op ic ne e (all obse e s), p ominen e hmoid bulla
(ENT esiden s. ENT su geon), and hickness and con ac o middle u bina e o o bi al lamina o e hmoidal
bone ( adiologis s. ENT su geon) (Figu e 1, Table 3).
3.3. In a-Obse e Ag eemen
We compa ed he deg ee o ag eemen be ween CTMD and CT3mm scans o he Lund-MacKay sco es and 43 o h-
M. Ka jalainen e al.
42
Table 1. Cha ac e is ics o he pa ien s.
Ch onic hinosinusi is pa ien s
n (59)
%
Gende
Male
27
45.8
Female
32 54.2
Age ≥45 yea s 35 59.3
<45 yea s 24 40.7
Smoking
No
30
50.8
Ex 11 18.6
Cu en 15 25.4
Unknown 3 5.1
Alle gic hini is
No
34
57.6
Yes 23 39.0
Unknown 2 3.4
As hma
No 45 76.3
Yes 12 20.3
Unknown 2 3.4
Nasal polyps
No 42 71.2
Yes 15 25.4
Unknown 2 3.4
Du a ion o symp oms
<1 yea 8 13.6
1 - 4.9 yea s 23 39.0
5 - 14.9 yea s 14 23.7
≥15 yea s 11 18.6
Unknown 3 5.1
An ibio ic cou ses du ing he pas 2 yea s
No 11 18.6
1 - 3 22 37.3
4 - 9 14 23.7
≥10 10 16.9
Unknown 2 3.4
Cu en use o in anasal co icos e oids
No 15 25.4
Yes 42 71.2
Unknown 2 3.4
≥1 pe o al co icos e oid cou se(s) du ing he pas 1 yea
No 55 98.2
Yes 1 1.8
Unknown 3 5.1
P e ious sinonasal ope a ion(s) No 45 76.3
Yes 14 23.7
Radiological signs in CT3mm scans o p e ious sinus ope a ion
No
46
78.0
Yes
13
22.0
Sinonasal ope a ion pe o med wi hin a yea a e he CT scans
No
25
42.4
Yes
32
54.2
Unknown
2
3.4
≥1 sinonasal ope a ion(s) du ing he 9-yea ollow-up
No
47
79.7
Yes
10
16.9
Unknown
2
3.4
Cu en symp oms by VAS, mean (min - max)
Sense o smell
3.2
(0 - 10)
Pos -nasal d ip
5.0
(0 - 10)
Obs uc ion
5.5
(0 - 10)
Facial pain
4.9
(0 - 10)
Abb e ia ion: VAS = isual analogue scale (0 - 10).
M. Ka jalainen e al.
43
Table 2. Medians and in e qua ile anges (Q1 - Q3) o he adiologic Lund McKay (LM) sco es e alua ed by he adiologis
om mul i-de ec o sinus compu ed omog aphy scans (CTMD) om 59 pa ien s su e ing om ch onic hinosinusi is.
LM sco e Righ Le
Median Q1 - Q3 Median Q1 - Q3
F on al sinus 0 0 - 1 0 0 - 1
An e io e hmoidal sinus 1 0 - 1 0 0 - 1
Os iomea al uni 0 0 - 2 0 0 - 2
Pos e io e hmoidal sinus 1 0 - 1 0 0 - 1
Sphenoid sinus 0 0 - 1 0 0 - 1
Maxilla y sinus 0 0 - 1 1 1 -1
LM sco e o al 3 1 - 6 4 1 - 5
Figu e 1. Two cases su e ing om ch onic hinosinusi is and who unde wen ou inely mul i-de ec o sinus compu ed o-
mog aphy scans (CTMD) ((A) (C) (E)). CTs wi h 3 mm slice hickness (CT3mm) we e econs uc ed om CTMD da a-se s in
o de o e alua e adia ion dose educ ion po en ial ((B) (D) (F)). (A) Pa ien 1, CTMD, F on al ecess and os ium. (B) Pa ien
1, CT3mm, ep esen a i e slice om F on al ecess and os ium wi h sligh ly limi ed isualiza ion. (C) Pa ien 1, CTMD, Os i-
omea al complex (D) Pa ien 1, CT3mm, ep esen a i e slice om Os iomea al complex wi h sligh ly limi ed isualiza ion. (E)
Pa ien 2, CTMD, An e io E hmoidal a e y sulcus (F) Pa ien 2, CT3mm, ep esen a i e An e io E hmoidal a e y sulcus, wi h
s ongly limi ed isualiza ion.
e s uc u es e alua ed by he adiologis . In gene al, he in a-obse e ag eemen was e y good (kappa 0.8 -
1.0) in he majo i y o s uc u es, such as Lund Mackay sco es (Table 4). Ye , se e al su gically impo an
s uc u es did no show adequa e ag eemen : OMC-a ea, on al and spheno-e hmoid ecess, lamina papy acae,
and op ic ne e (Table 4). An e io e hmoidal a e y was esponded o be unde ec able in 95% o he CT3mm
scans, bu only in 13% o he CTMD scans (Figu e 1).

M. Ka jalainen e al.
44
M. Ka jalainen e al.
45
M. Ka jalainen e al.
46
M. Ka jalainen e al.
47
3.4. Ope a i e Con idence o CT3mm Scans
ENT su geon and ENT esiden es ima ed ope a i e ce ain y o CT3mm in compa ison o CTMD. ENT su geon
esponded ha he ope a i e con idence is good in 69.6% o CTMD scans, and 64.3% o CT3mm scans. The in a-
obse e ag eemen in he g ade o ENT su geon’s con idence o CT3mm in compa ison o CTMD was mode a e
(kappa 0.4, P = 0.005). ENT esiden esponded ha he ope a i e con idence is good in 73.2% o CTMD scans,
bu only in 37.5% o CT3mm scans. The in a-obse e ag eemen in he g ade o ENT esiden ’s con idence o
CT3mm in compa ison o CTMD was hus poo (kappa 0.2, P = 0.035).
4. Discussion
This s udy was ca ied ou o e alua e whe he he used me hod would be eligible when s udying adia ion dose
educ ion po en ial. As an example, we compa ed wo echniques o pa anasal sinus CT scans. F om he CT
scans we e alua ed impo an s uc u es o CRS diagnos ics and ope a ion. By e o ma ing he CT3mm images,
we we e able o e alua e he same pa ien exac ly wi hou addi ional imagining. Pa ien selec ion was pe o med
andomly om a doc o ’s ecep ion. A he e ogeneous pa ien g oup made he ex apola ion o he esul s possi-
ble o clinical p ac ice
We ound ha in e -obse e ag eemen was only mode a e in CT3mm scans. This inding is in line wi h ou
unpublished da a ha in e -obse e ag eemen is a simila le el also in he con en ional CTMD scans. This
migh be due o he ac ha sinonasal ana omy is highly a iable. In a-obse e ag eemen be ween CT3mm and
CTMD scans was e y good in mos s uc u es. Howe e , i was ai o poo in se e al su gically impo an
s uc u es. The g ade o young su geons’ con idence was ela i ely good wi h CTMD scans whe eas he g ade o
con idence was poo wi h CT3mm scans. Taking oge he , CT3mm scans seemed no o be clinically ele an im-
aging me hod.
P e iously, i has been shown ha CRS diagnos ics is possible wi h o he educed adia ion CT- echniques.
Howe e , he su gical aspec s ha e no been deal wi h in hese s udies [7]-[9]. Ou cu en s udy showed ha i
was possible o educe he adia ion dose and inc ease slice hickness wi hou comp omising he excellen bone
and so issue con as . Howe e , success ul endoscopic sinus su ge y (ESS) equi es de ailed knowledge o he
highly a iable ana omy o he nasal ca i ies, he os iomea al uni , and he skull base. In addi ion, he i al
s uc u es especially he op ic ne e and he an e io e hmoid a e y and hei ela ionship o ope a ional a eas
need o be clea .
Mul i-de ec o CT echnology comp ises mul iple de ec o s ha a e equi alen o slices. Con en ional CTMD
comp ises consecu i e slices wi h a hickness o 0.6 - 1.5 mm and causes a adia ion dose o 0.5 mS on a e age.
Technological imp o emen has been apid esul ing in 4- o 8-, 16-, 32-, 40- and 64-de ec o machines. I is
known ha he e is no sa e adia ion dose. In all exposu e quan i ies, he e is a co esponding mu a ional isk
[11]. The head and neck egion, he eyes and hy oid gland a e he mos adiosensi i e o gans. In addi ion o he
cance isk, adia ion-induced ca a ac s a e possible a e mul iple exposu es [3]. Image quali y is a ec ed by
image noise o quan um mo le and is almos always di ec ly ela ed o he adia ion dose. Keeping his in mind,
he cone beam compu ed omog aphy seems o be ele an ecen implemen a ion in CT imaging. Essen ially, i
could be he bes comp omise so a be ween low adia ion and excellen image quali y [12]-[15].
As a conclusion, ou s udy demons a ed ha his me hodology was easy o use and migh ha e some use in
s udies aiming a adia ion dose educ ion. As was expec ed, 3 mm-slice- hick econs uc ion CT had poo e-
p oducibili y and su geon’s con idence. Mo e ecen me hods such as cone beam compu ed omog aphy scans
ha e nowadays mo e ele an dose educ ion po en ial.
Acknowledgemen s
The au ho s hank esea ch nu se Ma ja-Leena Oksanen o he excellen assis ance. The s udy was suppo ed in
pa by esea ch g an s om he Ahokas Founda ion, Compe i i e Resea ch Funding o he Tampe e Uni e si y
Hospi al (G an s 9H067, 9J108, 9L087), he Finnish Medical Socie y Duodecim, he Finnish Socie y o Alle -
gology and Immunology, Finnish Socie y o O o hinola yngology-Head and Neck Su ge y, Helsinki Uni e si y
Cen al Hospi al Resea ch Funds, he Ida Mon i Founda ion, he Jane and Aa os E kko Founda ion, he Väinö
and Laina Ki i Founda ion, and he Y jö Jahnsson Founda ion.