disease of the head and neck, or skull involvement in any of these patients. During a period of 18 mo 34 patients had. SPECT of the skull in addition to planar ...
CLINICAL
SCIENCES
Normal and Abnormal Single Photon Emission Computed Tomography ofthe Skull: Comparison with Planar Scintigraphy Ora Israel, Jacqueline Jerushalmi, Alexander Frenkel, Abraham Kuten, and Dov Front DepartmentsofNuclear Medicine and Oncology,Rambam Medical Center,and the Technion- Israel Institute ofrechnology, Haifa, Israel
Usinga rotatinggammacamerathenormalsinglephotonemissioncomputedtomography (SPECT) anatomy of the skull was defined in eight subjects. The value of SPECT as
compared with planar sdn@graphywas assessed in 34 patients with known or suspected
diseaseof theskull.Sevenpatientshadnormalplanarscintigraphy andSPECT.In12of 27 patientswithboneinvolvement SPECTandplanarscintigraphy showedessentially thesame
findings.In15 patientsSPECTwassuperiorto planarscintigraphy. Inthreeof thesepatients SPECT detected lesions while planar scan was normal. In the other 12 patients SPECT
showed better anatomiclocalizationand definedthe fullextent of the lesion. Thiswas most obviousinpatientswithinvolvement of sphenoid,petrous,divus,maxilla,andzygomatic bones. Our findings confirm the potential of SPECT to detect lesions in deep bones that are
overlapped bysuperficial bonystructuresthatcannotbevisualized cleatlywithplanar scintigraphy. J Nuci Med 29:1341—1346, 1988
he complex structure of the skull that is made of 22 bones may cause difficulties in detection and local ization of lesions on planar bone scintigraphy.
Single
photon emission computed tomography (SPECT) has the potential to obtain improved images of the bony structures. SPECT has been reported to be useful in some patients with facial bone lesions or malignant otitis externa (1—4).In the present study we define the normal anatomy of the skull as it appears on images obtained from a gamma camera with a rotating gantry
by a combination of clinical findings,radiologicevaluation, planarscintigraphyand follow-upof 6 mo or more. Planar bone scintigraphy of the skull was performed 2 to 3 hr after 25 mCi (925 MBq) oftechnetium-99mmethylenediphospho nate was injected i.v. A large field-of-view, digital camera (Elscint, Apex 415) was used to obtain 600,000 counts on a
256 x 256 bytematrixfor eachviewof the skullin the anterior,posterior,right,and left lateralprojections.SPECT examinations were performed with the same camera using a rotating gantry and a high resolution collimator. Our tech nique and equipment for SPECT have been described previ
and compare the value of SPECT with routine planar
ously(5.6)andonlythealterationsdoneforskullinvestigation
scintigraphy for the detection of skull lesions.
will be described here. Data were acquired during 20 mm over
MATERIALS
obtained in a 64 x 64 byte matrix using an angle step of 3.
AND METhODS
Eight patients examined for stress fractures were studied as controls to define the normal SPECT anatomy of the skull. There was no indication of generalized bone involvement, disease of the head and neck, or skull involvement in any of these patients. During a period of 18 mo 34 patients had SPECT of the skull in addition to planar scintigraphy. Table 1 lists the number of patients and their diagnosis. Twenty seven patients had involvement ofthe skull that was diagnosed Received Sept. 9, 1987; revision accepted Jan. 14, 1988.
For reprintscontact:Dov Front, MD, PhD, Dept. of Nuclear Medicine,RambamMedicalCenter,Haifa35254;Israel.
Volume29• Number8 • August1988
360 of rotation.Onehundredandtwentyprojections were Theacquisitiontimewas 10sec andcountdensity70,000for each projection. Reconstruction was accomplished by back projection using a Wiener filter (7,8). The filtered backprojec tion data was used to reconstruct 32 seven-mm-thick (1 pixel) transaxial,coronal, and sagittaltomograms. The reconstructed sections of the normal skull were compared with the cross sectional anatomy of cadavers to define the normal bone anatomy on SPECT (9,10). RESULTS
The normal SPECT anatomy of the skull in the transaxial,
sagittal and coronal
sections is shown in
I 341
TABLE 1
Correlation of AbnOrmalFindings in Planar Scintigraphy and SPECT of the Skull in 27 Patients Advantage of SPECT over planar
planar
SPECTDiagnosis
No.ofSimilar and patients(No.
Carcinomaof nasopharynx
pts)
6
1
(No. pts)
Bones - No.
5
Sphenoid -5 Petrous -2 Clivus-2 Maxilla-2 Zygoma-1
Carcinoma ofear
4
4
-4Sphenoid Petrous -1
2
Petrous -2
1
1
-1Sphenoid Sphenold
4
2
-2Petrous
1
1
-1Sphenoid
Malignant otitis extema
2
Carcinoma of mandible
2
2
Carcinoma of maxilla
2
Osteomyelitis of skull vauft
2 26
Othermalignant tumors Hlstlocytosest
2
-1
Fibrous displasia of vauft
I
1
.This group included one carcinoma ofthe orbit, carcinoma ofthe larynx, carcinoma ofthe lower lip, carcinoma ofthe palate,
malignantmelanomaof the eye, metastaticbreastcarcinoma. t Tl@
group
induded
the
patients
with
Histiocytosis
x and
eosinophillc
granuloma.
Figure 1. There was no area of increased activity in the skull in any ofthe eight control patients. Seven patients
and coronal sections is important for defining the extent of spread of lesions in patients with tumors or bone infections (Figs. 2—5). The improved contrast of SPECT with disease of the head and neck had normal planar and SPECT scintigraphy.The resultsofthe comparison enhances the high sensitivity of planar bone scintigra of planar scintigraphy and SPECT in 27 patients are phy. Lesions in three patients that were not evident on summarized in Table 1. In 12 patients with lesions in planarscintigraphywere detected by SPECT. The main the bony structures, SPECT and planar scintigraphy advantage of SPECT was in better visualization of showed essentially the same findings. SPECT was su penor to planar scintigraphy in providing better ana
tomic localization and in detecting the whole extent of the lesion in 15 patients. In three of these patients SPECT
detected
lesions in the sphenoid and petrous bone that are deeply situated in the skull and superimposed by other struc tures. SPECT did not show any advantage in the super
ficial bones, vault of the skull or the mandible.
lesions which were not visible on
planar scintigraphy. When involvement of individual
DISCUSSION
bones was analyzed in the 15 cases with superior SPECT
it was found that it included ten cases of involvement of the sphenoid, nine of the petrous, two of the civus, two ofthe maxilla and one ofthe zygomatic bone (Table
1, Figs. 2—5). Knowledge ofskull anatomy in the transaxial,sagittal
1342
Israel, Jerushalmi, Frenkel etal
Bone scintigraphy is an extremely sensitive method for detection of bone pathology and is considered the in vivo gold standard to which other methods are compared.
SPECT adds an improved contrast and bet
ter defines the anatomic localization and extent of bone
TheJournalof NuclearMediome
3@,f1IlI@ d@
75
‘@* .@9
@
4'
A
‘@* ‘V
—1214 169
18@j@@:71 $@I@J
a @ @
20 21
17
@il@II@
29 26‘
14@ 17@@I
27@6
—3
*18@@ FiGURE 1 NormalSPECTanatomyof the skull. 1. Frontalbone. 2. Frontalbone and
sinus.3. Temporalbone.4. OccipItal bone. 5. Zygomatlcbone. 6. Sphe noldal bone. 7. EthmOIdaIcells. 8.
b
Clivus.9.Petrousbone.10.MaxIllary sinus. 11. Infratemporalfossa. I2. Mastokiprocess.13.Temporo-man dibularjolntwith mandibularcondyle. 14
—4
14. Maxillary bone. 15. Remus of
I,
mandibula@ 16.OccipItalcondyle.17. MaIdbUIa.18.Firstandsecondcar vical vertebrae. 19. Ethmoidalbone andnasalconcheae.20. OrbItalroof.
21. Orbfta.22. Orbitalfloor.23. Fronto-zygomatic process. 24. Hard
palate.25. Parietalbone.26. Sphe noldsinus.27. Greaterwingof sphe noid. 28. OccipItal bone (foramen magnum).29. Zygomatic arch. a. Transaxial slices. b. Coronal slices. C
Volume 29 • Number 8 • August1988
C. Sagittal slices.
1343
@
a@'
.4
1
b
I
I
At'4T
‘‘a
c.@ FIGURE 2
d
SPECT visualizationof petrousboneinvolvement.a: Planarsontigraphyin a 64-yr-oldmalepatientwith malignantotitis
extemashowsonlyabnormaluptakein the righttemporalregion.b: SPECTShOWS a largelesionin the petrouspart of the temporal bone. Involvementof the petrous part, evident only on SPECT, was proven by surgery. C: Planar
scintigraphyin anotherpatientshows essentiallythe sameregionof abnormaluptake as in a. SPECT(d), however, showsthat the lesionin this patientis superficial.The normalpetrousbonewas detectedonly by SPECT.
$
@
,
I b FIGURE 3
SPECTvisualization of sphenoidboneinvolvement. Carcinoma of thenasopharynx withboneinvolvement in a 14-yr old boy. a: Planarscintigraphyshows abnormaluptake in the nght maxillaand zygomaticbone. b: SPECTShOWS
involvement of thenghtmaxillaandextendedinvolvement of thesphenoidbonewhichcannotbedetectedon planar scintigraphy.
1344
Israel, Jerushalmi, Frenkel etal
TheJournalof NudearMedlcine
Rt
Lt
I FIGURE 4 Carcinomaof the nasopharynxin a 65-yr-oldfemalewith involvementof the facialbones.a: Onlypoorlydefineduptake is seen on planar Scintigraphy. b: SPECT visualization of bone involvement of the maxilla, zygomatic and temporal bones.
I
,@
@1. Rt
Lt
@p_.,. @i
@
a
I
b
FIGURE 5 Histiocytosis X in a 35-yr-old male. a: Planar scintigraphy shows focal uptake in the frontal and temporal regions. The
exact locationof the bonelesioncannotbe defined.b: SPECTshowsthe involvementof the deeplysituatedsphenoid and petrousbones.
Volume29• Number8 • August1988
1345
lesions. Modern high resolution
rotating cameras pro
vide good definition of the normal anatomy by sepa ration ofthe various bones and regions ofthe skull (Fig.
1). Properpositioning ofthe head is important for good visualization
of bony structures
and shifting of the
patient's head can make interpretation difficult. Dc Roo et al. (1 1), using transaxial slices only, were also able to detect lesions that were not evident on planar scintigraphy. However, as we have shown, (Figs. 2-5) the sagittal and coronal planes are necessaryto
define the complete extent of the lesion. We believe that SPECT of the skull should be added to planar scintigraphy when: (a) a skull lesion is sus pected but planar scintigraphy appears normal or equiv ocal; (b) and when information is necessary about the exact localization and extent of a lesion involving deeply located bones before surgery or radiotherapy.
Med 1977;18:1184—1188.
2. CollierBD, CarreraGF, MesserEl, et al. Internal derangements ofthe temporomandibularjoint. Detec tion by single photon emission computed tomography.
Radiology 1983;149:557—561. 3. StrashumAM, NejatheimM, GoldsmithSJ.Malig nant otitis externa:early scintigraphicdetection. Ra diology1984; 150:541—545.
4. O'MaraRE. The roleof bonescanning in dentaland maxillofacial disorders. In: Freeman LM, Weissmann
HS, eds. Nuclearmedicineannual.New York:Raven Press, 1985:265—284.
5. Front D, Israel0, Even-SapirE, et al. Assessmentof superior sagittal sinus thrombosis with Tc-99m la belied red blood cells.. Radiology 1986; 158:453—456.
6. Front D, IosilevskyG, Even-SapirE, et al. In vivo quantitation using SPECT ofradiopharmaceutical up take by human meningiomas. Radiology 1987;
164:93—96. 7. King MA, Schwinger RB, Doherty PW, et al. Two dimensional filtering ofSPECT images using the Metz
and Wienerfilters.J NuclMed 1984;25:1234—1240. ACKNOWLEDGMENT The authors thank Lionel Liberman, MD, PhD, for kindly reviewingthis manuscript and providing many useful sugges
tions.
8. King MA, Doherty PW, Schwinger RB. A Wiener
filter for nuclear medicine images. Med Phys 1983; 10:876—880. 9. PotterGD. Sectionalanatomyand tomographyof the head. New York, London: Grune & Stratton, 1971.
10.Maue-DicksonW, DicksonDR. PullenFW, eds. Computed tomographic atlas of the head and neck.
REFERENCES
1. BrownML, KeyesJW, LeonardPF, and KinosTL. Facialbone scanningby emissiontomography.J Nucl
1346
Israel, Jerushalmi, Frenkel etal
Toronto:Little,Brownand Company, 1983. 11. Dc Roo M, Mortelmans L, Devos P, et al. Single photon
emission
computerized
tomography
of the
skull.Nucl Med Commun 1985;6:649—656.
TheJoumalof NudearMedicine
