1113

Reduction of Absorbed
Doses in Radiography of the
Facial Skeleton

Per JuIin1 Radiation absorbed doses from radiography of the paranasal sinuses and the facial
Thomas Kraepelien2 skeleton were measured with thermoluminescent dosimeters (TLD) on a phantom head
using high-sensitivity screens in an Orbix stand. The entrance doses to the skin of the
head ranged from 0.31 to 2.9 mGy per exposure. The absorbed dose from a full series
of sinus exposures averaged 0.33 mGy for the oral mucous membrane, 0.33 mGy for
the maxillary sinus mucous membrane, 0.11 mGy for the parotid gland, 0.15 mGy for the
submandibular gland, 0.61 mGy for the eye lens, and 0.75 mGy for the thyroid gland
region. A leaded soft collar adapted to the thyroid region reduced the thyroid doses by
more than one order of magnitude, but also reduced the image field. The mean energy
imparted from a full series of paranasal sinus projections was 4.8 mJ and from a total
series of the facial skeleton, 7.9 mJ.

In radiography of the facial skeleton, several critical organs such as the salivary
glands, the thyroid gland, and the eye lens are irradiated. It is therefore necessary
to reduce the absorbed doses to these organs as far as possible, bearing in mind
the necessary image quality [1 ]. Several approaches to this problem, such as high-
sensitivity screen-film combinations [2], increased tube voltage, and improved
collimation, have been recommended. It is of interest to investigate the absorbed
doses to different critical organs with such techniques and to evaluate the possi-
bilities of further dose reduction. At the Department of Oral Radiology of the
Huddinge Clinics, a Siemens Orbix stand has been used with extra primary
collimation, added filtration, and high-sensitivity screens. On the basis of our
experience over the past 2 years, the image quality has been satisfactory for
diagnostic purposes.
We investigated the doses to critical organs in the head and neck region from
radiography of the facial skeleton using this exposure technique. Furthermore, the
effect of radiation shielding of the thyroid was estimated.

Materials and Methods

Test Object
Received December 19, 1983; accepted after
revision July 9, 1984.
A phantom head type AT-Humanoid (Humanoid Systems) was used to determine the
1 Department of Oral Radiology. Karolinska In-
absorbed doses. This phantom was developed from the Rando phantom [3] and is specified
stitute, Odontologiska Klinikema, Fack, S-i 41 04
to be water equivalent (within 1%) for uses in radiation therapy. To check the absorption
Huddinge, Sweden. Address reprint requests to P.
Julin. properties at lower energies, depth-dose measurements were made by means of thermolu-
2 Department of Hospital Physics, Huddinge lkii- minescence dosimetry. The resufts were compared with corresponding studies on the Rando
varsity Hospital, S-141 86 Huddinge, Sweden. phantom and water performed by Shrimpton et al. [4]. At 70 kVp the relation between the
AJR 143:1113-1116, November 1984 effective absorption coefficients for the Humanoid material and water was 0.87, and at 90
0361-803X/84/1435-1 113 kVp it was 0.92. An x-ray fluorescence study showed the contents of mercury and barium to
C American Roentgen Ray Society be 0.02% and 0.01 respectively.
 1114 JULIN AND KRAEPELIEN AJR:143, November 1984

TABLE 1 : Details of Projection Factors transfer function [2]. All films were automatically processed (Pakorol
X-1 4 machine; Agfa-Gevaert G 138 and G 334 chemicals; total
Angle of T
Beam to ube
Half-
Entrance
processing time 2 mm; temperature: developer 29#{176}C,fixing and
value Extosure
Prolection Frankfurter voitage
Layer (mAs)
Dose rinsing 27#{176}C).
Contrast, density, and fog were regularly measured by
Plane (kvp) (mGy)
(0) (mm PJ) densitometry. Exposure values used in clinical work on adult patients
were used also with the phantom. Projections for the facial skeleton
Lateral 0 73 3.5 8 0.31
as recommended for the Orbix stand [6] were used (fig. 1). Angulation
PA +20 81 3.9 40 2.9
of the beam, exposure values, tube voltage settings, and half-value
PA, semiaxial . . +40 81 3.9 32 2.0
Axial -100 81 3.9 25 1.6 layer used for the different projections are shown in table 1 . The
lateral projection was exposed from the left side. The field diameter
was 0.1 4-0.1 5 m for examination of the paranasal sinuses and 0.16-
0.17 m for the facial skeleton exposures. The corresponding entrance
fields on the phantom were 0.09-0.1 2 m and 0.1 2-0.1 5 m, respec-
tively. A leaded collar of 0.3-mm lead equivalence was placed over
the thyroid region in the axial projection in one experiment.

Dosimetry

/// _ The dosimetry

and an ionization chamber
was performed
detector. The dosimeters
with thermoluminescent
were lithium-
dosimeters

--
fluoride ribbons (3.2 x 3.2 x 0.9 mm, Harshaw Chemical Co.). The
sensitivity varied from dosimeter to dosimeter, and an individual
relative calibration constant was determined for each [7]. The relative
calibration constants were then used to correct the measured ther-
a b moluminescence-signal values for differences in individual dosimeter
senitivity. In the experimental series, a standardized procedure for
handling the dosimeters was used. Before irradiation, the dosimeters
were thermally treated at 400#{176}Cfor 1 hr. After irradiation, the dosim-
eters were chemically treated with methanol containing 1 2 mol HCI/
m3 and then heated at 80#{176}C for 30 mm. Readout was done in a
Teledyne 7300.
:
To measure the absorbed doses, 21 dosimeters were used. In
each measurement of absorbed dose, the absolute sensitivity of the
dosimetry system was determined with four dosimeters. These were
irradiated to a known absorbed dose, which was given in a radiation
field of a reference x-ray machine (Siemens Stabilipan) with the same
C d radiation quality as in the experiments. At this calibration, an ionization
chamber calibrated at the Swedish National Primary Calibration
Fig. 1 -Projections for radiography of paranasal sinuses according to Centre of the National Institute of Radiation Protection was used.
R#{226}dberg
and Welander 161.a, Lateral. b, Posteroanterior (PA). c, PA, semiaxial. The choice of the four calibration dosimeters and background dosim-
d, Axial (submentovertical).
eters differed from one experiment to another.
The organ doses were measured with contralateral pairs of dosim-
eters in place during a full series of sinus exposures. For each
projection, a film was exposed to check the positioning of the head,
Radiography
the alignment and collimation of the beam, and the film density. Those
films were also used for estimation of image quality by two observers
An Orbix stand was equipped with an extra near-focus primary
simultaneously. Thereafter, the exposure was replicated twice. The
collimator, which reduced the maximum diameter of the radiation field effect on the thyroid dose of a leaded apron combined with soft
(at a magnification of 1 .25) to 0.27 m. The extra near-focus collimator leaded collar was studied in the same way in a separate series of
was 7 mm in diameter at 3 mm from the tube window and 1 9 mm at axial (submentovertical) exposures. The entrance doses were meas-
the other end (23 mm from the tube window). A tube with rotating ured with two dosimeters placed on the skin of the phantom head in
anode and 0.6-mm focal spot (Siemens Bi 125/30/50 R) was used the center of the beam. These dosimeters were replaced after each
with a six-pulse generator(Siemens Tridoros Optimatic). The inherent projection. All measurements were independently repeated two to
filtration was stated by the manufacturer to be at least 1 mm Al. The five times. The mean energy imparted was calculated from the
original near-focus filtration was 1 .5 mm Al. To this was added an entrance doses [8].
extra 1-mm aluminum disk.
The tube voltage (kVp) was measured with an electronic penetram-
eter (Digi-X), and the half-value layer (HVL) was determined with pure Results
aluminum sheets and ionization chamber detector. The HVL values
The accuracy of the tube voltage was found to lie within ±
are given in table 1 . Within a voltage variation of ±5%, the total
filtration tabulated from the Medical Physics Data Book [5) was 5.0 5% below 100 kVp. The density, contrast, and spatial reso-
± 0.4 mm. lution of the test images were estimated to be equivalent to
Lanthanide screen-film combinations in speed group 4 (Siemens those of images accepted as satisfactory in clinical routine.
SE 4 with Kodak X-Omat L film) were used. The resolution of the The absorbed doses in regions of the salivary glands, the
screen-film combination was 2.5 line pairs/mm at 30% modulation eye, the maxillary antrum, and the thyroid gland from a full
AJR:143, November 1984 ABSORBED DOSE IN FACIAL RADIOGRAPHY 1115

TABLE 2: Mean Radiation Absorbed Doses from Radiographic The phantom was found not to be fully water equivalent at
Examination of the Paranasal Sinuses
energies used in diagnostic radiology. The effective absorp-
Measure- Dose SDt SD/Mean tion coefficient was about 1 0% lower than that of water. The
L oca t on
menis. (mGy) (mGy) (/o)
mean energy imparted and the thyroid dose found in this
Palatal mucous study were one order of magnitude less than the average
membrane 3 0.33 0.01 3 given by Bengtsson et al. [9, 10].
Parotid gland, medial
The low doses in our study can be explained mainly by the
part . 3 0.11 0.02 18
Submandibular gland . 3 0.15 0.03 19 high-sensitivity screen-film system. The organ that still re-
Eye lens at surface 3 0.61 0.1 5 24 ceives the highest dose is the thyroid gland. This gland is also
Eyeball, 1 0 mm poste- of major interest in risk calculations [1 1 ]. The thyroid dose is
rior to surface:
still high enough to be considered in risk estimations, espe-
Left . 3 0.45 0.02 4
Right 3 0.28 0.02 24 cially since sinus examinations are rather common [1 0, 12].
Maxillary sinus, basal Recent studies have shown considerably reduced thyroid
part . . 3 0.33 0.07 21 doses from other examinations of the face, such as intraoral
Thyroid region: dental radiography [1 0, 1 3], panoramic radiography [1 0, 14-
1 0 mm posterior to
1 6], and cephalography [1 7, 18].
surface, 30 mm be-
low chin 5 0.15 0.03 20 The thyroid dose from a sinus series with high-sensitivity
6 mm posterior to screens would be almost negligible if the submental projection
surface, 50 mm be- were excluded or if a special leaded collar were used. How-
low chin . . 2 0.57 0.01 2
ever, the submental projection is indispensable in many cases.
Skin surface, 30 mm
below chin 5 1 .54 0.1 4 9
The risk of screening off valuable information also renders
. Number of pairs of independenCy repeated measurements.
routine use of a leaded collar for the submental projection
t Standard deviation of a single determination.
questionable. In patients with good flexibility of the neck,
especially children and young people, the collar might be
worth trying, at least in patients needing several repeated
radiographic examinations of the paranasal sinuses.
The following conclusions can be drawn from the study:
The radiation doses to critical organs and the energy imparted
series of four exposures are shown in table 2.
from radiographic examination of the facial skeleton were low.
The minor glands in the palatal mucous membrane received
An Orbix stand tested with speed group 4 screens gave very
a higher dose than the major glands, which were mainly
low doses compared with techniques commonly used in the
outside the primary field. The differences between the left
1 970s. Further reduction of the dose to the thyroid gland
and right side were not statistically significant for the salivary
glands, the maxillary sinus, and the eye lens. The right eyeball from the submental projection would be desirable.
received 38% less than the left.
In the thyroid region, no significant difference between the REFERENCES
right and left side was found. The ratio of the skin doses 15
1 . International Commission on Radiological Protection. Recom-
mm from the midline to the doses 1 0 mm within the surface, mendations of the International Commission on Radiological Pro-
30 mm from the midline, and 30 mm below the chin was 1 :10. tection, publication 26. Oxford: Pergamon, 1977
The absorbed dose 50 mm below the chin, 6 mm within the 2. Swedish Planning and Rationalization Institute of the Health
surface, and 1 2 mm from the midline was 0.57 mGy. The Service. Intensifying screens for x-ray cassettes, project 6030.
mean dose from all sites in the thyroid region was 0.75 mGy. Stockholm: SPRI, 1977
The dominating contribution to the thyroid dose came from 3. Alderson SW, Lanzl L, Rollins M, Spira J. An instrumented
the axial (submental) projection. The leaded collar was found phantom system for analog computation of treatment plans. AJR
1962;87: 185-1 95
to decrease the thyroid dose by a factor of 20. However,
4. Shrimpton PC, Wall BF, Fisher ES. The tissue-equivalence of the
since the phanton head had an inflexible neck, the collar
Alderson Rando anthropomorphic phantom for x-rays of diag-
reduced the posterior part of the image field.
nostic qualities. Phys Med Biol 1981;26: 133-1 39
The mean energy imparted from a full series of sinus 5. National Bureau of Standards. Medicalphysics data book. Wash-
exposures was calculated to be 4.8 mJ. Using the larger field ington, DC: NBS, 1982:138
for examination of the facial skeleton, the energy imparted 6. R#{227}dberg
C, Welander U. Maxillofacial radiography with Orbix.
was 7.9 mJ. Solna, Sweden: Siemens Elema AB, 1977
7. Carlsson CA, M#{225}rtensson
BKA, AIm Carisson G. High precision
dosimetry using thermoluminescent LiF. In: Proceedings of the
Discussion Second International Conference on Luminescent Dosimetry.
Oakridge, TN: U.S. Atomic Energy Commission, 1968:396
The image quality with the high-sensitivity screen-film com-
8. Wachsman F, Drexler C. Graphs and tables in diagnostic radio!-
bination was estimated to be satisfactory in the test films as ogy. Heidelberg: Springer-Verlag, 1976
well as in clinical routine. The modulation transfer function of 9. United Nations Scientific Committee on the Effects of Atomic
the combination used [2] also shows that no significant Radiation. Sources and effects of ionizing radiation. New York:
blurring from the screens is to be expected. UNSCEAR, 1977:307, 31 5, 319
I 1 16 JULIN AND KRAEPELIEN AJR:143, November 1984

1 0. Bengtsson G, Blomgren PG, Bergman K, Aberg L. Patient ex- 14. Wall BF, Fisher ES, Hudson A, Bird PD. Doses to patients from
posures in Swedish diagnostic radiology, report SSI: 1976-013. pantographic and conventional dental radiography. Br J Radio!
Stockholm: Swedish National Institute of Radiology Protection, 1979;52:727-734
1976 15. Forsgren L, Julin P. Radiation dose reduction in panoramic
1 1 . Bengtsson G. Maxiio-facial aspects of radiation protection, fo- radiography. Swed Dent J 1982;6:225-231
cused on recent research regarding critical organs, report SSI: 16. Stenstrom B, Julin P, Richter S. Comparison between panoramic
1977-023. Stockholm: Swedish National Institute of Radiation radiographic techniques. Dentomaxiiofac Radio! 1982;1 1:107-
Protection, 1977 116
12. Berigtsson G, Blomgren PG, Bergman K, Aberg L. Patient ex- 17. Block AJ, Goepp RA, Mason EW. Thyroid radiation dose during
posures and radiation risks in Swedish diagnostic radiology. Acta panoramic and cephalOmetriC dental x-ray examinations. Angle
Radiol Onco! 1978;17:81-105 Orthod 1977;47: 17-23
13. Rehnmark-Larsson 5, Stenstrom B, Julin P, Richter S. Radiation 18. Eliasson 5, Julin P, Richter 5, Stenstrom B. Radiation absorbed
absorbed doses at radiographic examination of third molars. doses in cephalography. Swed Dent J 1984;8:21-27
Swed Dent J 1982;6:189-201