Acta Radiologica: Oncology

ISSN: 0349-652X (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ionc19

Relation Between Kerma and Absorbed Dose in

Photon Beams

B. Nilsson & A. Brahme

To cite this article: B. Nilsson & A. Brahme (1983) Relation Between Kerma and Absorbed Dose
in Photon Beams, Acta Radiologica: Oncology, 22:1, 77-85, DOI: 10.3109/02841868309134343

To link to this article: https://doi.org/10.3109/02841868309134343

Published online: 08 Jul 2009.

Submit your article to this journal

Article views: 10362

View related articles

Citing articles: 18 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ionc20

FROM T H E DEPARTMENT OF RADIATION PHYSICS, KAROLINSKA INSTITUTET, S-10401 STOCKHOLM, SWEDEN.
RELATION BETWEEN KERMA AND ABSORBED DOSE IN PHOTON BEAMS
B. NILSSONand A. BRAHME
The determination of the relation between kerma
and absorbed dose is one of the basic problems of
dosimetry and it is treated in most textbooks on
radiation physics and dosimetry (cf. WHYTE1959,
JOHNS& CUNNINGHAM 1969, GREENING 1981). De-
spite the importance of these concepts, there is no
consistency in the treatments as pointed out already
by ATTIX (1968, 1979). An error is also introduced
when the depth integral of the kerma is equated with
that of the absorbed dose (JOHNS& CUNNINGHAM,
ANDERSON 1976) at least when it is assumed that
the absorbed dose can be taken from experimental
depth dose curves (CORMACK & JOHNS1954, AN-
DERSSON, ALMONDet COll. 1978).
There are at least three reasons for having a good
knowledge of the relation between different kerma
concepts and absorbed dose. First, it is of impor-
tance for the fundamental understanding and the
conceptual clarity of the different quantities that
have the same unit: namely absorbed energy per
unit mass. Secondly, it is of interest to know how
these relations vary in different materials depending
on the basic physical constants governing the elec-
tron transport in the materials, principally: the mass
scattering power and the radiation and collision
mass stopping powers. Thirdly, the ratio of kerma
and absorbed dose is of practical importance in
dosimetry with air kerma or exposure calibrated
ionization chambers.
In the present paper, the relation between kerma
( K ) and absorbed dose (D) is determined on the
basis of the method of calculating the electron trans-
port in a photon beam developed by NILSSON&
BRAHME(1979, 1981) and on the partition of the
Actn Radiologica Oncology 22 (1983) Fuse. I
kerma in a collision part (Kcol)and a radiative part
(Krad)as introduced by ATTIX(1979).
Fundamentals
In a photon beam the kerma is defined as the
initial kinetic energy of all charged particles, mainly
electrons and positrons, liberated by photon interac-
tions per unit mass in a medium (ICRU 1980). Since
part of this kinetic energy may be converted back to
energetic photons mainly through bremsstrahlung
and annihilation in flight processes it is useful to
analyse that part of the kerma which remains as
kinetic energy of charged particles, namely the colli-
sion kerma, Kcol (ATTIX 1979).
Therefore, if a photon beam is impinging on a
semiinfinite half space of matter, the volume inte-
gral of the collison kerma over the half space must
by definition be equal to the absorbed dose integral
or the mean energy imparted, C (ICRU 1980), when
that small part of the collision kerma that may es-
cape through the entrance surface of the half space
can be disregarded:
E, = ///, fL1(r)e(r) d v = // J,D(r) e(r) d v
(1)
where e(r) is the density of the medium.
A more relevant geometry for the use of uniform
collimated photon beams in homogeneous media is
to look at a cylindric volume aligned with the beam
axis and located no less than one maximum electron
Accepted for publication 14 October 1982.
77
78 9. NILSSON AND A . BRAHME

range from the border of the photon beam. As a Interaction ymbol Name
lateral charged particle equilibrium exists in this
geometry, eq. (1) is still valid and may even be Primary and

simplified one step as the photon fluence was as- coherent scatter

sumed to be uniform and thus both the collision

kerma and the absorbed dose will only depend on
the depth along the beam: Incoherent
scatter

Bremsstrahlung

where A is the cross sectional area of the cylindric Annihilation

in flight
volume.
Thus, in this geometry the depth integral of the Annihilal ion
collision kerma is equal to that of the absorbed dose 'at r e s t '

-
assuming again that the charged particle energy Flourascent

backscattered through the front surface can be dis- I

-photon -electroi
Total fluence
regarded. -
c ----posit roi

It is interesting that the relative importance of the

backscattered electrons can be determined by meas-
uring the relative build down at the back surface of a
phantom. If spectral changes are disregarded the
relative dose decrease at the back surface of a phan-
tom, when a saturating backscattering block behind
the phantom is removed, should be exactly equal to
the extrapolated relative surface dose at the front
surface of the phantom.
In order to evaluate the integrals in eqs (1) and (2)
it is important to be able to describe the depth
dependence of energy fluence, kerma and absorbed
dose in a correct and consistent way. By definition
the energy fluence of primary photons, Yp, de-
creases at a rate given by the total linear attenuation
coefficient, p:
Yp(z)= Y&O)e-pz (3 a)
This relation holds exactly for a monoenergetic
photon beam. If the coherently scattered photons
are included in the fluence of primaries a corre-
sponding reduction of p has to be introduced. For
beams with a finite width of the photon spectrum, p
will necessarily vary with depth. This is so as the
absorption of the different photon energies may be
evaluated in terms of a depth dependent linear at-
tenuation coefficient, p(z) defined by
Yp(z)=
1 Yp(O,E ) e-p(E)zdE= Y P(0) e-p(z)'z
Fig. 1. Illustration of the photon interactions contributing to the
total fluence of photons: the primaries Yp, the most important
build-up part (Ys+Yb+YJand the essentially isotropic part Y.
regard to scattered photons and charged particles,
all generations of photons are included, mainly scat-
tered photons, bremsstrahlung and annihilation
quanta, but also fluorescent photons (Fig. 1). This
implies that the depth dependence will be given by
the mean linear energy absorption coefficient of the
total energy fluence:
YJZ) = ~ ~ (e-'e0nZ) (4)
For finite beam cross sections, scattered photons
will necessarily escape from the beam and the true
mean attenuation coefficient, p , will be somewhere
between penand ,u (Fig. 2). The shape of the beam
cross section will therefore determine the exact val-
ue of p and the depth variation of the total energy
fluence:
Y,(z) = w,(O) e-pz (5)
Depending on the shape of the primary photon
spectrum the total fluence spectrum may change
slowly with depth. This implies that p and pen may
have a depth dependence in the above relations (cf.
eq. 3 b). This dependence is generally quite weak,

(3 b)
In the total photon fluence of an infinitely broad
beam, where there is a lateral equilibrium both with
especially for bremsstrahlung spectra and at large
depths where an equilibrium is reached in the ab-
sorption and production of scattered photons. How-
ever, for 6oCo it was found necessary to consider
the slow change in p with depth.
 KERMA A N D ABSORBED DOSE IN PHOTON BEAMS
t 1
I I I
Depth Att. Beam
dependence coeff. Fluence width
Pen Yt
- 1
P Yt
0.5
P Yp
I ,
0 1 2 3
z /cm
Fig. 2. Schematic illustration of the interrelation between differ-
ent attenuation coefficients.
Based on eqs (3) to (5) the depth dependence of
kerma, collision kerma and absorbed dose may now
be written:
Here ptr, the linear energy transfer coefficient,
expresses the fraction of the photon energy trans-
ferred to kinetic energy of charged particles. The pen
value is a reduced p,, value taking into account the
loss of kinetic energy of generated charged particles
due to forward directed bremsstrahlung, annihila-
tion in flight photons and to a less extent to essen-
tially isotropic fluorescent photons. It is important
that these photons are included in the total energy
fluence of photons (Fig. l), as they are dosimetrical-
ly difficult to distinguish from the primary photons.
Furthermore, if the depth dependence of the total
energy fluence is determined by dosimetric meas-
urements according to eqs (5) and (8) the mean
attenuation coefficient determined will automatical-
ly include them.
The correction factor p(z) (cf. LOEVINGER 1981)
which takes the motion of the charged particles into
account will be treated in considerable detail in the
following, based on calculations of the charged par-
ticle transport using the diffusion equation and the
basic equality given by eq. (2).
79
Method
In order to calculate the integral of the collision
kerma the effective attenuation coefficient for the
Infinite photons has to be determined. This may be done by
using build-up factors (ATTIX1979). However, the
build-up factors normally published do not include
Broad
the contribution from bremsstrahlung and annihila-
1
Narrow
tion in flight photons and must consequently be
corrected. An estimation of this correction for 6
Point
mono- MeV photons in aluminium gives a change in the
direc- effective attenuation coefficient of 2 per cent. For
tional
lower energies and lower atomic number materials
the difference is smaller.
An alternative way to determine the effective at-
tenuation coefficient is by using experimental depth
dose curves for broad beams corrected to infinite
SSD. A detailed analysis of the experimental data
shows that for 6oCo the change of the p value with
depth must be considered. This was done by fitting
the TAR values in water (Brit. Inst. Radiol. 1972)
with three-exponential functions according to
TAR(z) = k[e- mze-Pi + ( I -e-mz) e-'2'] (9)
where k, m, p l and p2 are field size dependent
(7) factors determined to minimize the mean square
deviations from experimental data. With this func-
tion the agreement with tabulated data was better
than 0.5 per cent and most of the deviations were
due to random errors in the tabulated TAR values.
For a 4 cm x 4 cm field the factors obtained were
1.02305, 0.14918 cm-', 0.02668 cm-' and 0.05875
cm-', respectively. This will give a mean extrapo-
lated effective attenuation coefficient of 0.029 cm- '
in the build-up region in agreement with experimen-
tally obtained attenuation coefficients for the wall in
ionization chambers (HOLTet coll. 1979, CUNNING-
HAM & SONTAG 1980).
For 21 MV roentgen radiation only a single p
value of 0.0209 cm-' for the whole depth range was
sufficient. Data for air and aluminium have been
obtained by scaling data for water by the mass
energy absorption coefficient.
The effective attenuation coefficient can be used
to calculate the depth integral of the absorbed dose
for depths beyond the depth of transient equilibri-
um. The main problem is thus to determine p(z) or
the shape of absorbed dose curve in the build-up
region.
In the present work the absorbed dose in the
build-up region was obtained by using the method of
80 B. NILSSON AND A. BRAHME

9.0

8.0

7.0 -
6.0 -

5.0 -

4.0 -

3.0 -

1.0

, o
0 5 10 15 20

0/degree
Fig. 3. Approximation of the incoherent electron angle distribu- gies of the groups 1, 2 and 3 are 4.51 MeV, 2.33 MeV and 2.33
tions according to the Klein-Nishina equation with three Gaus- MeV, respectively.
sian distributions. Photon energy: 5 MeV. Mean electron ener-

NILSSON& BRAHME(1979, 1981) to calculate the = mean mass energy transfer coefficient in
contribution to the dose distribution for each photon water,
energy and group of secondary electrons according = mean attenuation coefficient of the beam,
to: = mean extrapolated range of the secondary
electrons,
= coefficients describing the transmission of
a parallel beam of electrons as taken from
TABATA & IT0 (1974),
= initial mean square scattering angle of the
D(z) = absorbed dose due to secondary electrons electrons emitted in a photon interaction,
at depth z , the factor (1+B$/2) modifies the trans-
K(0) = kerma at phantom surface, mission functions to take the angular dis-
(7 =the cross section for production of sec- tribution of emitted electrons into ac-
ondary electrons belonging to the group count,
under consideration (cf. Fig. 3), = the ratio of fluence to planar fluence for
N = the density of atoms in the medium, the secondary electrons.
SC& =mean collision mass stopping power for
the secondary electrons, fr is calculated from the equation
 KERMA AND ABSORBED DOSE tN PHOTON BEAMS
0.2 - eq. (10) should ideally give the absolute value of the
where g ( 8 ) = l/)cos81 for l 8 - x / 2 1 ~8, and g(8) =
Usin8, elsewhere (BRAHME 1975) and $(z)
= t h e mean square scattering angle at depth z.
The cut off angle for cos 8 was taken from 8; = Td,
where To = the linear scattering power, A = mean
free path of the electrons.
In order to take the energy and angular distribu-
tions of the electrons into account the following
assumptions were made. For incoherent scattering
the differential cross section for the recoil of a free
electron at angle 8 with respect to the direction of
the photon (dddQ) was approximated by a sum of
three Gaussian distributions (Fig. 3). Thus, it was
possible to divide the recoil electrons in one high
energy small scattering angle group and a low ener-
gy large angle group (cf. BRAHME1982).
For the pair production interactions the electron
energy varies only slightly with the emission angle
as seen from the angular dependence of the mean
positron energy (Fig. 4). For low photon energies
the mean energy was calculated from the differential
6-838101
81
for 21 MV roentgen radiation the spectrum was
approximated by three energy groups.
Beside the absence of a true d-particle equilibrium
absorbed dose. However, by using very few energy
groups the calculations include some uncertainty,
especially through the energy dependence of the S,,,
also without normalization.
After normalization for each energy the contribu-
tion from the different electron energies were
weighted and added together and the absorbed dose
in the build-up region was obtained. The absorbed
dose distribution for larger depths was then added
and the total absorbed dose distribution could be
obtained.
An alternative method to using a fluence build-up
factorf, and the planar fluence transmission eq. (10)
is to use the energy deposition algorithm suggested
by TABATA& ITO, eq. (10) then becomes
(12)
where I(z)dz is taken from the energy deposition at
depth z , according to TABATA& ITO, but corrected
for the mean initial angular spread of the emitted
electrons. This method was used for comparison in
a few cases.
Results and Discussion
In Fig. 5 the absorbed dose distribution in the
build-up region for 6oCo y-rays is plotted for three
different materials. A noticeable difference was
82 B. NILSSON AND A. BRAHME

1.0

-
D (11
K (01

0.5

z/gcm2
Fig. 5 . Absorbed dose distribution in water (-), air (---) and in water are also included. They cannot, however, be distin-
aluminium (---) in the build-up region for 6oCo y-rays calcualted guished in this scale. Data are normalized to kerma at surface.
according to eq. (10). The kerma andcollision kerma distributions

1.0

D (21
-
K 101

0.5

0
0 0.1 0.2 0.3 0.4 0.5 0.6

Fig. 6. Absorbed dose distribution in water in the build-up region function model eq. (13) (GREENING 1981, . .) and the Monte
for
oC '@ y-rays calculated by different methods: the present Carlo method using the electron energy deposition distributions
electron diffusion model eq. (10) (-), the energy deposition algo- according to Berger (cf. ALLISY1967) (J).
rithm model eq. (12) (TABATA & ITO 1974, ---), the exponential
 KERMA AND ABSORBED DOSE IN PHOTON BEAMS 83

1.0 found between air, water and aluminium, principally

-
D IZl
depending on the differences in scattering and stop-
K 101
ping powers of these materials. Aluminium has high-
er mass scattering power than water and air which
implies that the dose rises faster. The slightly deeper
penetration in air compared with water is due to the
smaller mass stopping power of air which results in
a somewhat larger range of the electrons as the mass
0.5 scattering power of these materials is almost identi-
cal.
In Fig. 6, eq. (10) is compared with data obtained
by using the energy deposition algorithm given by
eq. (12). This algorithm seems to give a faster rise in
absorbed dose. In the figure is also plotted data
obtained by integrating the energy deposition distri-
bution for a plane parallel 6oCo compton electron
0
source in air calculated by Berger (cf. ALLISY1967).
0 0.1 0.2 0.3 0.4 The later data give a lower dose in the first two mm.
z /gem= Another way to express the depth absorbed dose
Fig. 7. Comparison between calculated absorbed dose in the distribution is by two-exponential functions (JOHNS
build-up region, eq. (10). with experimental data for mCo y-rays. et coll. 1949, BRAHME& SVENSSON 1979, GREEN-
The experimental data are measured for small fields and include
collimator scattered electrons which are not included in the cal- ING). The relation between absorbed dose and
culated curve. Compton electrons from the air are, however, kerma then becomes:
included. RICHARDSON et coll. (I954 A).SMITH & SUTHERLAND
(1976 0 ) . LEUNG et coll. (1976 w). Present investigation (4.
WZ) - Pe (e-pz-e-~e~
)
N O ) P-Pe (13)
where pe is an effective linear attenuation coeffi-
1.0
cient for the electrons. pe was determined by linear
D fzl
- exponential regression to 8.55 cm-'. This gives a
K (0)
quite good agreement (cf. Fig. 6) with the calculated
depth dose distribution.
For comparison some experimental data are in-
cluded in Fig. 7 (RICHARDSON et COIL 1954, SMITH
& SUTHERLAND 1976, LEUNGet coll. 1976) where
contamination of air electrons (NILSSON & BRAHME
1979) and backscattered electrons and photons have
0.5 been added to the results obtained by eq. (10). A
linear absorption of the air electrons in the phantom
has been assumed. The contribution from backscat-
tered electrons was taken into account by using
moments method calculations of SPENCER (1959)
giving the absorbed dose at the surface. The calcu-
lated curve seems to rise too fast during the first
mm. The differences are, however, small and the
agreement is good considering the approximations
0 made.
0 2 .o 4 .O 6.0 8.0
In Fig. 8 the kerma ( K ) and collision kerma (I&,)
z/gcm2
for water and 21 MV roentgen radiation are plotted
Fig. 8. Absorbed dose distribution in water and air in the build-up together with the absorbed dose (D)normalized to
region for 21 MV roentgen radiation calculated according to eq. 1.00 for the kerma at the surface. The ratios of
(10). The kerma and collision kerma distributions in water are also
included. Data are normalized to kerma at surface. absorbed dose to collision kerma and kerma are
84 B. NILSSON AND A. BRAHME

Table
Absorbed dose and kerma ratios

Photon Material Field size DIK,,, DIK km

energy (m2)

1.25 MeV*
(60co) Water l o x 10 1.005 1 1.0011 0.997
4x4 1.0059 1.0019 0.992
Air l o x 10 1.005 1 1.0011 0.997
4x4 1.0059 1.0019 0.991
(1.007)
Aluminium l o x 10 1.0043 0.998 0.997
4x4 I ,0049 0.997 0.991
(1.006)
6 MeV** Aluminium Broad 1.017 0.966 0.948
beam (1.018) (0.967)
21 MV Water 20x20 1.023 1.001 0.910
Air 20x20 1.022 1.000 0.908

* Data in parentheses from LOEVINGER

(1981).
** Data in parentheses from ATTIX(1979).

tabulated in the Table. The DIK,,, values for the 4
cm x 4 cm field agree well with those calculated by
LOEVINGER using the mean distance, X, that the
secondary electrons travel before they deposit their
energy as suggested by ROESCH(1958, 1968). LOE-
VINGER used a narrow beam attenuation coefficient
in his calculations, which holds for the primary pho-
tons and is rather close to the effective attenuation
for a small beam. For the 10 cm x 10 cm field the
values are somewhat lower depending on the larger
effective attenuation coefficient.
Using the present model, R can easily be calculat-
ed from the relationIn[P(z)l,/,l(z). It is interesting to
note that because /3(z)stays constant independent of
depth beyond the depth of transient equilibrium, X
will decrease slowly with depth due to the slow
increase with depth of p(z).
The absorbed dose and the kerma practically co-
incide in the transient equilibrium region because
for 6oCo(1-g) is 0.996 and /3 is 1.005 (REICH& TRIER
198 1 , GREENING).
When calibrating an air equivalent ionization
chamber free in air, a correction for the photon
attenuation in the wall has to be made. An interest-
ing quantity for dosimetry is therefore the ratio be-
tween the absorbed dose at maximum build-up and
the collision kerma at the surface as the latter quan-
tity can easily be obtained from an exposure or air
kerma calibration and the former quantity is equal to
the absorbed dose to the air in the cavity. The ratio
between the absorbed dose at maximum build-up
and the collision kerma at the surface is given in the
Table under the heading k,,,. These values can be
used to correct for wall attenuation when using
plane parallel chambers in a phantom.
When using a cylindric ionization chamber free in
air this ratio is not strictly applicable because the
geometry is somewhat different. However, the at-
tenuation factor for the 4 cm x 4 cm field is 0.991 for
air close to what is obtained experimentally for cy-
lindric chambers (JOHANSSON et coll. 1978, HOLTet
coll.) and by Monte Carlo methods (BONDet coll.
1978). This indicates that the effect of the cylindric
shape of the thimble chamber is of small importance
for the value of the attenuation factors.
The data can also be used to investigate the mate-
rial dependence of the attenuation factor. According
to the present results the material dependence is
very weak (Table) with a difference of less than 0.1
per cent, as also found experimentally by JOHANS-
SON et COll.
For 21 MV roentgen radiation very similar results
are obtained for water and air. Surprisingly enough,
also at this high energy the kerma still practically
coincides with the absorbed dose for depths beyond
transient equilibrium as (1-g)/3 is still very close to
unity.
In order to compare with the results obtained by
 KERMA AND ABSORBED DOSE IN PHOTON BEAMS
ATTIX (1979) calculations were made for 6 MeV
photons in aluminium. Practically the same results
were obtained with both methods (Table).
SUMMARY
The relation between kerma and absorbed dose has been
calculated using a diffusion equation approximation for
the electron transport in photon beams and by equating
the depth integral of the collision kerma with that of the
absorbed dose. The results show that the absorbed dose in
the transient equilibrium region practically coincides with
the total kerma in water both for *Co and 2 1 MV roentgen
rays. The ratio between the absorbed dose at maximum
build-up and the collision kerma at the surface (katt)was
calculated. The results show that the dependence of k,,
on the atomic number is very small. For @'Co kartwas
determined for water and air to 0.992 and 0.991, respec-
tively.
REFERENCES
ALLISYA.: Contribution a la mesure de I'exposition pro-
duite par les photons emis par le @'Co. Metrologia 3
(1967), 41.
ALMOND P. R., MENDEZA. and BEHMARD M.: Ionization-
chamber-dependent factors for calibration of
megavoltage X-ray and electron beam therapy ma-
chines. In: International Symposium on National and
International Standardization of Radiation Dosimetry.
Vol. 11. IAEA-SM-222/29. Vienna 1978.
ANDERSOND. N.: Conversion from exposure to dose for
megavoltage beams under 3 MeV. Phys. Med. Biol. 21
(1976), 524.
A ~ I F.
X H.: Basic y-ray dosimetry. Hlth Phys. 15 (1968),
49.
- The partition of kerma to account for bremsstrahlung.
Hlth Phys. 36 (1979), 347.
BONDJ. E., NATHR. and SCHULZR. J.: Monte Carlo
calculation of the wall correction factors for ionization
chambers and A,, for *Co y-rays. Med. Phys. 5
(1978), 422.
BRAHME A.: Investigations on the applications of a micro-
tron accelerator for radiation therapy. Thesis, Stock-
holm 1975.
- Correction for the angular dependence of a detector in
electron and photon beams. To be published in Acta
radiol. Oncology 22 ( 1983).
- and SVENSSON H.: Radiation beam characteristics of a
22 MeV microtron. Acta radiol. Oncology 18 (19791,
244.
BRITISHINSTITUTEOF RADIOLOGY.Central axis depth
dose data for use in radiotherapy. Brit. J. Radiol.
(1972) Suppl. No. 1 1 .
CORMACK D. V. and JOHNSH. E.: The measurement of
high energy radiation intensity. Radiat. Res. 1 (19541,
133.
CUNNINGHAM J. R. & SONTAG M. R.: Displacement cor-
rections used in absorbed dose determination. Med.
Phys. 7 (1980), 672.
GLUCKSTERN R. L. and HULL JR M. H.: Polarization
85
dependence of the integrated bremsstrahlung cross
section. Phys. Rev. 90 (1953), 1030.
GREENING J . R.: Fundamentals of radiation dosimetry, p.
66. Adam Hilger Ltd, Bristol 1981.
HOLTJ. G., FLEISCHMAN R. C., PERRYD. J. and BUFFA
A.: Examination of the factors A, and AE, for cylindri-
cal ion chambers used in cobalt-60 beams. Med. Phys.
6 (1979), 280.
ICRU: Radiation quantities and units. Report No. 33,
1980.
JOHANSSON K.-A., MATTSSONL. O., LINDBORG L. and
SVENSSON H.: Absorbed dose determination with ion-
ization chambers in electron and photon beams with
energies between 1 and 50 MeV. In: International
Symposium on National and International Standard-
ization of Radiation Dosimetry, Atlanta 1977. IAEA-
SM-222135. Vienna 1978.
JOHNSH. E. and CUNNINGHAM J. R.: The physics of
radiology. Third edition, p. 277. Charles C. Thomas,
Springfield 1969.
- DARBYE. K., HASLAMR. N. H. KATZL. and HAR-
RINGTON E. L.: Depth dose data and isodose distribu-
tions for radiation from a 22 MeV betatron. Amer. J.
Roentgenol. 62 (1949), 257.
LEUNGP.M., SONTAG M. R., MAHARAJ H. and CHENERY
S.: Dose measurements in the build-up region for Co-
balt-60 therapy units. Med. Phys. 3 (1976), 169.
LOEVINGER R.: A formalism for calculation of absorbed
dose to a medium from photon and electron beams.
Med. Phys. 8 (1981), 1 .
NILSSONB. and BRAHME A.: Absorbed dose from second-
ary electrons in high energy photon beams. Phys.
Med. Biol. 24 (1979), 901.
- - Contamination of high-energy photon beams by
scattered photons. Strahlentherapie 157 (1981), 181.
REICHH. and TRIERJ. 0 . : Which is the quantity best
suited to replace exposure when switching form old
units to SI units? In: Biomedical dosimetry. Physi-
cal aspects, instrumentation, calibration. IAEA-SM-
249158. Vienna 1981.
RICHARDSON J. E., KERMAN H. D. and BRUCERM.: Skin
dose from a Cobalt 60 teletherapy unit. Radiology 63
(1954), 25.
ROESCHW. M. C.: Dose for nonelectronic equilibrium
conditions. Radiat. Res. 9 (1958), 399.
- Mathematical theory of radiation fields. In: Radiation
dosimetry, p. 229. Edited by F. H. Attix and W. M . C.
Roesch. Academic Press, New York 1968.
SAUTER F.: Uber die Bremsstrahlung schneller Elek-
tronen. Ann. Physik 20 (1934), 404.
SCHIFFL. I.: Energy angle distribution of thin target
bremsstrahlung. Phys. Rev. 83 (1951), 252.
SMITH C. W. and SUTHERLAND W. H.: Electron contami-
nation of telecobalt beams. Brit. J. Radiol. 49 (1976),
563.
SPENCER L. V.: Energy dissipation by fast electrons. NBS
Monograph 1, Washington 1959.
TABATA T. and IT0 R.: An algorithm for the energy depo-
sition by fast electrons. Nucl. Sci. Eng. 53 (1974), 226.
WHYTEG. N.: Principles of radiation dosimetry, p. 62.
John Wiley & Sons, New York 1959.