Int. J. Radiation Oncology Biol. Phys., Vol. 59, No. 3, pp.

905–910, 2004
Copyright © 2004 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/04/$–see front matter

doi:10.1016/j.ijrobp.2004.02.030

PHYSICS CONTRIBUTION

COMPARISON OF ELECTRON BEAM CHARACTERISTICS FROM

MULTIPLE ACCELERATORS

DAVID S. FOLLOWILL, PH.D., DENISE S. DAVIS, M.S., AND GEOFFREY S. IBBOTT, PH.D.
Department of Radiation Physics, The University of Texas M. D. Anderson Cancer Center, Houston, TX

Purpose: To determine the relationships between electron beam depth dose characteristics, depth of maximum
dose (dmax), depth of 80% dose (d80), and depth of 50% dose (d50), and the nominal energy designation of electron
beams from multiple linear accelerators for the purpose of electron beam treatment planning and quality
assurance.
Methods: The Radiological Physics Center Staff, during its on-site dosimetry review visits to institutions
participating in clinical trials, measured depth dose characteristics for more than 2000 electron beams. Mea-
surements were performed on Varian, Siemens, and Elekta/Philips accelerators generating beams with nominal
energy values ranging from 4 –22 MeV. The depth dose data were determined at the nominal source-to-skin
distance with the reference cone size in accordance with recommendation of the American Association of
Physicists in Medicine Task Group 25 report.
Results: The important depth dose characteristics dmax, d80, and d50 varied in a predictable fashion when plotted
against the true beam quality indicator, R50. However, d80 and d50 values overlapped considerably when plotted
against the manufacturers’ nominal electron energy values. For a specific nominal electron energy value, the
values of dmax, d80, and d50 varied by as little as 3 mm for low energy levels to nearly as much as 20 mm for high
energies.
Conclusions: The manufacturer’s nominal electron energy value does not adequately describe the depth dose
characteristics of an electron beam for treatment planning purposes. Clinicians and physicists should determine
and use only the specific depth dose data for their clinical beams and not the manufacturer’s nominal value.
© 2004 Elsevier Inc.

Electron beam dosimetry, Electron beam planning, Quality assurance.

INTRODUCTION spot-checking the depth dose data at the depth of maximum

In 1980, the Radiological Physics Center (RPC) staff began
routinely reviewing megavoltage electron beams through
on-site dosimetry review visits to institutions participating
in cooperative clinical trials. The RPC’s review included an
independent verification of beam calibration, depth dose
characteristics, and cone ratios both at nominal source-to-
skin distance (SSD) and extended SSD at each electron
energy. Initially, the RPC evaluated the depth dose charac-
teristics by measuring 8 –10 points along the depth dose
curve. However, after collecting data on approximately 100
electron beams, Kirby et al. (1) showed that data from
electron beams of the same mean incident energy and the
same make and model were very consistent. As a result, the
RPC in the mid 1980s modified its technique and began
dose (dmax), depth of 80% dose (d80), and depth of 50%
dose (d50). In 1984, the RPC implemented the American
Association of Physicists in Medicine (AAPM) Task Group
21 (TG-21) calibration protocol for photon and electron
beam calibrations and began applying a replacement cor-
rection factor (Prepl) and a restricted stopping power ratio
(L៮ /␳) corrections to its depth-ionization readings for the
calculation of electron depth dose data (2). Since that time,
the RPC has used consistent and reproducible techniques to
measure depth dose characteristics of more than 2100 elec-
tron beams from nearly 500 linear accelerators (3).
The electron beam energy spectrum emerging from the
accelerator treatment head is complex and may differ ap-
preciably among accelerators from different manufacturers

Reprint requests to: David S. Followill, Ph.D., The University of
Texas M. D. Anderson Cancer Center, Department of Radiation
Physics, Unit 547, 1515 Holcombe Boulevard, Houston, TX
77030.Tel:(713)745-8989;Fax:(713)794-1364;E-mail:dfollowi@
mdanderson.org
Presented in part at the 37th Annual Meeting of the American
Association of Physicists, 1995, in Boston, MA.
This work was supported by Public Health Service Grant no. CA
10953, awarded by the National Cancer Institute, Department of
Health and Human Services.
Acknowledgments—The authors thank the many physicists of the
Radiological Physics Center for assisting in gathering the data as
part of the Center’s quality audit program. The authors further
acknowledge Morris Tatcher, Alexander Nevelsky, and Raquel
Bar-Deroma for useful discussions regarding these data.
Received Aug 22, 2003, and in revised form Feb 13, 2004.
Accepted for publication Feb 16, 2004.

905
906 I. J. Radiation Oncology ● Biology ● Physics Volume 59, Number 3, 2004

Table 1. Average depth dose characteristics for electron beams from Varian accelerators

Nominal energy (MeV) Average Standard Range of

(R50 [cm]) N Parameter depth (cm) deviation (cm) depths (cm)

4 (1.4) 3 dmax 0.7 0.2 0.6–0.9

d80 1.2 0.1 1.1–1.2
d50 1.4 0.1 1.4–1.5
6 (2.3) 244 dmax 1.1 0.2 0.8–1.6
d80 1.9 0.1 1.3–2.2
d50 2.3 0.1 1.8–2.7
9 (3.5) 250 dmax 1.7 0.3 1.2–2.3
d80 2.9 0.2 2.3–3.4
d50 3.5 0.2 2.9–4.1
12 (4.8) 246 dmax 2.4 0.3 1.5–3.1
d80 4.0 0.2 3.5–4.6
d50 4.8 0.2 4.2–5.4
15–16 (7.2) 246 dmax 2.4 0.6 1.0–3.7
d80 5.3 0.3 4.5–5.9
d50 6.4 0.4 5.5–7.0
18–20 (7.9) 232 dmax 1.8 0.5 1.0–2.8
d80 6.4 0.4 5.5–7.1
d50 7.9 0.5 6.9–8.7
22 (8.9) 16 dmax 1.7 1.0 1.0–2.3
d80 7.0 0.3 6.6–7.7
d50 8.8 0.2 8.4–9.2

and among models of accelerator from the same manufac-
turer, depending on the design of the treatment head. The
manufacturers’ displayed nominal electron energy values
are generally defined in terms of the most probable (modal)
energy of the discrete electron energy spectrum before it
leaves the vacuum window of the accelerating structure
within the head of the accelerator (4). Over the years, the
manufacturers of accelerators have used different methods
to describe electron beam energies (5). Because of variabil-
ity in the manufacturers’ nominal electron energy specifi-
cations, the nominal electron energy value is not meaningful
for physics or dosimetry purposes. The AAPM calibration
protocol (6), an AAPM electron dosimetry report (7), and an
independent publication (8) all state that electron depth dose
values best define the actual electron energy of the spectrum
reaching the patient. In particular, the d50 along the central
axis adequately characterizes the electron beam quality for
calibrations and dosimetry calculations (6, 7, 9).
In contrast, clinicians do not have a standard method of
specifying therapeutic electron beam energy. Traditionally,
clinicians are instructed that the prescription depth (in cm)
is equal to either one fourth or one third of the manufactur-
er’s nominal energy (MeV), corresponding to the approxi-
mate depths of the 90% or 80% isodose lines, respectively.
Clinicians are also taught that the approximate range of the
electrons (cm) is equal to one half of the nominal electron
energy value (MeV), and this value is often used as a guide
for protecting sensitive underlying, normal tissues, such as
lung (10). These rules of thumb can be misleading, because
the manufacturer’s nominal electron energy does not fully
describe the dosimetry characteristics of the specific elec-
tron beam. The depths of 80% and 90% dose depend very
strongly on the electron field size, in particular for field sizes
less than 10 cm ⫻ 10 cm. The depth dose curves, however,
do not differ significantly for field sizes 10 cm ⫻ 10 cm or
larger.
We used measured data obtained in more than 20 years of
on-site dosimetry review visits to establish the relationship
between the manufacturer’s nominal electron energy and
actual beam energy and dose distribution. We report mea-
sured depth dose data for electron beams from three man-
ufacturers: Elekta/Philips, Siemens, and Varian. The data
presented in this report provide clinical physicists a useful
quality assurance tool to provide redundant checks of their
electron beam dosimetry parameters. In addition, clinicians
will benefit from the data by better understanding the spe-
cific dosimetry characteristics associated with the manufac-
turers’ nominal electron energy values. Electron beam ther-
apy planning should be based on the actual beam energy and
dose distribution, not on the manufacturer’s nominal value.
METHODS AND MATERIAL
Beam characteristics of accelerators installed at institu-
tions participating in cooperative clinical trials were mea-
sured by the staff of the RPC. The RPC is responsible for
providing an independent quality assurance audit program
to institutions entering patients into clinical trials. The qual-
ity assurance audit program includes on-site dosimetry re-
view visits to many of the participating institutions by RPC
physicists. During these visits, the RPC staff perform a
series of dosimetry measurements to ensure that the dosim-
etry data used to calculate beam-on time for patients are
accurate. As a part of the measurements, each electron beam
output is calibrated, and the depth dose curve is verified at
dmax, d80, and d50.
 Comparison of electron beam characteristics ● D. FOLLOWILL et al. 907

Table 2. Average depth dose characteristics for electron beams from Siemens accelerators

Nominal energy (MeV) Average Standard Range of

R50 [cm] N Parameter depth (cm) deviation (cm) depths (cm)

5 (1.9) 18 dmax 1.0 0.1 0.8–1.1

d80 1.6 0.1 1.3–1.8
d50 1.9 0.1 1.7–2.1
6 (2.3) 72 dmax 1.2 0.1 1.0–1.5
d80 1.9 0.1 1.0–2.2
d50 2.3 0.1 2.0–2.7
7 (2.7) 65 dmax 1.4 0.2 0.9–1.9
d80 2.2 0.2 1.5–2.9
d50 2.7 0.2 2.1–3.5
8 (3.1) 41 dmax 1.7 0.2 1.2–1.9
d80 2.6 0.2 2.2–2.9
d50 3.1 0.2 2.6–3.5
9 (3.5) 50 dmax 1.9 0.2 1.4–2.2
d80 3.0 0.2 2.3–3.3
d50 3.5 0.2 2.7–3.7
10 (4.0) 75 dmax 2.2 0.2 1.8–2.5
d80 3.3 0.2 2.7–3.8
d50 4.0 0.2 3.3–4.5
11 (4.1) 13 dmax 1.9 0.3 1.6–2.5
d80 3.4 0.3 2.9–3.7
d50 4.1 0.3 3.6–4.5
12 (4.7) 106 dmax 2.5 0.2 1.8–3.0
d80 4.0 0.2 3.5–4.4
d50 4.7 0.2 4.2–5.2
14 (5.3) 25 dmax 2.7 0.2 2.3–3.0
d80 4.5 0.1 4.3–4.7
d50 5.3 0.1 5.0–5.5
15 (6.1) 54 dmax 2.5 0.4 1.7–3.2
d80 5.1 0.2 4.8–5.5
d50 6.0 0.2 5.7–6.8
18 (7.4) 49 dmax 2.3 0.6 1.4–3.5
d80 6.1 0.2 5.4–6.5
d50 7.3 0.2 6.6–7.8
20 (8.2) 6 dmax 2.4 0.5 1.5–3.0
d80 6.7 0.1 6.5–6.9
d50 8.2 0.1 8.0–8.4
21 (8.3) 19 dmax 2.0 0.6 1.2–3.0
d80 6.8 0.2 6.5–7.3
d50 8.3 0.2 8.0–8.6

Measurements were made using electron beams from
Varian accelerators (Clinac models 12, 18, 20, 1800, 2100
series, 2300 series, and 2500), Siemens accelerators (mod-
els 12, 20, 67, 74, 77, MD, MD2, KD, MXE, and Primus),
and Elekta/Philips accelerators (models SL75, SL18, SL20,
SL25, and Precise). The data consist of 2127 electron beams
ranging in energy from 4 –22 MeV.
The RPC measurements were made in a water phantom
that provided adequate lateral scatter and backscatter
using a Farmer-type 0.6 cm3 ion chamber- either NE
model 2571 (NE Technology, Ltd., Berkshire, England)
or PTW model N23333 (PTW, Freiburg, Germany), and
read with a Keithley model 602 electrometer (Keithley
Instruments, Inc., Cleveland, OH). All of the depth dose
values were measured at nominal SSD and used a refer-
ence cone size in accordance with the AAPM TG-25
report (7). The reference cone size was 10 cm ⫻ 10 cm
for 93% of the measurements while the remaining data
were taken with a cone size of 15 cm ⫻ 15 cm. The SSD
was 100 cm for nearly all (99%) of the measurements. An
SSD of 95 cm was used for a few of the Elekta/Philips
machines. All electron beam measurements incorporated
a shift of 0.5rcav to the effective point of measurement.
Values of dmax were measured in increments of 1 mm or
2 mm for electron beam energies less than 10 MeV. For
electron energies greater than or equal to 10 MeV, the
institution’s value of dmax was used because a flat plateau
region around dmax makes the precise location less crit-
ical for these higher energies. Exact values of d80 and d50
were interpolated from shallower (⬎80%) and deeper
(⬍50%) measurements.
The average d80 values should aid clinicians in determin-
ing the effective therapeutic depth for different electron
energies while the average d50 values are important to
physicists in determining the beam quality of each electron
beam for calibration purposes.
908 I. J. Radiation Oncology ● Biology ● Physics Volume 59, Number 3, 2004

Table 3. Average depth dose characteristics for electron beams from Elekta/Philips accelerators

Nominal energy (MeV) Average Standard Range of

(R50 [cm]) N Parameter depth (cm) deviation (cm) depths (cm)

6 (2.4) 28 dmax 1.2 0.1 1.0–1.5

d80 1.9 0.2 1.5–2.2
d50 2.4 0.2 1.8–2.8
8 (3.1) 32 dmax 1.6 0.2 1.2–1.9
d80 2.6 0.2 2.2–3.0
d50 3.1 0.3 2.7–3.6
10 (3.9) 31 dmax 2.2 0.3 1.6–2.5
d80 3.3 0.2 2.9–3.8
d50 3.9 0.3 3.5–4.6
12 (4.8) 29 dmax 2.3 0.3 1.6–2.8
d80 4.0 0.2 3.7–4.4
d50 4.8 0.3 4.4–5.4
14 (5.6) 19 dmax 2.1 0.5 1.4–3.0
d80 4.6 0.3 4.2–5.1
d50 5.5 0.3 5.1–6.0
15 (6.0) 12 dmax 2.9 0.4 2.3–3.4
d80 5.1 0.3 4.7–5.5
d50 6.0 0.4 5.6–6.8
17 (7.0) 13 dmax 2.3 0.5 1.4–3.3
d80 5.7 0.4 5.0–6.3
d50 7.0 0.4 6.3–7.8
18 (7.1) 6 dmax 2.3 0.4 1.8–3.0
d80 5.9 0.2 5.8–6.2
d50 7.1 0.1 6.9–7.3
20 (7.9) 12 dmax 2.1 0.7 1.4–3.5
d80 6.5 0.3 5.9–6.5
d50 7.9 0.4 7.4–8.7

RESULTS ered, R50 exhibited a continuous spectrum of values, with no

apparent grouping around the nominal energy values. This
Tables 1–3 list the average dmax, d80, and d50 values for
is in distinct contrast to photon beams, for which the beam
Varian, Siemens, and Elekta/Philips accelerators, respec-
qualities appear to correspond well with nominal energies
tively. Tables 1–3 also show the standard deviations and
across different accelerator makes and models (11). If,
ranges for each dmax, d80, and d50 value. As one would
instead, the depth data are compared with the manufactur-
expect, the average d80 and d50 values increased with en-
er’s nominal electron energy, as shown in Fig. 2, discrete
ergy, and the dmax values for energies greater than 12 MeV
clumping of the data are seen but the ranges of d80 values in
exhibited a large standard deviation. The increased standard
most cases overlapped the ranges of d50 values. Beams with
deviation for the dmax values can be explained by the broad
equal nominal energies showed a range of dmax, d80, and d50
peak of the depth dose curve for these higher energies. The
values that varied by as little as a few millimeters at the low
range of dmax, d80, and d50 values overlapped between
energies to nearly 20 mm for the higher energies.
nominal energies, however, the average values plus or mi-
The values of d80 and d50 varied almost linearly with R50
nus 1 standard deviation overlapped with adjacent nominal
for the three makes of accelerators shown in Fig. 1. No
energies very little.
distinction among accelerator manufacturers was observed
Figure 1 shows the dmax, d80, and d50 values as a function
for the values of d80 and d50 as shown in Fig. 1, and the
of R50, the depth of 50% dose, which is defined by the
beams showed comparable dosimetry characteristics when
AAPM TG-51 report as the electron beam quality specifi-
energy was expressed in terms of R50.
cation (6). R50 is also recommended by the International
Atomic Energy Agency (IAEA) (9) to characterize electron
beam quality and is preferred over the outdated ICRU 35 DISCUSSION
nomenclature (12). R50 and d50 have the same definition and
originate from the same 50% ionization value; however, The data presented here demonstrate clearly that nominal
they are determined differently. The beam quality specifi- energy is not a unique or specific indicator of the penetration of
cation, R50, is calculated from an empiric fit outlined in a clinical electron beam. Tables 1–3 show that for any nominal
TG-51, whereas a beam’s d50 is determined by applying energy, the commonly used indicators of clinically relevant
Prepl and L៮ /␳ corrections to its depth ionization readings for beam penetration, d80 and d50, varied considerably. For exam-
the calculation of electron depth dose according to TG-25 ple, the d80 values for nominal 12 MeV electron beams from
(7). Surprisingly, when all beams in this study are consid- Varian accelerators fell in the range of 3.5– 4.8 cm. The d80
 Comparison of electron beam characteristics ● D. FOLLOWILL et al. 909

Fig. 1. Electron beam depth dose characteristics of dmax, d80, and d50 depths plotted against electron beam quality, R50
for Varian, Siemens, and Elekta/Philips accelerators. The function fit and its R2 value are listed.

value is used frequently to determine the prescription depth,
meaning that the choice of energy is based on agreement
between the greatest depth of the target volume and the depth
of the 80% isodose line. Our data, however, show that this
agreement does not hold. Indeed, we show that the use of
estimators or rules of thumb may lead to significant errors in
dose prescription. In the example above, the rule of thumb
would indicate the prescription depth for a 12 MeV beam
should be 12/3, or 4 cm. Measurements showed that the actual
depth of the 80% isodose line varied from 0.5 cm shallower to
0.8 cm deeper than the estimate.
In addition, the data in Tables 1 and 2 show that the depth
of the 50% isodose line for a 12 MeV beam was as little as 4.2
cm. This suggests that the use of a rule of thumb may result in
the selection of an electron beam having a dose at the prescrip-
tion depth of only slightly more than 50%. In fact, because of
the wide range of d80 and d50 values for the 12 MeV beams, 1
linac’s d80 is another’s d50. This observation is true for the
Varian, Siemens, and Elekta/Philips accelerators.
However, the data in Fig. 1 demonstrate a reproducible
and linear relationship between R50, the TG-51 electron
beam quality specification, and the d80 and d50 values. The
linear relationship between R50 and d50 is used by the
TG-51 calibration protocol for the calibration of electron
beams. This linear relationship is also used by RPC physi-
cists during their on-site dosimetry review visits as a redun-
dant check of the electron depth dose ion chamber measure-
ments. For clinical purposes, specification of prescription
depth should be based on actual measured data, not on rules
of thumb or other estimators. Reliance on nominal electron
energies and rules of thumb can be very misleading, as
shown in Fig. 2.
Physicists should find the data in Fig. 1 and Tables 1–3
useful as a quality assurance tool for the verification of
measured electron depth dose data. Once electron depth
dose curves are measured, institutional data can be com-
pared to the data presented here as a redundant check to
identify potential errors in the dosimetry data.
CONCLUSION
The manufacturer’s nominal energy is not an appropriate
electron beam indicator for clinical radiotherapy. Beams of the
same nominal electron energy from different accelerators may
exhibit significant differences in electron beam penetration,
even for the reference 10 cm ⫻ 10 cm field size.
Radiation oncologists and physicists should be familiar with
the specific depth dose characteristics for each electron beam
used in their clinic. Clinical decisions as to which electron
910 I. J. Radiation Oncology ● Biology ● Physics
Fig. 2. The range of electron beam depth dose characteristics of dmax, d80, and d50 depths plotted against manufacturer’s
nominal electron energy value for Varian, Siemens, and Elekta/Philips accelerators.
beam is best used for an electron treatment should be made
based on the actual measured depth dose data rather than rules
of thumb. Clinicians are cautioned when using these results, as
REFERENCES
1. Kirby TH, Gastorf RJ, Hanson WF, et al. Electron beam
central axis depth dose measurements. Med Phys 1985;12:
357–361.
2. American Association of Physicists in Medicine, RTC Task
Group 21. A protocol for the determination of absorbed dose
from high-energy photon and electron beams. Med Phys 1983;
10:741–771.
3. Davis DS, Followill DS, Kennedy P, Hanson WF. Electron
percent depth dose and cone ratio data from various machines.
Med Phys 1995;22:1007.
4. Klevenhagen SC. Physics and dosimetry of therapy electron
beams. Madison, WI: Medical Physics Publishing, 1993.
5. International Electrotechnical Commission (IEC). Interna-
tional standard—medical electrical equipment. Geneva: IEC
60601-2-1; 1998.
6. Almond PR, Biggs PJ, Coursey BM, et al. AAPM’s TG-51
protocol for clinical reference dosimetry of high-energy pho-
ton and electron beams. Med Phys 1999;26:1847–1870.
7. Khan FM, Doppke KP, Hogstrom KR, et al. Clinical electron-
beam dosimetry: Report of AAPM Radiation Therapy Com-
mittee Task Group 25. Med Phys 1991;18:73–109.
Volume 59, Number 3, 2004
the data presented in this work are for a single field size and do
not contain the effects of smaller field sizes, beam obliquity, or
increased treatment distances.
8. Brahme A, Svensson H. Specification of electron beam quality
from central-axis depth absorbed-dose distribution. Med Phys
1976;3:95–102.
9. International Atomic Energy Agency. Absorbed dose deter-
mination in external beam radiotherapy. An international code
of practice for dosimetry based on standards of absorbed dose
to water. Technical report series no. 398. Vienna: International
Atomic Energy Agency, 2000.
10. Vaeth JM, Meurk ML. Electron beam therapy in clinical
practice. In: Leibel SA, Phillips TL, editors. Textbook of
radiation oncology, 1st ed. Philadelphia: WB Saunders, 1998:
p. 216.
11. Kennedy PM, Hanson WF. A review of high-energy photon
beam characteristics measured by the Radiological Physics
Center. Med Phys 1992;19:838.
12. International Commission on Radiation Units and Measure-
ments. Radiation dosimetry: Electron beams with energies
between 1 and 50 MeV. ICRU report 35. Bethesda, MD:
International Commission on Radiation Units and Measure-
ment, 1984.