Quantifying the predictability of diaphragm motion during respiration
with a noninvasive external marker
S. S. Vedam
Department of Biomedical Engineering and Department of Radiation Oncology, Virginia Commonwealth
University, Richmond, Virginia
V. R. Kini and P. J. Kealla)
Department of Radiation Oncology, Virginia Commonwealth University, Richmond, Virginia
V. Ramakrishnan
Department of Biostatistics, Virginia Commonwealth University, Richmond, Virginia
H. Mostafavi
Ginzton Technology Center, Varian Medical Systems Inc., Palo Alto, California
R. Mohan
Department of Radiation Oncology, Virginia Commonwealth University, Richmond, Virginia
and Department of Radiation Physics, M.D. Anderson Cancer Center, Houston, Texas
!Received 11 June 2002; accepted for publication 13 January 2003; published 17 March 2003"
The aim of this work was to quantify the ability to predict intrafraction diaphragm motion from an
external respiration signal during a course of radiotherapy. The data obtained included diaphragm
motion traces from 63 fluoroscopic lung procedures for 5 patients, acquired simultaneously with
respiratory motion signals !an infrared camera-based system was used to track abdominal wall
motion". During these sessions, the patients were asked to breathe either !i" without instruction, !ii"
with audio prompting, or !iii" using visual feedback. A statistical general linear model was formulated to describe the relationship between the respiration signal and diaphragm motion over all
sessions and for all breathing training types. The model parameters derived from the first session for
each patient were then used to predict the diaphragm motion for subsequent sessions based on the
respiration signal. Quantification of the difference between the predicted and actual motion during
each session determined our ability to predict diaphragm motion during a course of radiotherapy.
This measure of diaphragm motion was also used to estimate clinical target volume !CTV" to
planning target volume !PTV" margins for conventional, gated, and proposed four-dimensional
!4D" radiotherapy. Results from statistical analysis indicated a strong linear relationship between
the respiration signal and diaphragm motion (p!0.001) over all sessions, irrespective of session
number (p"0.98) and breathing training type (p"0.19). Using model parameters obtained from
the first session, diaphragm motion was predicted in subsequent sessions to within 0.1 cm !1 #" for
gated and 4D radiotherapy. Assuming a 0.4 cm setup error, superior–inferior CTV–PTV margins of
1.1 cm for conventional radiotherapy could be reduced to 0.8 cm for gated and 4D radiotherapy.
The diaphragm motion is strongly correlated with the respiration signal obtained from the abdominal wall. This correlation can be used to predict diaphragm motion, based on the respiration signal,
to within 0.1 cm !1 #" over a course of radiotherapy. © 2003 American Association of Physicists
in Medicine. $DOI: 10.1118/1.1558675%
I. INTRODUCTION
Reproducibility of target position during and between subsequent fractions of a radiotherapy treatment procedure plays a
critical role in increasing the accuracy of the treatmentplanning process. The different components that affect such
reproducibility of target position include the beam-bony
anatomy alignment, displacement of internal organs between
fractions, and internal organ motion within a fraction. Depending on the site being treated, each of the abovementioned components contributes, in varying proportions,
to the margins that are to be added around the clinical target
volume !CTV",1,2 creating the planning target volume !PTV"
to ensure adequate coverage throughout the entire course of
505
Med. Phys. 30 „4…, April 2003
the treatment. One of the principal causes of intrafraction
internal anatomy motion is respiration. Treatment sites that
are most significantly affected by such motion include the
lung, breast, and liver. Several studies, conducted to examine
the extent of diaphragm excursion due to normal respiration,
report the range of motion to be between 0.4 and 3.8 cm3–7
in the superior–inferior !SI" direction.
During conventional treatment planning to sites affected
by respiratory motion, a significant margin is added around
the CTV in order to ensure complete coverage of the CTV as
it moves due to respiration within a treatment fraction. It
therefore follows that a large amount of surrounding normal
tissue is irradiated, thus increasing the amount of healthy
0094-2405Õ2003Õ30„4…Õ505Õ9Õ$20.00
© 2003 Am. Assoc. Phys. Med.
505
506
Vedam et al.: Predicting diaphragm motion from a respiration signal
lung irradiated and limiting the maximum dose that can be
prescribed to the tumor itself. Accounting for such motion
during treatment, therefore, has the potential to reduce margins drawn around the CTV, resulting in a lower dose to
normal tissues !e.g., lung" and thus a lower risk of treatmentinduced complications.8 –14
Among the techniques that explicitly account for intrafraction motion are breath-hold,6,7,15–20 respirationgating,21–36 and 4D37 or tumor-tracking38 – 40 techniques.
Breath-hold techniques either actively or passively suspend
the patient’s respiration and allow treatment during this interval. A study41 examining intra-and interfraction reproducibility of diaphragm and liver positions during breath holds
has indicated that, while diaphragm position within a fraction
can be reproduced satisfactorily, daily imaging and repositioning are still required in order to achieve any appreciable
reduction in treatment margins. Respiration-gating methods
periodically turn the beam on when the patient’s respiration
signal is in a certain part of the breathing cycle !generally
end-inhale or end-exhale". 4D methods propose to track the
tumor with the radiation beam as the tumor moves during the
respiration cycle. These techniques require acquisition of
some form of respiration signal !infrared reflective markers,
spirometry, strain gauges, video tracking of chest outlines
and fluoroscopic tracking of implanted markers are some of
the techniques employed to date", which is assumed to be
correlated with internal anatomy motion. It is obvious that
fluoroscopic tracking of implanted markers is well correlated
with internal anatomy motion. However, due to the complexity of respiration, it is not so obvious whether the respiration
signal from an external marker is correlated with internal
anatomy motion.
Mageras et al.,42 in a six-patient fluoroscopic study, presented evidence of a strong correlation between diaphragm
motion and the respiration signal. It was also shown that
verbal instructions improved the regularity of the respiration
signal. Recently, an attempt43 was also made to analyze motion of the actual tumor in the frequency domain from multiple fluoroscopic videos. In Mageras’s study, skin markers
served as the respiration signal. Results obtained were on
similar lines with a tumor motion range of 0 to 5 cm and an
average breathing cycle of 2.8 s. A maximum time delay of
nearly a third of a breathing cycle was also observed. In a
further study,44 respiratory movement during gated radiotherapy was evaluated by employing film and electronic portal imaging. Among the several parameters examined in this
study were intra- and interfraction variability in diaphragm
position due to respiration motion. Ozhasoglu et al.,38 in a 5
patient study, examined breathing patterns by employing simultaneous fluoroscopic visualization of internal fiducial
markers and external respiration signal recordings !spirometry and optical markers" in real time. They reported significant variations in the breathing patterns, including transient
temporal shifts between breathing traces generated by the
external marker motion and the motion of the internal
anatomy !tumor".
The current literature dealing with respiration-induced organ motion has not statistically evaluated the relationship
Medical Physics, Vol. 30, No. 4, April 2003
506
between the diaphragm motion and the respiration signal for
multiple patients, breathing regimens and sessions simulating
a course of radiotherapy.
Thus, the main aims of our research were to:
!1" Examine the nature of the relationship between diaphragm motion and the respiration signal from multiple
repeat fluoroscopic sessions for different patients and
breathing training regimens.
!2" Use the relationship determined from one ‘‘simulation’’
session to predict diaphragm motion in subsequent
‘‘treatment’’ sessions.
!3" Quantify and compare the potential SI margins that need
to be added to the CTV for conventional, gated, and 4D
radiotherapy.
II. METHODS AND MATERIALS
A. Data acquisition
Sixty-three fluoroscopy movies, typically 30 s each, were
recorded from 5 lung cancer patients who participated in a
study protocol. Each patient underwent at least two and up to
five complete sessions !on different days, with not more than
a week separating subsequent sessions" of breathing monitoring !with and without training". Each breathing monitoring
session consisted of simultaneous recording of the respiration signal and the fluoroscopy movie, while the patient was
asked to breathe !1" normally !no instructions", !2" according
to audio prompting !periodic ‘‘Breathe in’’ and ‘‘Breathe
out’’ instructions" presented at a rate comfortable for each
patient, and !3" by observing a visual representation of the
motion trace on a screen and regulating the excursion of the
trace between two prespecified motion limits !visual feedback".
The patients were set up in the simulator room by aligning
the room lasers with the skin marks as per their treatment
chart. The Varian !Varian Medical Systems, Palo Alto,
CA 94304-1129" Real Time Position Management !RPM"
system was used to acquire !at 30 Hz" the respiration signal,
R(t), and also to simultaneously record a stream of lung
fluoroscopy images !at 10 Hz". During each breathing session, the infrared reflective marker was placed at the same
point on the patient’s surface, where the magnitude of respiration motion was observed to be high, generally midway
between the patient’s xyphoid process and umbilicus. Permanent marker was used to aid in the repositioning of the
marker from day to day. A typical patient setup for a study
session is shown in Fig. 1.
The respiration signal is the anterior–posterior !AP" component of the infrared reflective marker motion, as measured
by the RPM system. Custom-written image-analysis software
was developed to allow a user to identify the portion of the
diaphragm to be monitored on the first frame of the fluoroscopy movie. The software then automatically updates the SI
position of the selected portion of the diaphragm on subsequent frames, thus yielding a motion trace of the diaphragm,
F act(t). 31 For each fluoroscopy session, the apex and the lateral and medial aspects of the diaphragm were tracked when
507
Vedam et al.: Predicting diaphragm motion from a respiration signal
FIG. 1. Typical patient setup for simultaneous recording of the respiration
signal and fluoroscopy movie. A reflective marker block placed midway
between the patient’s xyphoid process and umbilicus is used in the recording
of the respiration signal. The headphones are for audio prompting, and the
LCD screen is for visual feedback. Inset is the patient view of the LCD
screen used for the visual feedback.
possible. To ensure the fidelity of the automatic diaphragm
position determination, the detected position was checked on
a frame-by-frame basis, and repeat measures also yielded
identical results.
B. Relationship of respiration signal to internal
anatomy motion: Statistical validation of a general
linear model
C. Prediction of diaphragm motion
If there is a linear relationship between the internal
anatomy motion as observed on fluoroscopy, F act(t), and the
respiration signal, R(t), then the following equation will
hold:
F act! t " #F act"c$ ! R ! t# & " #R " ,
!1"
where c is the average ratio between the internal motion and
respiration signal, and & is the time delay between the diaphragm motion and respiration signal. F act and R are the
mean !integrated over time" values of the internal motion and
Medical Physics, Vol. 30, No. 4, April 2003
respiration signal, respectively, calculated for each fluoroscopy session of interest. By including the normalization
terms F act and R , we are removing daily systematic differences in the mean position of the diaphragm, and instead
concentrating on intrafraction variations due to respiration
only.
The average ratio, c, was determined by calculating the
ratio of the time-integrated absolute value of the left-hand
side of Eq. !1" to the time-integrated absolute value of the
right-hand side of Eq. !1". The time delay for the simulation
session, & 0 , was determined by visually inspecting the overlay of the diaphragm motion and respiration signal traces and
searching for any systematic temporal shift between the two
traces. This time delay, which had been observed in previous
studies,31,38,42,43 was determined to be 0 for all the patients
enrolled in this study. Though the time delay was observed to
be zero for the patient data obtained here, it is still included
in the prediction method for generality.
For a given patient, the average ratio, c, and time delay, &,
values established from the initial !simulation" session can
be used in subsequent sessions to predict the motion of the
diaphragm, F pred(t), during subsequent !treatment" sessions.
The applicability of predicting the diaphragm motion from
the first session to subsequent sessions is the analogy of obtaining these parameters from a simulation session and applying them throughout the course of treatment. Thus, from
the respiration trace of subsequent sessions we can predict
the diaphragm motion using
F pred! t " #F pred"c 0 $ ! R ! t# & 0 " #R " ,
The first step was to statistically validate the assumption
that the relationship between the respiration signal and internal anatomy motion could be expressed satisfactorily with
the mathematical model of the form expressed in Eq. !1". A
linear model relating the two variables F act(t)#F act and
R(t# & )#R was statistically fitted. Since this correlation
may be influenced by the variability in breathing training
types and sessions, these variables were included in the linear regression model to obtain an adjusted estimate of the
correlation between the two variables. As the same subjects
were repeated over the sessions and training types, within
subject correlations had to be accounted for in the linear
model. This was achieved using the general linear model.
507
!2"
where c 0 and & 0 are the average ratio and time delay values
established from the initial session.
The validity of the predicted trace, F pred(t), to represent
the actual trace, F act(t), can be quantified by taking the standard deviation of the relative positional differences between
these traces, # F pred#F act, for a given session. Obviously, the
smaller the value of # F pred#F act, the better the internal motion
can be predicted from the respiration signal. This internal
motion prediction can be used in 4D radiotherapy to track
the tumor based on the respiration signal.
Also of relevance for gated radiotherapy is the quantification of the internal motion prediction during the gating window when the beam is on. This quantity, which is denoted as
# FGate
, indicates the variation between the predicted tupred#F act
mor motion and actual tumor motion within a specified gating window. For gated radiotherapy, respiration-induced motion occurs during the finite window. As the duty cycle
increases, the residual motion within the gating window, defined as # FGate , also increases. $The duty cycle, or efficiency,
act
as defined here is the ratio of beam on time !when the respiration signal falls within the gating window" to total beam
delivery time. Delivering 150 monitor units at 300 monitor
units per second without gating would normally take 30 s.
With a 25% duty cycle, this will take 2 min.% For this study,
the gating windows were chosen at 20%, 35%, and 50% duty
cycles for both inhale and exhale to cover the range of clinically used values.
Vedam et al.: Predicting diaphragm motion from a respiration signal
508
508
TABLE I. The standard deviation !1#" of the measured diaphragm motion,
# F act, for each patient and breathing training type, averaged over multiple
fluoroscopy sessions. All values are in centimeters.
Patient
Normal
breathing
Audio
prompting
Visual
feedback
1
2
3
4
5
All
0.60
0.27
0.34
0.24
0.44
0.36
0.65
0.46
0.53
0.70
1.00
0.71
0.48
0.29
0.38
0.23
0.82
0.47
D. Calculation of superior–inferior CTV to PTV
margins
Because intrafraction motion is one component of the
margin that needs to be added to the CTV to obtain the PTV,
gating and 4D methods that explicitly account for tumor motion during delivery will have a small CTV–PTV margin.
However, the scale of this margin reduction depends on the
relative magnitudes of the other margin components. In our
study, the margins were calculated in such a way that the
CTV lies within the PTV for 95% of the treatment duration.
This approach requires that the uncertainties are known.2
Thus, the following equations for conventional (M Conv),
gated (M Gate), and 4D (M 4D) margins were derived:
2
M Conv"2$ !# setup
% # F2 ,
!3"
2
% ! # FGate #F " 2 % ! # FGate" 2 ,
M Gate"2$ !# setup
pred
act
act
!4"
2
% ! # FGate #F " 2 ,
M 4D"2$ !# setup
pred
act
!5"
act
where # setup includes both the variation in patient position
!beam-bony anatomy alignment" and the bony anatomytumor alignment !bony anatomy-diaphragm in this case".
# F act represents the variation of the actual diaphragm motion
throughout the entire breathing session.
For this study, the setup error used was the realistic value
of 0.4 cm.45 However, because improved technology and
techniques may reduce setup errors in the future, setup error
values of 0.2 cm were also used for the margin calculations.
III. RESULTS
A. Quantification of diaphragm motion
For the purposes of margin calculation, the complex nature of the diaphragm motion $see, for example, Fig. 4!a"%
can be quantified by calculating the standard deviation of the
motion !assuming that the probability distribution of the position with time is approximately normally distributed".46
Table I provides values of the diaphragm motion !1 #" for
each patient and for the different types of breathing training,
averaged over multiple fluoroscopy sessions. Diaphragm motion during visual feedback was comparable to that obtained
during normal breathing. However, audio prompting inMedical Physics, Vol. 30, No. 4, April 2003
FIG. 2. Comparison of the average variation of diaphragm motion, # FGate
,
act
within different gate intervals at inhale and exhale for all patients during
normal respiration. Error bars represent a variation of 1#. The increase in
motion with duty cycle is due to the residual motion within the gating
window.
creased the range of/amplitude of diaphragm motion. This
result was found to be consistent with the observations made
in a previous study.42
Diaphragm motion within different gate intervals !20%,
35%, and 50% duty cycle at inhale and exhale" for the freebreathing data is shown in Fig. 2. In general, motion within
gate intervals at exhale was less than that at corresponding
gate intervals at inhale. An increase in the width of the duty
cycle, or gating window resulted in an increase in motion
extent, as shown in a previous study.31 Note that the motion
with a 100% gate interval is the same as that obtained over
the entire breathing session.
B. Relationship of respiration signal to internal
anatomy motion: Statistical analysis using
the repeated measures mixed effects model
A repeated measures mixed effects model was fitted using
the PROC MIXED procedure in SAS !SAS Institute Inc, Version 8.2e, 2000, Cary NC" to statistically evaluate Eq. !1". A
scatter plot of the diaphragm motion, F act , and the respiration signal, R, for every measurement point obtained
!'10 000" is shown in Fig. 3. An analysis of the data shows
that even though the multiple recording sessions and breathing instruction types changed the respiration patterns, the
relationship between respiration signal and diaphragm motion did not differ significantly over the different breathing
types (p"0.19) or session number (p"0.98), as shown in
Table II. However, the diaphragm motion is strongly correlated with the respiration signal (p!0.0001), and the use of
a linear model is justified. In addition, results obtained (p
!0.0001) when Time !instant of time during the breathing
cycle" is considered as a factor and those obtained for the
interaction variables—Time, Respiration signal$Time,
509
Vedam et al.: Predicting diaphragm motion from a respiration signal
509
FIG. 3. Scatter plots representing !a"
the overall relationship between diaphragm position, F act , and respiration
signal, R, for all data points !'10 000"
from all !63" of the fluoroscopic sessions and !b" patient-specific relationships between diaphragm position,
F act , and respiration signal, R. The
line of best fit in each case and the
correlation values are also displayed.
Note that positive values on both axes
correspond with the inhale phase of
the breathing cycle.
Respiration signal$Session number, and Respiration signal
$Breathing training type—do suggest the existence of a significant variation in the slope of the line of fit with respect to
time instant during the breathing cycle, session number, and
breathing training type.
To interpret the statistical results in a more visual manner,
the diaphragm motion and respiration signal traces, representing the session with the lowest diaphragm motionrespiration signal correlation !an outlier" and the highest correlation, are shown in Fig. 4. This figure shows that even for
the lowest observed correlation !0.51", similarities still exist
between the diaphragm motion and respiration signal. Note
that the mean correlation observed for individual sessions
was 0.94.
Medical Physics, Vol. 30, No. 4, April 2003
TABLE II. Estimation of coefficients for the general linear model !p values",
expressing the relationship between respiration signal and diaphragm motion. The session number was the session for each patient !1–5" corresponding to the different days on which subsequent fluoroscopy studies were
performed. The breathing training type is either normal breathing, audio
prompting, or visual feedback.
Parameter
p value
Respiration signal
Session number
Breathing training type
Time
Respiration signal$Time
Respiration signal$Session number
Respiration signal$Breathing training type
!0.0001
0.98
0.19
!0.0001
!0.0001
!0.0001
!0.0001
Vedam et al.: Predicting diaphragm motion from a respiration signal
510
510
FIG. 4. The diaphragm position !dotted line", F act , and
respiration signal !solid line", R, plotted against time for
!a" the lowest correlation coefficient value !0.51" and
!b" the highest correlation coefficient value !0.99". Note
that positive values along the position axis correspond
with the inhale phase of the breathing cycle.
C. Prediction of diaphragm motion
IV. DISCUSSION
The ability of the respiration signal to predict the diaphragm motion based on Eq. !2" is summarized in Table III.
These results were obtained by including all sessions of the
same training type for all of the patients in the analysis. The
results show a consistent ability to predict the diaphragm
motion to within '0.1 cm !1 #", independent of breathing
training.
These results represent our prediction ability over the entire breathing cycle and are applicable to 4D treatments. A
similar analysis performed for gated treatments is shown in
Fig. 5, again showing the ability to predict the diaphragm
motion to within 0.1 cm at different duty cycle values.
Many gated, breath-hold and 4D radiotherapy techniques
rely on using a respiration signal as a surrogate for or predictor of tumor motion. Verifying this assumption was an
aim of this research. In this study, diaphragm motion and
respiration signal !abdominal wall motion" were correlated in
over 60 fluoroscopy sessions. Statistical analysis showed that
these two quantities are strongly correlated. Ozhasoglu
et al.38 also report in their work a similar correlation between
the superior–inferior displacement of a pancreas tumor and
the anterior–posterior displacement of an external chest
marker. They also suggest that if such a stationary relationship can be established, then a correlation between the external chest marker and the tumor motion can be derived from
three or four randomly timed observations of both signals
and hence be extrapolated into the future. We adopted a similar approach in our study when we first established such a
stationary relationship between the external respiration signal and the diaphragm !as a surrogate for tumor" motion. The
correlation derived, based on such a stationary relationship
for a ‘‘simulation’’ session, was then used to predict dia-
D. Calculation of superior–inferior margins for
different treatment modalities
The values obtained for predicted diaphragm motion were
entered into Eqs. !3"–!5" to determine the CTV–PTV margins that would be needed to have the CTV within the PTV
95% of the time. The results are shown in Fig. 6. All margin
reductions from the conventional values shown are significant at the 1% level, except for the results of gated exhale for
the 50% duty cycle.
With gated therapy, the degree of margin reduction decreased with an increase in the width of the gate !duty cycle"
due to residual motion within the gating window. The potential margin reduction was greater as setup error became less
significant. 4D radiotherapy allows the greatest margin reduction, since there is no residual motion within the gating
window.
TABLE III. Measure of ability to predict !1#" diaphragm motion, # F pred#F act,
in multiple sessions !cm".
Patient
Normal
breathing
Audio
prompting
Visual
feedback
1
2
3
4
5
All
0.1
0.08
0.05
0.09
0.11
0.09
0.1
0.09
0.07
0.11
0.1
0.09
0.08
0.07
0.15
0.07
0.14
0.11
Medical Physics, Vol. 30, No. 4, April 2003
FIG. 5. Average ability to predict the diaphragm position, # FGate
, within
pred#F act
different gating windows at inhale and exhale for all patients during normal
respiration. Error bars represent a variation of 1#. Note the use of the same
scale as in Fig. 2, in order to illustrate consistency in the ability to predict
diaphragm motion within different gating widows.
511
Vedam et al.: Predicting diaphragm motion from a respiration signal
FIG. 6. Comparison of projected population averaged superior–inferior !SI"
margins to be added to the CTV for conventional (M Conv), gated (M Gate),
and 4D (M 4D) treatment modalities for patients during normal respiration.
The gated margins are calculated for several different duty cycle !DC" values. Projected margins are lower for gated and 4D radiotherapy than conventional therapy. Gating at exhale allows slightly smaller margins than
gating at inhale.
phragm motion in future ‘‘treatment’’ sessions, based on the
respiration signal from that same treatment session alone, to
within 0.1 cm !1 #" over a course of radiotherapy. Thus,
methods such as gated and 4D radiotherapy that rely on a
relationship between the internal anatomy motion and the
respiration signal have the potential for safe margin reduction as well as the potential for target dose escalation and/or
lower treatment-related complications.
Intrafraction respiratory motion and the potential for reduction of such motion during respiratory gated radiotherapy
has been studied by Ford et al.44 using fluoroscopy, respiratory gated film, and electronic portal imaging data acquired
from eight patients. In their study, the patient averaged intrafraction diaphragm range of motion of 0.69 cm during
normal respiration was reduced to 0.26 cm with respiratory
gating. Our corresponding results, presented as standard deviations of motion, are 0.36 and '0.11 cm, respectively, and
are therefore consistent with the range of motion values reported by Ford et al.44
A limitation of the current work is that only the motion of
one anatomical structure !the diaphragm" and only one respiration signal !abdominal wall motion at the midway point
between the umbilicus and xyphoid process" have been correlated. However, breathing patterns show a lot of variations
over the course of time. Hence, longer treatment times directly influence the stationary nature of the correlation between the respiration signal and tumor/diaphragm motion.
Also, the presence of a phase shift31,38,42,43 !transient or otherwise" between the external and internal motion signals, as
reported in the literature, has to be accounted for. In the
current study, we observe a phase shift. Such a shift, if and
when it occurs, will affect the results obtained from this
study. Of most importance, therefore, is correlating the tumor
motion with the respiration signal. This correlation may be
Medical Physics, Vol. 30, No. 4, April 2003
511
possible with the development of respiration-correlated
‘‘4D’’ CT methods.47–53
Patient training using audio and visual aids does affect
amplitude of breathing patterns, as is evident from results in
Sec. III A. Audio instructions tend to increase the amplitude
of breathing while visual aids could potentially be used to
limit diaphragm excursion to predetermined levels. However,
training does not seem to have a significant effect on our
ability to predict the diaphragm motion during treatment sessions.
Intrafraction motion, which was the main focus of our
work, is just one of several components that contribute to
increased margins. The issues of interfraction motion and
setup error have already been studied and reported in literature. Balter et al.54,55 have analyzed diaphragm position on
daily radiographs from eight patients to assess setup variations in liver position during breath-hold treatments with the
ABC device. Analysis of repeat 3D CT scan data obtained
from this method indicated a reduction in setup errors !1 #"
from 6.7 to 3.5 mm !superior–inferior, SI", 4 to 2.1 mm
!left–right, LR", and 3.8 to 2.3 mm !anterior–posterior, AP"
respectively. Dawson et al.,41 using a similar technique, have
reported interfraction SI reproducibility !1 #" of 4.4 mm. In
a study of interfraction diaphragm variation during respiratory gated radiotherapy, Ford et al.44 reported patient averaged interfraction diaphragm variability of 2.8&1.0 mm,
based on data obtained from gated localization films. The
study further concluded that most of the variability in diaphragm position occurs due to setup errors rather than respiratory motion. Thus, while gated and 4D treatments exhibit a
potential for margin reduction as compared with conventional treatments, this margin reduction is still limited by the
magnitude of the setup error.
ACKNOWLEDGMENTS
The authors acknowledge Varian Medical Systems for the
loan of an RPM gating system, engineering support, and a
Varian Medical Systems internship. This research was aided
in part by Grant No. IRG-100036 from the American Cancer
Society, and Grant No. R01 CA 93626 from the National
Cancer Institute. Thank you to Devon Murphy for reviewing
and improving the clarity of this manuscript.
a"
Author to whom all correspondence should be addressed; electronic mail:
[email protected]
1
‘‘Prescribing, recording and reporting photon beam therapy,’’ ICRU Report No. 50 !International Commission on Radiation Units and Measurements, Bethesda, MD, 1993".
2
‘‘Prescribing, recording and reporting photon beam therapy,’’ ICRU Report No. 62 !Supplement to ICRU Report No. 50" !International Commission on Radiation Units and Measurements, Bethesda, MD, 1999".
3
P. H. Weiss, J. M. Baker, and E. J. Potchen, ‘‘Assessment of hepatic
respiratory excursion,’’ J. Nucl. Med. 13, 758 –759 !1972".
4
H. W. Korin, R. L. Ehman, S. J. Riederer, J. P. Felmlee, and R. C. Grimm,
‘‘Respiratory kinematics of the upper abdominal organs: A quantitative
study,’’ Magn. Reson. Med. 23, 172–178 !1992".
5
S. C. Davies, A. L. Hill, R. B. Holmes, M. Halliwell, and P. C. Jackson,
‘‘Ultrasound quantitation of respiratory organ motion in the upper abdomen,’’ Br. J. Radiol. 67, 1096 –1102 !1994".
6
J. Hanley, M. M. Debois, D. Mah, G. S. Mageras, A. Raben, K. Rosenzweig, B. Mychalczak, L. H. Schwartz, P. J. Gloeggler, W. Lutz, C. C.
512
Vedam et al.: Predicting diaphragm motion from a respiration signal
Ling, S. A. Leibel, Z. Fuks, and G. J. Kutcher, ‘‘Deep inspiration breathhold technique for lung tumors: The potential value of target immobilization reduced lung density in dose escalation,’’ Int. J. Radiat. Oncol.,
Biol., Phys. 45, 603– 611 !1999".
7
J. M. Balter, K. L. Lam, C. J. McGinn, T. S. Lawrence, and R. K. Ten
Haken, ‘‘Improvement of CT-based treatment-planning models of abdominal targets using static exhale imaging,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 41, 939–943 !1998".
8
S. L. Kwa, J. V. Lebesque, J. C. Theuws, L. B. Marks, M. T. Munley, G.
Bentel, D. Oetzel, U. Spahn, M. V. Graham, R. E. Drzymala, J. A. Purdy,
A. S. Lichter, M. K. Martel, and R. K. Ten Haken, ‘‘Radiation pneumonitis as a function of mean lung dose: An analysis of pooled data of 540
patients,’’ Int. J. Radiat. Oncol., Biol., Phys. 42, 1–9 !1998".
9
M. V. Graham, J. A. Purdy, B. Emami, W. Harms, W. Bosch, M. A.
Lockett, and C. A. Perez, ‘‘Clinical dose-volume histogram analysis for
pneumonitis after 3D treatment for non-small cell lung cancer
!NSCLC",’’ Int. J. Radiat. Oncol., Biol., Phys. 45, 323–329 !1999".
10
L. Marks, ‘‘Dosimetric predictors of radiation-induced lung injury,’’ Int.
J. Radiat. Oncol., Biol., Phys. 54, 313–316 !2002".
11
M. K. Martel, R. K. Ten Haken, M. B. Hazuka, A. T. Turrisi, B. A. Fraass,
and A. S. Lichter, ‘‘Dose-volume histogram and 3-D treatment planning
evaluation of patients with pneumonitis,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 28, 575–581 !1994".
12
D. Oetzel, P. Schraube, F. Hensley, G. Sroka-Perez, M. Menke, and M.
Flentje, ‘‘Estimation of pneumonitis risk in three-dimensional treatment
planning using dose-volume histogram analysis,’’ Int. J. Radiat. Oncol.,
Biol., Phys. 33, 455– 460 !1995".
13
J. G. Armstrong, M. J. Zelefsky, S. A. Leibel, C. Burman, C. Han, L. B.
Harrison, G. J. Kutcher, and Z. Y. Fuks, ‘‘Strategy for dose escalation
using 3-dimensional conformal radiation therapy for lung cancer,’’ Ann.
Oncol. 6, 693– 697 !1995".
14
M. L. Hernando, L. B. Marks, G. C. Bentel, S. M. Zhou, D. Hollis, S. K.
Das, M. Fan, M. T. Munley, T. D. Shafman, M. S. Anscher, and P. A.
Lind, ‘‘Radiation-induced pulmonary toxicity: A dose-volume histogram
analysis in 201 patients with lung cancer,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 51, 650– 659 !2001".
15
H. D. Kubo, P. M. Len, S. Minohara, and H. Mostafavi, ‘‘Breathingsynchronized radiotherapy program at the University of California Davis
Cancer Center,’’ Med. Phys. 27, 346 –353 !2000".
16
K. E. Rosenzweig, J. Hanley, D. Mah, G. Mageras, M. Hunt, S. Toner, C.
Burman, C. C. Ling, B. Mychalczak, Z. Fuks, and S. A. Leibel, ‘‘The
deep inspiration breath-hold technique in the treatment of inoperable nonsmall-cell lung cancer,’’ Int. J. Radiat. Oncol., Biol., Phys. 48, 81– 87
!2000".
17
D. Mah, J. Hanley, K. E. Rosenzweig, E. Yorke, L. Braban, C. C. Ling, S.
A. Leibel, and G. Mageras, ‘‘Technical aspects of the deep inspiration
breath-hold technique in the treatment of thoracic cancer,’’ Int. J. Radiat.
Oncol., Biol., Phys. 48, 1175–1185 !2000".
18
E. A. Barnes, B. R. Murray, D. M. Robinson, L. J. Underwood, J. Hanson, and W. H. Roa, ‘‘Dosimetric evaluation of lung tumor immobilization using breath hold at deep inspiration,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 50!4", 1091–1098 !2001".
19
J. W. Wong, M. B. Sharpe, D. A. Jaffray, V. R. Kini, J. S. Robertson, J. S.
Stromberg, and A. A. Martinez, ‘‘The use of active breathing control
!ABC" to reduce margin for breathing motion,’’ Int. J. Radiat. Oncol.,
Biol., Phys. 44, 911–919 !1999".
20
J. S. Stromberg, M. B. Sharpe, L. H. Kim, V. R. Kini, D. A. Jaffray, A. A.
Martinez, and J. W. Wong, ‘‘Active breathing control !ABC" for
Hodgkin’s disease: Reduction in normal tissue irradiation with deep inspiration and implications for treatment,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 48, 797– 806 !2000".
21
C. J. Ritchie, J. Hseih, M. F. Gard, J. D. Godwin, Y. Kim, and C. R.
Crawford, ‘‘Predictive respiratory gating: A new method to reduce motion artifacts on CT scans,’’ Radiology 190, 847– 852 !1994".
22
M. Mori, K. Murata, M. Takahashi, K. Shimoyama, T. Ota, R. Morita,
and T. Sakamoto, ‘‘Accurate contiguous sections without breath-holding
on chest CT: Value of respiratory gating and ultrafast CT,’’ Am. J. Roentgenol. 162, 1057–1062 !1994".
23
H. D. Kubo and B. C. Hill, ‘‘Respiration gated radiotherapy treatment: A
technical study,’’ Phys. Med. Biol. 41, 83–91 !1996".
24
G. S. Mageras, ‘‘Interventional strategies for reducing respiratoryinduced motion in external beam therapy,’’ in The Use of Computers in
Medical Physics, Vol. 30, No. 4, April 2003
512
Radiation Therapy, edited by W. Schlegel and T. Bortfeld !Springer,
Heidelberg, 2000", pp. 514 –516.
25
K. Ohara, T. Okumura, T. Akisada, T. Inada, T. Mori, H. Yokota, and M.
B. Calguas, ‘‘Irradiation synchronized with respiration gate,’’ Int. J. Radiat. Oncol., Biol., Phys. 17, 853– 857 !1989".
26
T. Tada, K. Minakuchi, T. Fujioka, M. Sakurai, M. Koda, I. Kawase, T.
Nakajima, M. Nishioka, T. Tonai, and T. Kozuka, ‘‘Lung cancer: Intermittent irradiation synchronized with respiratory motion—results of a
pilot study,’’ Radiology 207, 779–783 !1998".
27
D. Scaperoth, ‘‘Why respiration gated radiotherapy?,’’ presented at the
RPM Respiration gating training, Knoxville, TN, 2000.
28
C. R. Ramsey, D. Scaperoth, D. Arwood, and A. L. Oliver, ‘‘Clinical
efficacy of respiratory gated conformal radiation therapy,’’ Med. Dosim
24, 115–119 !1999".
29
H. Shirato, S. Shimizu, T. Kunieda, K. Kitamura, M. van Herk, K. Kagei,
T. Nishioka, S. Hashimoto, K. Fujita, H. Aoyama, K. Tsuchiya, K. Kudo,
and K. Miyasaka, ‘‘Physical aspects of a real-time tumor-tracking system
for gated radiotherapy,’’ Int. J. Radiat. Oncol., Biol., Phys. 48, 1187–
1195 !2000".
30
S. Minohara, T. Kanai, M. Endo, K. Noda, and M. Kanazawa, ‘‘Respiratory gated irradiation system for heavy-ion radiotherapy,’’ Int. J. Radiat.
Oncol., Biol., Phys. 47, 1097–1103 !2000".
31
S. S. Vedam, P. J. Keall, V. Kini, and R. Mohan, ‘‘Determining parameters for respiration gated radiotherapy,’’ Med. Phys. 28, 2139–2146
!2001".
32
S. Shimizu, H. Shirato, S. Ogura, H. Akita-Dosaka, K. Kitamura, T. Nishioka, K. Kagei, M. Nishimura, and K. Miyasaka, ‘‘Detection of lung
tumor movement in real-time tumor-tracking radiotherapy,’’ Int. J. Radiat.
Oncol., Biol., Phys. 51, 304 –310 !2001".
33
H. Shirato, S. Shimizu, K. Kitamura, T. Nishioka, K. Kagei, S. Hashimoto, H. Aoyama, T. Kunieda, N. Shinohara, H. Dosaka-Akita, and K.
Miyasaka, ‘‘Four-dimensional treatment planning and fluoroscopic realtime tumor tracking radiotherapy for moving tumor,’’ Int. J. Radiat. Oncol., Biol., Phys. 48, 435– 442 !2000".
34
D. J. Kim, B. R. Murray, R. Halperin, and W. H. Roa, ‘‘Held-breath
self-gating technique for radiotherapy of non-small-cell lung cancer: A
feasibility study,’’ Int. J. Radiat. Oncol., Biol., Phys. 49, 43– 49 !2001".
35
H. D. Kubo and L. Wang, ‘‘Introduction of audio gating to further reduce
organ motion in breathing synchronized radiotherapy,’’ Med. Phys. 29,
345–350 !2002".
36
P. J. Keall, V. R. Kini, S. S. Vedam, and R. Mohan, ‘‘Potential radiotherapy improvements with respiratory gating,’’ Australas. Phys. Eng.
Sci. Med. 25, 1– 6 !2002".
37
P. J. Keall, V. Kini, S. S. Vedam, and R. Mohan, ‘‘Motion Adaptive X-ray
Therapy: A feasibility study,’’ Phys. Med. Biol. 46, 1–10 !2001".
38
C. Ozhasoglu and M. J. Murphy, ‘‘Issues in respiratory motion compensation during external-beam radiotherapy,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 52, 1389–1399 !2002".
39
C. Ozhasoglu, M. J. Murphy, G. Glosser, M. Bodduluri, A. Schweikard,
K. Forster, D. P. Martin, and J. R. Adler, ‘‘Real-time tracking of the tumor
volume in precision radiotherapy and body radiosurgery—A novel approach to compensate for respiratory motion,’’ presented at the Computer
Assisted Radiology and Surgery, San Francisco, 2000.
40
A. Schweikard, G. Glosser, M. Bodduluri, M. J. Murphy, and J. R. Adler,
‘‘Robotic motion compensation for respiratory movement during radiosurgery,’’ Comput. Aided Surg 5, 263–277 !2000".
41
L. A. Dawson, K. K. Brock, S. Kazanjian, D. Fitch, C. J. McGinn, T. S.
Lawrence, R. K. Ten Haken, and J. Balter, ‘‘The reproducibility of organ
position using active breathing control !ABC" during liver radiotherapy,’’
Int. J. Radiat. Oncol., Biol., Phys. 51, 1410–1421 !2001".
42
G. Mageras, E. Yorke, K. E. Rosenzweig, L. Braban, E. Keatley, E. Ford,
S. A. Leibel, and C. C. Ling, ‘‘Fluoroscopic evaluation of diaphragmatic
motion reduction with a respiratory gated radiotherapy system,’’ J. Appl.
Clin. Med. Phys. 2, 191–200 !2001".
43
Q. S. Chen, M. S. Weinhous, F. C. Deibel, J. P. Ciezki, and R. M. Macklis, ‘‘Fluoroscopic study of tumor motion due to breathing: Facilitating
precise radiation therapy for lung cancer patients,’’ Med. Phys. 28, 1850–
1856 !2001".
44
E. Ford, G. Mageras, E. Yorke, K. E. Rosenzweig, R. Wagman, and C. C.
Ling, ‘‘Evaluation of respiratory movement during gated radiotherapy
using film and electronic portal imaging,’’ Int. J. Radiat. Oncol., Biol.,
Phys. 52, 522–531 !2002".
45
L. Ekberg, O. Holmberg, L. Wittgren, G. Bjelkengren, and T. Landberg,
513
Vedam et al.: Predicting diaphragm motion from a respiration signal
‘‘What margins should be added to the clinical target volume in radiotherapy treatment planning for lung cancer?,’’ Radiother. Oncol. 48,
71–77 !1998".
46
T. Bortfeld, K. Jokivarsi, M. Goitein, J. Kung, and S. B. Jiang, ‘‘Effects
of intra-fraction motion on IMRT dose delivery: Statistical analysis and
simulation,’’ Phys. Med. Biol. 47, 2203–2220 !2002".
47
E. Ford, G. Mageras, E. York, and C. C. Ling, ‘‘Respiration-correlated
spiral CT: A method of measuring respiratory-induced anatomic motion
for radiation treatment planning,’’ AAPM Meeting, Montreal, 2002.
48
D. Low, J. Bradley, J. Dempsey, D. Politte, K. Islam, S. Mutic, G. Christensen, and J. Deasy, ‘‘4D !time-dependent" measurements of breathing
motion using multislice CT scanning,’’ AAPM Meeting, Montreal, 2002.
49
S. Vedam, P. Keall, V. Kini, R. Mohan, H. Mostafavi, and H. Shukla, ‘‘4D
CT imaging: Feasibility and preliminary results,’’ AAPM Meeting, Montreal, 2002.
50
P. Zygmanski, ‘‘Experimental quantification and analysis of organ motion
artifacts in conventional CT-simulation: Conditions of 4D CT data acquisition,’’ AAPM Meeting, Montreal, 2002.
Medical Physics, Vol. 30, No. 4, April 2003
51
513
G. S. Mageras, ‘‘Respiration correlated CT techniques for gated treatment
of lung cancer,’’ presented at the 21st Annual ESTRO meeting, Praha,
2002.
52
M. van Herk, C. Scheider, J. J. Sonke, E. Damen, and K. De Jaeger,
‘‘Respiration-correlated CT of lung cancer patients,’’ presented at the 21st
Annual ESTRO meeting, Praha, 2002.
53
S. Vedam, P. Keall, V. Kini, H. Mostafavi, H. Shukla, and R. Mohan,
‘‘Acquiring a 4D CT data set using an external respiratory signal,’’ Phys.
Med. Biol. 48, 45– 62 !2003".
54
J. M. Balter, L. A. Dawson, S. Kazanjian, C. McGinn, K. K. Brock, T.
Lawrence, and R. Ten Haken, ‘‘Determination of ventilatory liver movement via radiographic evaluation of diaphragm position,’’ Int. J. Radiat.
Oncol., Biol., Phys. 51, 267–270 !2001".
55
J. M. Balter, K. K. Brock, D. W. Litzenberg, D. L. McShan, T. S.
Lawrence, R. Ten Haken, C. J. McGinn, K. L. Lam, and L. A. Dawson,
‘‘Daily targeting of intrahepatic tumors for radiotherapy,’’ Int. J. Radiat.
Oncol., Biol., Phys. 52, 266 –271 !2002".