•The measurement‐based approach uses the same fields and number of MUs as applied

for the patient irradiation, but requires an additional treatment plan made and dose
calculation with the CT data of the phantom used for these measurements
•Because the beam‐patient geometry is different from the beam‐phantom geometry, an
additional check of the effect of the patient anatomy on the dose calculation is often
performed with an independent MU/dose verification programme.
•Depending on the philosophy of the institution, the intensity (energy fluence)
distribution of IMRT fields are sometimes checked with a 2‐D detector array without a
phantom in the beam
•Some institutions are using in vivo dosimetry, generally with an EPID, to verify the
complete planning – delivery procedure for the beam‐patient geometry
•In the per‐beam approach a 2‐D detector is placed at a certain depth in a phantom to
measure the 2‐D dose distribution for each beam separately
•During a cumulative dose verification procedure the 2‐D detector may stay at the same
depth in the phantom as during the per‐beam dose verification approach, but now all
beams are delivered under the actual gantry angle
•For both approaches it is necessary to make a new phantom plan, in addition to the
actual patient plan
• Various commercial 2‐D array devices are available using diodes or ionisation
chambers: MapCHECK, MatriXX, Compass, Seven29
• IMRT beams are often designed to add together to produce a uniform dose
distribution in the target volume. Changing gantry angles to 0 degrees, may cause
high dose gradients in the measurement area
• MLC leaf positions at treatment angle may affect plan delivery in a different way
than seen at 0 degrees
•Per‐beam and cumulative dose measurements are both applied for IMRT verification
•For VMAT verification only a check of the cumulative (composite) dose distribution is
possible
•Not all 2‐D detectors are suitable for a measurement of the cumulative dose
distribution because of the angular dependence of the response of the detector system
•For that reason a number of 3‐D detector‐phantom systems have been designed
•These are also needed for the verification of VMAT deliveries
•Anthropomorphic phantoms are useful as a final check of a newly developed
technique, a class solution or a specific complicated field arrangement
•It is often more difficult to position detectors inside an anthropomorphic phantom than
in a slab phantom
•Because of its closer resemblance to the patient situation than when using a slab
phantom, for instance with respect to the presence of tissue heterogeneities,
anthropomorphic phantoms are sometimes preferred, particularly for the verification of
IMRT of lung cancer and cancer in the head‐and‐neck region
•For the routine verification of a large number of patients, a simple flat phantom in
combination with a 2‐D detector (film, matrix array or EPID) is often used
•The determination of accurate absolute dose values from film measurements is
difficult
•Film dose measurements are therefore often normalised on ionisation chamber data
•Details about the characteristics of dosimeters and phantoms and their use for IMRT
verification can be found in: Daniel A. Low, Jean M. Moran, James F. Dempsey, Lei Dong,
Mark Oldham. Dosimetry tools and techniques for IMRT. Med. Phys. 38: 1313 – 1338,
2011
•Setup of film measurements at an accelerator for a per‐beam dose verification
•Calculated dose distribution (left) and film measurement (right)
•At the left side of the film the results of the film calibration are shown
•Example of the variation in film calibration over a period of time
•This slide shows how the cumulative (composite) dose distribution of an IMRT prostate
irradiation is checked
•A new plan is made for the phantom irradiation with the same beam directions and
number of MUs as applied for the patient irradiation
•Note that the phantom represents the pelvic area rather closely
•This slide shows the dose distributions for the patient and phantom irradiation
•Note that the dose distributions are almost the same because the patient and phantom
dimensions are very similar
•In this example the film is positioned in the central plane of the phantom
•In a separate measurement the absolute dose is determined with a calibrated
ionisation chamber positioned at the isocentre
•The ‐evaluation technique is used to compare two dose distributions
•It compares dose differences at specific points, and determines the distance to the
nearest point having the same dose value
•The dose difference criterion is important in low dose gradient regions, while the
distance‐to‐agreement criterion yields valuable information in regions having a high
dose gradient
•Dose difference and distance‐to‐agreement criteria can be selected and often values of
3%/3mm are used
•Points lying inside the ellipse with axes having the criteria values have a gamma value
equal or smaller than one
•The method has been introduced by: D.A Low, W. B. Harms, S. Mutic, and J. A. Purdy.
(1998). “A technique for the quantitative evaluation of dose distributions.” Med Phys 25:
656–661.
•This slide shows the result of a gamma evaluation (bottom) of the dose differences
between a film measurement (top right) and a calculated dose distribution (top left)
using 3%/3mm criteria
•The colour scale of the gamma evaluation at the right indicates the gamma value at a
specific point. In this particular case, the blue colour indicates that all dose values are
within the 3%/3mm criteria
•Note that the colour scale does not allow to assess if the measured dose is higher or
lower than the calculated dose
•In this example the isodose lines of the plan and the film measurement are compared
and the difference computed
•Note that in this example the sign of the difference is also indicated by a difference in
colour; i.e. it is possible to see immediately if the measured dose is higher or lower than
the calculated dose
•Commercial tools have different options for presenting dose differences
•By performing film measurements in more than one plane, e.g. two orthogonal planes,
it is possible to obtain information about the delivery of the 3D dose distribution in the
phantom
•There exist no guidelines for the use of pass/fail decisions for patient‐specific physics
QA of IMRT based on gamma evaluation criteria
•Systematic errors discovered by patient‐specific physics QA of IMRT due to non‐optimal
modeling of the TPS should, however, be corrected
•The decreasing availability of well‐maintained film processors, and the increasing
prevalence of digital imaging resulted in the growing use of 2‐D matrix detectors instead
of film for patient‐specific physics QA of IMRT
•Details about the characteristics of 2‐D matrix detectors and their use for IMRT
verification can be found in: Daniel A. Low, Jean M. Moran, James F. Dempsey, Lei Dong,
Mark Oldham. Dosimetry tools and techniques for IMRT. Med. Phys. 38: 1313 – 1338,
2011
•Diodes have small physical size (0.8X0.8 mm)
•As a consequence, if diodes are placed at an MLC leaf junction (default position) the
result may exaggerate localized dose errors caused by the tongue‐and‐groove effect
•Solution: offset the array by 0.25 cm to avoid the tongue‐and‐groove effect
•Diodes degrade as they accumulate dose; radiation hardening helps, but it is necessary
to regularly check and update the array calibration annually or as needed
•This 2‐D device uses ionisation chambers, which has certain advantages compared to
the use of diodes such as lack of detector ageing effects
•Each chamber volume is 0.08 cm3 , a height of 5 mm and a diameter of 4.5 mm
•The spatial resolution of the detector system is 7.5 mm
•By shifting the array 3 times the number of measurement points can be increased to
2916
•As discussed in the presentation about delivery techniques for IMRT(Presentation # 8)
volumetric‐modulated arc radiotherapy is an IMRT technique where an optimal 3‐D dose
distribution is delivered during continuous gantry rotation
•The verification methods used for conventional iMRT delivery techniques may not work
for VMAT and therefore a number of new phantom‐detector systems have been
developed

27
•The MatriXX 2‐D detector can be inserted in the Multicube phantom and dose in
coronal and sagittal planes can be measured depending on the phantom setup
•The detector‐phantom combination has an angular dependence of its response with a
maximum error of 8% at 90 degrees
•By measuring its angular response, a correction factor as a function of gantry angle can
be applied
•Due to its rotational option thousands of measurement points are obtained over the
entire phantom volume
•The detector is always aligned perpendicular to the beam
•The software allows 3‐D volume analysis and patient‐specific CT overlay
•The largest effective measurement field is 27 cm x 27 cm
•A cylinder‐shaped PMMA phantom with 2 imbedded orthogonal crossing detector
planes
•1069 diode detectors are positioned in two planes
•The dose is recorded in 2 planes and a 3‐D dose distribution is reconstructed for
comparison with the calculated plan
•The behaviour of these systems for verifying VMAT plans calculated with two
treatment planning systems was evaluated
•A number of characteristics, such as the angular dependence (MapCHECK) and
atomic composition (Delta4), were studied in more detail
•This CIRS thorax phantom has lungs with an electron density (E.D.) relative to water of
0.207
•Bone, muscle and adipose inserts can be used with E.D values of 1.506, 1.042 and
0.946, respectively
•A discussion about these dose reconstruction models and various other aspects of EPID
dosimetry can be found in: van Elmpt, W., L. McDermott, S. Nijsten, M. Wendling, P.
Lambin, and B. Mijnheer. A literature review of electronic portal imaging for
radiotherapy dosimetry. Radiother. Oncol. 88: 289‐309, 2008.
•This slide illustrates the procedure for in vivo dosimetry using EPID dosimetry in
combination with a back‐projection algorithm
•The patient plan is directly compared with the reconstructed dose distribution using
the patient CT data
•The same method can be applied for pre‐treatment verification when using a phantom
instead of the patient data
•For per‐treatment verification it is, however, necessary to make a new plan based on
the phantom CT data
•This slide shows an example of a patient record of an in vivo dose verification as
applied routinely in The Netherlands Cancer Institute of a five‐field prostate IMRT
treatment
•A number of clinically relevant errors have been discovered by means of in vivo EPID
dosimetry in The Netherlands Cancer Institute
•In this example the deviation of the isocenter dose was outside the clinical criterion of
5%
•Other examples of serious errors discovered by means of EPID in vivo dosimetry during
IMRT verification has been discussed by: A. Mans, M. Wendling, L.N. McDermott, J.‐J.
Sonke, B. Mijnheer, M. van Herk and J.C. Stroom. Catching errors with in vivo dosimetry.
Med. Phys. 37: 2638‐2644, 2010.
•The reason for the observed difference between delivered and planned dose
distribution was that at the time of treatment the filled air cavity when making the
planning CT scan was emptied, causing a 6% higher dose at the isocentre.
•Independent MU/dose calculations are performed by means of simple models using
tabulated values obtained during the commissioning process of an accelerator
•The traditional method, based on tables and interpolation rather than physical
modelling

43
•In this ESTRO Booklet No. 6 some practical examples are given how to apply the ESTRO
formalism for the calculation of monitor units in high energy photon beams
•The basis for that procedure is the determination of the absorbed dose per monitor
unit under reference conditions: 10 cm depth, SDD = 100 cm and a 10 cm x 10 cm field
size
•The traditional MU calculation using dosimetric quantities referring to dose maximum
have been replaced by a formalism which applies quantities referring to measurements
at 10 cm depth for all beam qualities
•The reason for this change is that the maximum dose depends on the degree of
electron contamination that varies critically with change in beam geometry
•In this AAPM report recommendations are presented on how to perform
verification of monitor unit calculations in a modern clinic
•Furthermore, recommendations are provided on establishing action levels for
agreement between primary calculations and verification, and to provide
guidance in addressing discrepancies outside action levels
•Performing independent MU/dose calculations for IMRT is no longer possible by using
look‐up tables
•The traditional method, based on tables and interpolation rather than physical
modelling

46
•Independent MU/dose calculations for IMRT require a sophisticated dose calculation
mode
•Commercial products for IMRT verification are nowadays available
•These products vary in sophistication in their dose calculation, in the comparison with
the primary calculation, and
in the presentation of the results
•These are companies providing software for independent MU/dose calculation for
IMRT
•This figure, taken from ESTRO Booklet No. 9, shows the difference between the patient
dose calculations performed with the clinically applied treatment planning system
(Eclipse) and the independent dose calculation software (RadCalc) for the verification of
single field IMRT head‐and‐neck treatments (blue data)
•The data in purple show the differences when the RadCalc data were compared with
phantom dose calculation performed with Eclipse and the green data are the differences
with ionisation chamber measurements
•The similarity between the blue, purple and green data indicate that the patient
anatomy, including tissue heterogeneities, have only a small effect on the accuracy of
the independent dose calculation
•This slide shows schematically how independent MU/dose calculations (MapCALC) can
be combined with dose calculations performed by the treatment planning system (TPS),
using comparison software available from a measurement device (MapCHECK)
•This is an example of another commercial product for independent MU/dose
calculations for IMRT showing its 3‐D dose calculation possibilities and the way the
comparison results are presented
•In order to have a meaningful comparison with dose calculations performed with the
clinically applied TPS, independent MU/dose calculation software should fulfill several
requirements.
• This booklet presents in detail beam fluence modelling of linacs and dose
distributions in homogeneous slab geometry, as well as methods to analyse
observed deviations found by an independent dose calculation

•It describes analytical models for independent dose calculation of virtually any
beam configuration with very small calculation uncertainty, together with
detailed description of the error propagation

•The booklet also describes methods of applying independent dose calculations

•Points in high dose region of plan, but not necessarily for each individual beam.

•From: Georg et al.: Patient‐specific IMRT verification using independent fluence‐based