| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Medical Physics, University Hospital of North Staffordshire, Princes Road, Hartshill, Stoke-on-Trent, Staffordshire ST4 7LN, UK
Correspondence: Craig Edwards, Medical Physics Department, University Hospital of North Staffordshire, Princes Road, Hartshill, Stoke-on-Trent, Staffordshire ST4 7LN, UK. E-mail: craig.edwards{at}uhns.nhs.uk
 |
Abstract |
|---|
 Top Abstract References |
|---|
Recommendations for the adoption and implementation of in vivo dosimetry (IVD) have been made in several recent reports as a means of checking the dose and providing an early identification of an incorrect dose administration [1–3]. However, its clinical role, value and cost effectiveness as a routine procedure for most treatments have generated comment in this journal [4–8] and have provided the topic for structured debate at recent national conferences (e.g. the IPEM Biennial Radiotherapy Physics Meeting, Bath, 2–4 September 2008, and the UK Radiation Oncology Conference, Cardiff, 6–8 April 2009). These deliberations have tended to concentrate on the justification for implementing IVD as a routine practice and on the resources required. However, less attention has been paid to the physical characteristics and limitations of the technique. These features will be governed largely by the properties of the detector used for the IVD system.
A recent survey found that diodes and LiF thermoluminescent dosemeters (TLDs) were the most commonly used detectors, and that the number of radiotherapy centres that routinely measured critical organ dose outside the main treatment field was about twice the number that measured the central axis dose [9]. This latter finding was not consistent with the Manual of Cancer Service Standards [10] and with recommendations from other sources [1, 11, 12] that concentrate on tumour (i.e. central axis) IVD. Therefore, although a considerable amount of information has been published on the properties of diodes and LiF TLDs, an evaluation of the effect of these properties on the accuracy and reliability of IVD must include an examination of their response outside the main treatment field, as well as on the central axis.
The response of both types of detector will be dependent to varying degrees upon factors such as ambient temperature, angle of incidence, dose rate, accumulated radiation damage and X-ray energy. As the usual practice is to calibrate the detector on the central axis of a treatment beam, errors in central axis IVD due to a non-uniform energy and incident angle response will be virtually eliminated, and the residual uncertainty for diode measurements (about ±4% for 6 MV and 10 MV X-rays) will be dominated by ambient temperature effects (+0.3% per °C) [13]. It has been pointed out that consideration should be given to skewing the action level positively [9] because the diode is most likely to be calibrated at room temperature but used to record dose on the patient's skin at a temperature of typically 32°C. The overall uncertainty can be minimised by the adoption of a suitably robust calibration regime, and good agreement has been obtained between the response of a diode and an ionisation chamber when used to record entrance and exit dose measurements [14–16]. Studies investigating the accuracy of critical organ dose measurements have evaluated IVD performed during treatment by total body irradiation [17–19], and differences between the expected dose and that recorded by the IVD only exceeded ±5% when insufficient build-up material was used with TLD detectors [17].
However, when critical organ dose measurements are made outside the treatment field, much greater errors will be introduced by the non-uniform energy response of diodes and LiF TLDs [20], as there will be a difference between the X-ray spectrum at measurement positions outside the main beam and the spectrum at the calibration position on the central axis. Compared with their response on the central axis of a 6 MV X-ray beam, these spectral differences have been found to produce an overestimate of X-ray dose by a diode of between 10% and 70% at distances of 1–10 cm from the edge of a field ranging in size from 4 x 4 cm2 to 15 x 15 cm2 [21]. The ratio of the response of a LiF TLD at the two positions for the same range of distances and field sizes was about unity, indicating that more reliable measurements of X-ray dose outside the main field will be produced by a LiF TLD than by a diode [21].
The previous discussion relates only to the near surface dose from photons, and it ignores the contribution to the overall dose from contaminating charged particles (electrons and positrons) present in the X-ray beam. These charged particles are generated mainly by interactions between the primary X-ray beam and the components in the head of the linear accelerator (e.g. the flattening filter) [22]. Their intensity and energy spectrum will be dependent on the X-ray energy and on the design and type of materials used in the head. Hence the relative contribution of all three types of radiation (photons, electrons, positrons) to the surface dose and to the IVD detector response is different inside and outside the treatment field.
It has been shown that, for a 6 MV X-ray beam from a linear accelerator, there is negligible contribution to the surface dose from positrons inside and outside the main beam for field sizes from 10 x 10 cm2 to 40 x 40 cm2 [23]. Strict numerical comparisons between different publications reporting the contribution of contaminating electrons to the total surface dose should be treated with caution because of the differences in accelerator type, head design, field size, beam energy and method of dose estimation (i.e. by calculation and the type of Monte Carlo code, or by a measurement that usually relied on magnetic separation). However, this contribution to the total surface dose and to the response of a diode when used for IVD cannot be ignored [23, 24]. For a 6 MV X-ray beam, the contaminating electron contribution to the total diode response (without a build-up cap) at the surface on the central axis has been estimated by Monte Carlo calculations to be 
7–8% for field sizes ranging from 4 x 4 cm2 to 15 x 15 cm2 [24]. However, the mean energy of the contaminating electrons at the surface on the central axis is less than 1 MeV for a 6 MV X-ray beam [24], and therefore it will have a short penetration in tissue (a practical range of 5 mm). Hence the contribution to a diode measurement with a build-up cap on the central axis will be even less and can probably be ignored, thereby allowing a dose at greater depths to be derived from the measurement.
Outside the main field, the contribution of contaminating electrons to the total diode response has been estimated to increase with distance from the field edge and, at any given position, it increased with a decrease in field size, e.g. giving a value of 58% at 10 cm from the edge of a 4 x 4 cm2 6 MV X-ray beam [24]. Although the mean energy of the contaminating electrons on the surface outside the main field is also less than 1 MeV for a 6 MV X-ray beam, this much larger relative contribution to a diode response coupled with the low mean energy of the photon spectrum at these positions (typically 0.3 MeV depending on field size and distance from outside the field edge [21]) means that it cannot be ignored for the interpretation of IVD outside the field even with a build-up cap, and that IVD outside a field with a diode has to be interpreted with caution. Moreover, derivation of a dose at depth would require data on the electron and X-ray depth–dose variations specific to that position (i.e. specific to the incident energy spectrum for both types of radiation).
As previously highlighted [9], an overestimation of dose outside the main field may afford an element of protection to a critical organ, but could result in subsequent confusion over the correctness of recorded doses. The need for detailed guidance on, as well as adequate resources for, the treatment circumstances and method for implementation of IVD has also been identified as a factor that should help to satisfy recommendations for the technique to become routine. The limitations discussed above for estimating the dose to critical organs outside the main beam are based on very little data. There is a clear need for further studies on the effects of X-ray energy and electron contamination on both diode and LiF TLD responses outside the main field before recommendations can be made on (i) values of the relative X-ray and electron contribution to detector response and dose, and (ii) a matrix of position-dependent detector sensitivity correction factors that will also be dependent on other aspects, such as field size and beam energy.
Moreover, the effects of the X-ray energy response and contaminating electrons on diode measurements of the dose profile and of the in vivo surface dose for dynamically wedged, conformal and intensity-modulated X-ray beams need to be explored. The recent development of a multiwire ionisation chamber, such as DAVID (PTW-Freiburg, Germany) and COMPASS (IBA Dosimetry Gmbh, Schwarzenbruck, Germany), inserted into the accessory tray of an accelerator may eliminate these uncertainties in fluence maps. However, it should be noted that in vivo dose verification by such a chamber will only determine the fluence map of a beam and will neglect the patient-specific contribution to the surface dose distribution inside and outside the main field, which can only be included by an in vivo diode or TLD measurement.
Received for publication March 30, 2009. Revision received June 4, 2009. Accepted for publication June 9, 2009.
|
References |
|---|
|
TopAbstractReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| BJR | DMFR | IMAGING | ALL BIR JOURNALS |
