«CALIBRATION OF P-TYPE SILICON DIODES FOR IN-VIVO DOSIMETRY IN 6 0 Co BEAMS Iva Mrčela, Tomislav Bokulić, Mirjana Budanec and Zvonko Kusić ...»
VI. simpozij HDZZ, Stubičke Toplice, L.. HR0500068
CALIBRATION OF P-TYPE SILICON DIODES FOR
IN-VIVO DOSIMETRY IN 6 0 Co BEAMS
Iva Mrčela, Tomislav Bokulić, Mirjana Budanec and Zvonko Kusić
Department of Oncology and Nuclear Medicine,
University Hospital "Sestre milosrdnice",
Vinogradska c. 29, HR-10000 Zagreb
Radiation treatment accuracy is expressed as a comparison between prescribed and delivered dose. Several studies suggest 3.5%, 1SD [1-2] as the overall accuracy required and achievable in radiation treatment, based on radiobiological studies and measurements by in vivo dosimetry in clinical conditions. Semiconductor diodes as detectors for in vivo dosimetry are considered as very useful tool in clinical practice. Their main advantage over other detectors, such as TLDs, is a possibility of immediate readout and detection of errors while patient is still on a treatment couch. Moreover, diodes are known for their high sensitivity, small size, simplicity of operation and mechanical stability. However, for accurate dosimetry, diodes have to be individually characterised for conditions other than referent.
In this work we present first results in implementation of in-vivo dosimetry in our department by calibration and characterisation of diodes designed for use in Co beams. It is known from the literature [1,3] that ideal diode should have small dependence, of about 1-2 %, on field size, source to skin distance (SSD) and use of beam modifying devices. These correction factors originate from the dependence of diode response on beam energy, dose per pulse, dose rate, temperature and direction of beam.
MATERIALS AND METHODSThree Scanditronix EDE-5 p-type silicon diodes connected to a DPD-3 electrometer were calibrated for measuring entrance dose. EDE-5 has an effective thickness of measuring volume of 60 um and 1.5 mm detector diameter.
Hemispherical build up cap consists of polystyrene and epoxy plastic and it is equivalent to 5 mm of water, which is the depth of maximum dose for 60 Co. Diodes were preirradiated with 10 MeV electrons to 8 kGy by manufacturer. Technical specifications state 1% signal deviation for changes in field sizes from 5x5 to 30x30 cm2 and 0.4% per °C sensitivity variation with temperature.
A Farmer ionisation chamber 0.6 cm3 PTW type 30002 connected to the PTW Unidos electrometer was used as a reference detector for calibration. All VI. simpozij HDZZ, Stubičke Toplice, 2005.
measurements were performed on a polystyrene phantom PTW type RW3 with special slab adapted in depth and dimensions of opening for this chamber.
Dimensions of slices used are 30x30 cm and 1 cm thickness. Prior to calibration, intrinsic precision of diodes was measured as signal reproducibility of ten consecutive irradiations at same dose when diode is placed on top of the phantom.
In addition, to confirm linear dose response we tested our diodes in dose interval from 0.5 Gy to 8 Gy.
The calibration of diodes for entrance dose measurements and evaluation of various correction factors were performed according to the procedure recommended by ESTRO [4-5] and Leunens . Entrance dose is defined as the
dose at the d max from the incident plane on the beam axis:
(1) Here, Rdiode is a diode reading, Fcal is calibration factor and CFi are correction factors.
Calibration factors were determined for each diode as the ratio of dose measured by ionisation chamber placed at dmax in plastic phantom and signal from diode placed at phantom surface, at standard reference conditions (10x10 cm2 field size at isocenter, SSD=80 cm, gantry angle 0°).
(2) J R diode Jref.con.
Calibration has to be repeated on regular basis, because diode sensitivity changes with accumulated dose. Some authors advise recalibration after accumulated dose of ikGy for p-type diodes .
Correction factors (CFs) for different field sizes, SSDs, wedges, tray and gantry angles were determined. Field size (denoted by FS = a) and SSD (r) correction factors were measured as a ratio of chamber and diode reading in given
condition normalised to the reference conditions :
Measurements for ten wedges available for our 60 Co unit were normalised to appropriate open fields (open FS) to avoid double inclusion of field size correction, where a" is a wedge angle and a is a wedge length. The same approach was for tray
For evaluation of directional correction, we placed diodes in the field centre at ref.
conditions and measured response for different gantry angles (GA = 8). CFs are given as ratio of diode signal at particular angle and at GA = 0°.
Finally, to confirm that applying all correction factors will give true entrance dose we have simulated clinical conditions on polystyrene phantom and compared diode readings with the expected dose calculated with treatment planning system.
RESULTSAll diodes showed acceptable intrinsic precision, less than recommended SD of 0.5% . Linearity of diode dose response was very good in dose interval that is typically used in patient treatment (0.5-8 Gy). Results for each diode together with calibration factors are given in Table 1.
Diode 1 Diode 2 Diode 3 Table 1.
Intrinsic precision - standard deviation 0.12% 0.10% 0.15% Linearity - correlation coefficient r2 0.999998 0.999998 0.999998 Calibration factor ± stand, deviation % 0.082±0.07% 0.0848±0.01 0.0845±0.01% Difference between week cal. factors 0.19% 0.03% 0.20% Diodes were regularly recalibrated every week during one month and they showed very small sensitivity variation. That can be explained with less than lkGy of accumulated dose in one week.
Variation in diode sensitivity with field size was very small, producing about 0.5% overestimation of dose, for 25 cm square field (Figure la). This is expected result for 60 Co beams and EDE diodes [7,8] in contrast to the high-energy photon beams where correction can be about 2% [1,3]. Due to the electron contamination of primary beam, originated from head scatter, diodes measure larger surface dose than dose at dmax, measured by chamber. That is the reason for decreasing the CFFS for larger collimator openings.
VI. simpozij HDZZ, Stubičke Toplice, 2005.
Changing the SSD from 70 to 100 cm increases the CF SD by about 1% S
(Figurelb). For smaller SSD there is a larger number of head scatter electrons that reaches the diode and the ratio of the chamber reading is decreasing .
Additionally, increase in SSD results in lowering the dose rate, which is another reason for underestimation of diode signal.
Figure 1. a) CFFs decreasing with increasing the collimator opening,
b) CFSSD increasing with SSD, D1-D3 represents three diodes used Effect of wedge filters on diode response is shown in Table 2. It ranges from about 1% for small wedges 15°, 30° for all three diodes and all field sizes, to 2.8 % in sensitivity variation for 60° wedge. Inserting the wedge in a beam decreases the dose rate and changes the beam quality. Therefore, diodes read smaller dose than expected, and CFs, are greater than one. The use of tray for supporting blocks in the beam alters the diode response by producing the electrons. This electron yield is greater for large collimator openings, causing about 0.7% overestimation of dose for 25 cm square field. We have measured correction factors for 0.5 cm thick PMMA tray with metal construction for block fixation, placed on 54.5 cm from the source. Tray CFs are given in Table 3.
It is known that diodes with hemispherical build up caps and ground plate have larger directional correction factors than cylindrical ones . Our results give largest correction of about 4.5% for 60° angles, which is even smaller than reported for EDE diodes .
In order to verify use of CF according to the relation (1), we have carried out set of phantom measurements simulating the different beam arrangements used in actual patient treatment. For example, beam set up with 8x8 cm2 field size, SSD=75 cm, gantry angle of 30°, with tray and 30°/8 cm wedge and dose specification of 100 cGy to the isocenter. Expected entrance dose was calculated with Theraplan Plus 1000 treatment planning system. Measured entrance doses were about 2.5% less than the expected, which is within the required accuracy.
CONCLUSIONThe aim of this work was to characterise new diodes intended for use in clinical radiotherapy, as a part of a quality assurance programme. We have evaluated stability, linearity calibration and correction factors. Results were within expected values for this type of diodes giving acceptable agreement in dose delivered and the expected dose. In future, we expect to investigate other parameters such as stability of correction factors with accumulated dose, temperature correction, calibration for exit dose measurements, midline dose calculations and finally, to carry out a patient studies for different treatment localisations.
REFERENCES Essers M, Mijnheer BJ. In vivo dosimetry during external photon beam radiotherapy.
Int J Radiat Oncol Biol Phys 1999; 43:245-259.
 Loncol T, Greffe JL, Vynckier S, Scalliet P. Entrance and exit dose measurement with semiconductors and TLD: a comparison of methods and in vivo results.
Radiother Oncol 1996; 41:179-187.
 Jornet N, Ribas M, Eudaldo T. In vivo dosimetry: Intercomparison between p-type and n-type based diodes for the 16-25 MV range. Med Phys 2000; 27:1287-1293.
 Van Dam J, Marinello G. Methods for in vivo dosimetry in external radiotherapy.
Physics for clinical radiotherapy. ESTRO Booklet No.l; 1994.
 Huyskens D. et al. Practical guidelines for the implementation of in vivo dosimetry with diodes in external radiotherapy with photon beams (entrance dose). Physics for clinical radiotherapy. ESTRO Booklet No.5; 2001.
 Leunens G, Van Dam J, Dutreix A, van der Schueren E. Quality assurance in radiotherapy by in vivo dosimetry. 1. Entrance dose measurements, a reliable procedure. Radiother Oncol 1990; 17:141-151.
 Nilsson B, Ruden B-I, Sorcini B. Characterization of silicon diodes as patient dosemeters in external radiation therapy. Radiother Oncol 1988; 11:279-288.
 Voordeckers M, Goosens H, Rutten J, Van den Bogaert W. The implementation of/« vivo dosimetry in a small radiotherapy department. Radiother Oncol 1998; 47: 45-48.
Acknowledgement: This work has been supported in part by International Atomic Energy Agency (Research Project Contract No. 13115).