Design and characterisation of the Octa dosimeter

Towards an ideal dosimeter

for

small x-ray beams in modern radiation therapy

Background

Small x-ray beams in modern radiation therapy

Radiation therapy aims to optimise the conformity of lethal dose to the tumour target while limiting as much as possible radiation-induced side effects to surrounding healthy tissue. Consequently, there has been an escalation in the type of treatments that use a combination of small, irregularly-shaped high-dose beams of megavoltage (MV) X-rays (see, for example, What is Stereotactic Radiotherapy?).

The treatment unit (a medical linear accelerator, ‘linac’) and the software for dose planning (the treatment planning system) are commissioned and routinely tested for the accuracy of dose calculation as compared to the measurements, following a consensus procedure. For instance, see commissioning [1] and quality assurance [2,3] of linac, and beam modelling in a planning system [4]. Measurements are typically performed by scanning a 0.6 cm³-type ionization chamber in a water tank (Figure 1), and include percentage central axis dose (PDD), tissue maximum ratios (TMR), off-axis-ratios (OAR) and the linac’s output factor (OF).

However, if the X-ray field is small (less than 3 cm in diameter), the sensitive volume of a 0.6 cm³-type ionisation chamber is too large, resulting in measurement artefacts (Figure 2; Figure 3) [5–7]. Ensuring artefacts are negligible is challenging because the x-ray source is partially occluded (Figure 4.), and dose distributions are characterized by a lack of charged-particle equilibrium.

During the early period when small-field radiotherapy was first introduced, physicists commissioned and tested linacs and treatment planning systems with much the same procedure and dosimeters used for broad fields. The inaccuracies of small field dosimetry were poorly appreciated, and there was no clear criteria for how a small-field dosimeter should be designed.

---

It was not until 2017 after

• IAEA TRS-483 guideline, Dosimetry of Small Static Fields Used in External Beam Radiotherapy [7], and
• ICRU Report 91, Prescribing, Recording and Reporting of Stereotactic Treatments with Small Photon beams [6]

were published, that the equipment, techniques and requirements were standardized.

Dosimeters for small X-ray fields now have rigorous requirements (Table 1), and it is good practice to assume that a dosimeter suitable for measurements in broad fields will not be appropriate for measurements in small fields.

At the time this work began in 2016, there were only a few small-field dosimeters commercially available. None satisfied ICRU 91 requirements [6] and a measurement to dose correction factor had to be applied to readings. It was an accepted practice to take measurements with at least two types of dosimeters – one with a correction factor above and the other below unity. The method provided a means of cross-checking the consistency of results [11], as recommended by the ICRU [6]. Examples of such dosimeters were micro-ionization chambers, radiochromic films and micro-diamonds [7].

Property	Guideline	Notes
Stability (aka linearity with accumulated dose)	better than 0.1% for a total accumulated dose of many hundreds of kGy, from multiple exposures	corrections can be made provided the effect is consistent and recalibration is not frequently required
Linearity with dose	better than 0.1% over a dose range of at least 3 orders of magnitude
Linearity with instantaneous dose rate (aka dose per pulse linearity)	better than 0.1% over the range of interest, typically from 0.2 mGy to 2.0 mGy per pulse
Energy dependence	minimized in the energy range of interest, typically 60Co to 10 MV	ideally, energy independent with interaction coefficients ( uen/⍴ for photons, S/⍴ for electrons) having a constant ratio to those of water
Spatial resolution	trade-off between a high signal-to-noise ratio and a small sensitive volume	requirement is set by the dose gradients in the radiation field of interest
Size of the sensitive volume	correction for volume averaging better than 5%
Directional dependence (aka angular dependence)	better than 0.5% for angles <60° between the beam axis and the detector axis	corrections can be made to minimize the effect, or beam incidence can be kept fixed
Background signal (aka leakage signal)	at least 3 orders of magnitude lower than the detector response
Environmental factors (temperature, …)	corrections can reduce any influence to better than 0.3% over the full range of working conditions

Based on these shortcomings when measuring dose for small-field MV X-ray beams, my research work aimed to design and characterize an ideal prototype dosimeter and was named ‘the Octa’.

The Octa

Concept & design

The Octa is a 2D silicon dosimeter. It has a total sensitive area 38.7 mm × 38.7 mm and is covered by a thin layer of epoxy resin (Figure 5.) to provide a tissue-equivalent protection against moisture and accidental damage. There are 512 sensitive elements (diodes), arranged along 4 intersecting orthogonal linear arrays and oriented at 45 degrees with respect to each other. All the sensitive elements have the same 0.032 mm² area and are elongated rectangular shape (0.04 mm × 0.80 mm) – except for the 9 sensitive elements at the intersection of the four arrays, which are of 0.16 mm × 0.20 mm. The pitch between the sensitive elements is sub-millimetre, 0.3 mm along the vertical and horizontal arrays and 0.43 mm along the diagonals.

This specially designed layout has unique features for small-field dosimetry. It has sub-millimetre spatial resolution for when simultaneously measuring the linac’s OF, and cross-plane, in-plane and the two diagonal OAR.

Unfortunately, silicon dosimeters have a field-size dependence in small fields due to perturbation of the particle fluence as the field size is reduced. To convert readings to dose, a  correction factor is derived from Monte Carlo simulations or from experimental cross-checks with other dosimeters [6].

Ideally, a ‘correction-free’ dosimeter (or one requiring a consistent, close to unity correction) is needed. Charles et al suggested that this may be possible by adding a low-density medium to the high-density silicon [12]. The Octa achieved this by introducing a small air gap on top of the sensitive diode elements [13] (Figure 6.). A more complete description of the data acquisition system (hardware and software) can be found in my thesis [14].

During my PhD thesis work, I have carried out megavoltage photon beam performance tests of the Octa on a range of linear accelerators. They were:

• Varian Clinac iX® [14],
• Varian TrueBeam STx™ [15], and
• Elekta Axesse™ [16].

The performance results of the Octa to measure the physical characteristics of an Accuray CyberKnife® system [17], are described in this article.

Case study

The CyberKnife accelerator

The CyberKnife (Figure 7) is a machine dedicated to providing small-field stereotactic megavoltage X-ray treatments down to 5 mm in diameter and with sub-millimetre positional accuracy [18,19].  The linear accelerator is mounted on a robotic arm capable of directing the X-ray beam 3-dimensionally. There’s no flattening filter and the beam is shaped using either fixed circular cones or the variable aperture Iris™ collimator system (Figure 8) [18,20].

For CyberKnife quality assurance tests, the high spatial resolution Gafchromic film was the preferred dosimeter [21,22]. But the use of Gafchromic film has disadvantages. The time taken after film exposure to develop and analyse it, is an inconvenient delay. Further, there can be large uncertainties between film batches due to:

• film density measurements affected by variations in chemical sensitivity;
• film polarization;
• non-uniformity across the film;
• errors during scanning detector measurements; and
• variations in film handling and developing techniques [11].

The Octa diode dosimeter I developed at Wollongong University was thoroughly tested in small X-ray beams prior to the CyberKnife tests [15,16]. The Octa specification tests showed it had:

• a stable and linear response with absorbed dose;
• near megavoltage photon energy independence (i.e. the signal to dose ratio varying with depth in water was constant),
• high sensitivity; and
• the small sensitive diode elements provided high spatial resolution.

The initial Wollongong University specification tests indicated that the Octa could potentially offer comparable performance to Gafchromic film. However, the Octa had the decided advantage to film of being able to obtain real-time, more reliable results with reduced uncertainties.

To verify this hypothesis, Octa measurements of the CyberKnife M6 at the Sir Charles Gairdner Hospital (Nedlands, WA) were compared against a PTW SRS diode 60018 and Monte Carlo calculations. I performed the Monte Carlo calculations with Geant4 (GEometry ANd Tracking 4) [23], a tool-kit for simulating the passage of particles through matter [24,25].   

The experimental setup is shown in Figure 9. The Octa was placed on the treatment couch with:

• 10 cm of ‘solid water’ slabs (i.e. solid material with equivalent radiation absorption characteristics to water) for adequate X-ray beam backscatter; and
• additional solid water slabs placed on top of the Octa for the required water-equivalent depth for each measurement.

Octa measurements were obtained for OF, OAR, PDD and TMR.

Results

Figure 10(a) shows the comparative measurements of the Octa and SRS diode for fixed cone OF and Figure 10(b) shows similar comparisons plus Monte Carlo dose calculations for the Iris collimator. The percentage difference is shown in the lower panels of Figure 10. The Monte Carlo OF calculation described the dose deposited in a voxel of solid water whose dimensions were those of the Octa’s central element. Uncertainty bars, calculated as 3 standard deviations, did not exceed the symbol size for both measurements and calculations.

OAR measurements with the Octa and the SRS diode are shown in Figure 11 for Iris collimated fields. OAR is aligned such that the origin lies at the coordinate corresponding to the 50% response. Uncertainty bars of 3 standard deviations did not exceed the symbol size.

Figure 12 shows the PDD measured with:

• the Octa in solid water;
• the SRS diode in water; and
• Monte Carlo calculations assuming solid water.

The X-ray field was a 60 mm diameter and collimated with the Iris.

Figure 13 shows the TMR measured with the Octa in solid water, the SRS diode in water and Monte Carlo calculations in solid water. The X-ray fields were of 5 and 60mm diameter and collimated with the Iris.

In each Figure, the Octa percentage differences, with respect to the benchmark dose measurements, are shown in the lower panels. The 3 standard deviation uncertainty bars did not exceed the symbol size for the measurements or calculations.

Conclusion

The use of the Octa during this CyberKnife study demonstrated:

• Data can be simultaneously acquired of OAR at 0°, 45°, 90° and 135°, and those at 15° and 105° upon rotation (Figure 8., Figure 9.);
• The measurement time to complete vendor quality assurance protocols can be significantly reduced [20].
• There can be a more robust implementation of the requirements when including OAR along directions not currently considered.
• OF and OAR of all Iris fields investigated can be measured using the Octa in less than 10 minutes;
• Each PDD and TMR was measured in approximately 25 minutes.
• PDD and TMR measured with the Octa were accurate to within 3% with respect to both the SRS diode and Monte Carlo calculations for all fields investigated.

Summary

In summary, the Octa was shown to be:

• A ‘correction-free’ dosimeter suitable for routine quality assurance of the CyberKnife;
• A reliable real-time read-out device; and
• Unique for dosimetry applications, such as evaluating long-term stability of the Iris collimator when used in dynamic mode.

References

[1]      Das I J, Cheng C-W, Watts R J, Ahnesjö A, Gibbons J, Li X A, Lowenstein J, Mitra R K, Simon W E and Zhu T C 2008 Accelerator beam data commissioning equipment and procedures: Report of the TG-106 of the Therapy Physics Committee of the AAPM Med. Phys. 35 4186–215

[2]      Klein E E, Hanley J, Bayouth J, Yin F-F, Simon W, Dresser S, Serago C, Aguirre F, Ma L, Arjomandy B, Liu C, Sandin C and Holmes T 2009 Task Group 142 report: Quality assurance of medical accelerators Med. Phys. 36 4197–212

[3]      Smith K, Balter P, Duhon J, White G A, Vassy D L, Miller R A, Serago C F and Fairobent L A 2017 AAPM Medical Physics Practice Guideline 8.a.: Linear accelerator performance tests J. Appl. Clin. Med. Phys. 18 23–39

[4]      Ma C M C, Chetty I J, Deng J, Faddegon B, Jiang S B, Li J, Seuntjens J, Siebers J V. and Traneus E 2020 Beam modeling and beam model commissioning for Monte Carlo dose calculation-based radiation therapy treatment planning: Report of AAPM Task Group 157 Med. Phys. 47 e1–18

[5]      Pappas E, Maris T G, Zacharopoulou F, Papadakis A, Manolopoulos S, Green S and Wojnecki C 2008 Small SRS photon field profile dosimetry performed using a PinPoint air ion chamber, a diamond detector, a novel silicon-diode array (DOSI), and polymer gel dosimetry. Analysis and intercomparison Med. Phys. 35 4640–8

[6]      ICRU 2017 Report 91 J. Int. Comm. Radiat. Units Meas. 14 1–160

[7]      IAEA 2017 Technical Reports Series No. 483 – Dosimetry of Small Static Fields Used in External Beam Radiotherapy An International Code of Practice for Reference and Relative Dose Determination

[8]      Wuerfel J U 2013 Dose measurements in small fields Med. Phys. Int. 1 81–90

[9]      Sánchez-Doblado F, Hartmann G H, Pena J, Roselló J V, Russiello G and Gonzalez-Castaño D M 2007 A new method for output factor determination in MLC shaped narrow beams Phys. Medica 23 58–66

[10]    IPEM 2010 Small Field MV Photon Dosimetry

[11]    Tyler M, Liu P Z Y, Chan K W, Ralston A, McKenzie D R, Downes S and Suchowerska N 2013 Characterization of small-field stereotactic radiosurgery beams with modern detectors Phys. Med. Biol. 58 7595–608

[12]    Charles P H, Crowe S B, Kairn T, Knight R T, Hill B, Kenny J, Langton C M and Trapp J V. 2013 Monte Carlo-based diode design for correction-less small field dosimetry Phys. Med. Biol. 58 4501–12

[13]    Utitsarn K, Alrowaili Z A, Stansook N, Biasi G, Carolan M, Lerch M and Rosenfeld A 2019 The effect of an air gap on a 2D monolithic silicon detector for relative dosimetry J. Instrum. 14 P06018–P06018

[14]    Biasi G 2019 Design of a solid-state array detector prototype for small-field dosimetry (University of Wollongong)

[15]    Biasi G, Petasecca M, Guatelli S, Hardcastle N, Carolan M, Perevertaylo V, Kron T and Rosenfeld A B 2018 A novel high-resolution 2D silicon array detector for small field dosimetry with FFF photon beams Phys. Medica 45 117–26

[16]    Biasi G, Al Shukaili K, Petasecca M, Corde S, Guatelli S, Perevertaylo V L, Kron T and Rosenfeld A B 2019 Today’s monolithic silicon array detector for small field dosimetry: the Octa Journal of Physics: Conference Series vol 1154 p 012002

[17]    Biasi G, Petasecca M, Guatelli S, Martin E A, Grogan G, Hug B, Lane J, Perevertaylo V, Kron T and Rosenfeld A B 2018 CyberKnife ® fixed cone and Iris^TM defined small radiation fields: Assessment with a high-resolution solid-state detector array J. Appl. Clin. Med. Phys. 19 547–57

[18]    Kilby W, Dooley J R, Kuduvalli G, Sayeh S and Maurer C R 2010 The CyberKnife® Robotic Radiosurgery System in 2010 Technol. Cancer Res. Treat. 9 433–52

[19]    Taylor M L, Kron T and Franich R D 2011 A contemporary review of stereotactic radiotherapy: inherent dosimetric complexities and the potential for detriment Acta Oncol. (Madr). 50 483–508

[20]    Echner G G, Kilby W, Lee M, Earnst E, Sayeh S, Schlaefer A, Rhein B, Dooley J R, Lang C, Blanck O, Lessard E, Maurer C R and Schlegel W 2009 The design, physical properties and clinical utility of an iris collimator for robotic radiosurgery Phys. Med. Biol. 54 5359–80

[21]    Dieterich S and Sherouse G W 2011 Experimental comparison of seven commercial dosimetry diodes for measurement of stereotactic radiosurgery cone factors Med. Phys. 38 4166–73

[22]    Blanck O, Masi L, Damme M C, Hildebrandt G, Dunst J, Siebert F A, Poppinga D and Poppe B 2015 Film-based delivery quality assurance for robotic radiosurgery: Commissioning and validation Phys. Medica 31 476–83

[23]    Allison J, Amako K, Apostolakis J, Arce P, Asai M, Aso T, Bagli E, Bagulya A, Banerjee S, Barrand G, Beck B R, Bogdanov A G, Brandt D, Brown J M C, Burkhardt H, Canal P, Cano-Ott D, Chauvie S, Cho K, Cirrone G A P, Cooperman G, Cortés-Giraldo M A, Cosmo G, Cuttone G, Depaola G, Desorgher L, Dong X, Dotti A, Elvira V D, Folger G, Francis Z, Galoyan A, Garnier L, Gayer M, Genser K L, Grichine V M, Guatelli S, Guèye P, Gumplinger P, Howard A S, Hřivnáčová I, Hwang S, Incerti S, Ivanchenko A, Ivanchenko V N, Jones F W, Jun S Y, Kaitaniemi P, Karakatsanis N, Karamitrosi M, Kelsey M, Kimura A, Koi T, Kurashige H, Lechner A, Lee S B, Longo F, Maire M, Mancusi D, Mantero A, Mendoza E, Morgan B, Murakami K, Nikitina T, Pandola L, Paprocki P, Perl J, Petrović I, Pia M G, Pokorski W, Quesada J M, Raine M, Reis M A, Ribon A, Ristić Fira A, Romano F, Russo G, Santin G, Sasaki T, Sawkey D, Shin J I, Strakovsky I I, Taborda A, Tanaka S, Tomé B, Toshito T, Tran H N, Truscott P R, Urban L, Uzhinsky V, Verbeke J M, Verderi M, Wendt B L, Wenzel H, Wright D H, Wright D M, Yamashita T, Yarba J, et al 2016 Recent developments in GEANT4 Nucl. Instruments Methods Phys. Res. Sect. A Accel. Spectrometers, Detect. Assoc. Equip. 835 186–225

[24]    Larsson S, Svensson R, Gudowska I, Ivanchenko V N and Brahme A 2005 Radiation transport calculations for 50 MV photon therapy beam using the Monte Carlo code GEANT4 Radiat. Prot. Dosimetry 115 503–7

[25]    Faddegon B A, Asai M, Perl J, Ross C, Sempau J, Tinslay J and Salvat F 2008 Benchmarking of Monte Carlo simulation of bremsstrahlung from thick targets at radiotherapy energies Med. Phys. 35 4308–17

G Biasi 25 October 2021