Two‐dimensional solid‐state array detectors: A technique for in vivo dose verification in a variable effective area

Abstract Purpose We introduce a technique that employs a 2D detector in transmission mode (TM) to verify dose maps at a depth of dmax in Solid Water. TM measurements, when taken at a different surface‐to‐detector distance (SDD), allow for the area at dmax (in which the dose map is calculated) to be adjusted. Methods We considered the detector prototype “MP512” (an array of 512 diode‐sensitive volumes, 2 mm spatial resolution). Measurements in transmission mode were taken at SDDs in the range from 0.3 to 24 cm. Dose mode (DM) measurements were made at dmax in Solid Water. We considered radiation fields in the range from 2 × 2 cm2 to 10 × 10 cm2, produced by 6 MV flattened photon beams; we derived a relationship between DM and TM measurements as a function of SDD and field size. The relationship was used to calculate, from TM measurements at 4 and 24 cm SDD, dose maps at dmax in fields of 1 × 1 cm2 and 4 × 4 cm2, and in IMRT fields. Calculations were cross‐checked (gamma analysis) with the treatment planning system and with measurements (MP512, films, ionization chamber). Results In the square fields, calculations agreed with measurements to within ±2.36%. In the IMRT fields, using acceptance criteria of 3%/3 mm, 2%/2 mm, 1%/1 mm, calculations had respective gamma passing rates greater than 96.89%, 90.50%, 62.20% (for a 4 cm SSD); and greater than 97.22%, 93.80%, 59.00% (for a 24 cm SSD). Lower rates (1%/1 mm criterion) can be explained by submillimeter misalignments, dose averaging in calculations, noise artifacts in film dosimetry. Conclusions It is possible to perform TM measurements at the SSD which produces the best fit between the area at dmax in which the dose map is calculated and the size of the monitored target.


considers point-dose measurements
performed with an ionization chamber 4 and dose distribution measurements performed with an electronic portal imaging device (EPID), 5-7 a phantom-based electronic array [8][9][10][11] or films. However, time-consuming pretreatment QA is typically considered only once before the first treatment session; potential changes or errors in all sessions will remain unaddressed and/or undetected. 12,13 An in vivo verification approach validates, in real time, accuracy, and integrity of treatment plans; parameters monitored include, for instance, the output of a medical linear accelerator (linac) and the position and/or movement of the leaves of a multileaf collimator (MLC). [12][13][14][15] Solutions for in vivo monitoring include 16  Transit detectors such as EPIDs are placed so that the beam penetrates the patient first, and then the detector. [17][18][19] QA with transit EPIDs is challenging; their response is energy dependent and there is additional scatter from the patient; also, they are not able to discriminate between changes in signal due to changes in fluence incident on the patient from changes in signal due to anatomical variations within the patient. 20 Transmission detectors are, instead, placed between the linac head and the patient. Commercially available options include the Device for Advanced Verification of IMRT Delivery (DAVID) system (PTW, Freiburg, Germany), a flat, multiwire transmission-type ionization chamber 21,22 ; the Dolphin detector with the COMPASS software (IBA Dosimetry, Germany), which uses 1513 air-vented plane parallel ionization chambers, 23-25 the integral quality monitoring (IQM) system (iRT Systems GmbH, Koblenz, Germany), a large-area wedge ionization chamber 12,13,26,27 ; the Delta 4 Discover (ScandiDos AB, Uppsala, Sweden), a 2D solid-state array. 16 Several prototypes have also been proposed in the literature, including optical attenuation-based scintillating fibers 28 ; 2D solid-state arrays, such as the MP121 29,30 and the MP512. 31 Transmission detectors allow for independent monitoring of the output of a linac, and of the position and/or movement of the leaves of an MLC. 16 However, they have limitations. Any device placed in the beam path affects beam quality and introduces beam attenuation, 12 and as such has to be modeled in the treatment planning system (TPS). 27 Also, transmission detectors may increase surface dose 16,23 and their efficacy for beam monitoring is limited by their shape, active area, and spatial resolution.
The present study introduces a novel technique for using a 2D solid-state array prototype, the MP512 (512 diode-sensitive volumes, 2 mm spatial resolution). The MP512 was used in transmission mode (TM) to verify dose maps at a depth of d max in Solid Water. TM measurements were taken at different surface-to-detector distances (SDDs) in order to adjust the area at d max where the dose map is calculated.

2.B | The MP512 system
The MP512 is a prototype of a monolithic silicon-array detector; it was developed at the Centre for Medical Radiation Physics (University of Wollongong, NSW, Australia). The prototype has 512 diodesensitive volumes; these have an area of 0.5 × 0.5 mm 2 and are uniformly distributed with a pitch of 2 mm over an active area of 52 × 52 mm 2 . The MP512 is operated in passive mode (i.e., no external bias is applied); its associated readout electronics has a high temporal resolution (pulse-by-pulse signal acquisition). 32 In the literature, the MP512 has been characterized as a phantom-based detector for quality assurance in modern radiotherapy; it was demonstrated to be an accurate dosimeter for the measurement of output factors, percentage depth dose distributions, and lateraldose profiles; furthermore, its angular dependence was investigated and corrected for, making it a suitable candidate for quality assurance in arc deliveries. [33][34][35] The use of the MP512 as a transmission detector was also assessed. 31 In that study, it was reported that the MP512 in TM increases the surface dose by <25% for a SDD in the range from 0.3 to 18 cm, and by <5% for SDD >18 cm. 31 The transmission factor, at d max depth in Solid Water, 100 cm SSD, was in the range from 1.020 to 0.997 for SDDs from 0.3 to 24 cm. 31 2.C | Gafchromic™ EBT3 films and Farmer ionization chamber  were pre-and post-scanned (24 hrs after irradiation) six times maintaining a consistent orientation and using only the last three optical density maps. Films were calibrated using absolute dose measurements with the Farmer chamber. 36 Film analysis methodology was the same as that used by Aldosari et al. 37

2.D | Measurements in transmission mode and in dose mode
The MP512's active area was made light-tight using a black plastic sheet of thickness 80 µm. An equalization procedure, performed prior to all measurement, was used to address a nonuniformity in the integral response of the MP512's sensitive volumes. 38 Also, to convert readings to absolute dose, the MP512 was calibrated using measurements of response linearity with dose; those measurements were performed in jaws-defined fields of 10 × 10 cm 2 , at a depth of d max in Solid Water, 100 cm SSD. Delivered MUs were in the range from 1 to 1000 MU, at a fixed dose rate of 600 MU/min. The Farmer chamber was used for the absolute dose measurements at a depth of d max in Solid Water. 36 For TM measurements, the MP512 was sandwiched between two protective slabs of PMMA of thickness 3 mm. To minimize the resulting composite thickness, each slabs had an opening, centered on the axis of the MP512's active area, of 9.5 × 9.5 cm 2 ( Figure 1).
The MP512 was then lodged into a movable holder of PMMA; by moving the holder, the SDD could be varied in the range from 0.3 to 24 cm ( Figure 2). The effective area (A eff ), at a depth of d max in Solid Water, was defined as a function of SDD as:   fields, the response of the MP512 in DM at d max was then calculated using the relationship between DM and TM measurements derived as described in the previous section. Note that these fields were not part of those used to obtain the relationship in the first place. As the field of 1 × 1 cm 2 was smaller than the smallest field used for the fit, the calculated response in DM was extrapolated.
The Equivalent square fields (A eq ) of IMRT fields were calculated using 43 : As above, in each of these fields, the response of the MP512 in DM at d max was then calculated using the relationship between DM and TM measurements. Calculated dose distributions were compared with TPS calculations and with DM measurements with the MP512 itself and with Gafchromic™ EBT3 films. The comparison was performed with a gamma index analysis with the following acceptance criteria: 1%/1 mm, 2%/2 mm, and 3%/3 mm; a global threshold of 10% was applied. Using the averaged M (0.0196) and the B A0 value corresponding to any given field size, the dose in DM at d max was calculated using the TM measurement at a given SDD as:

3.A | Measurements in transmission mode and in dose mode
For an arbitrary radiation field of area A, B A0 could be found from the piecewise polynomial fit (adjusted regression coefficient R 2 = 1) (Figure 4):   When considering a gamma index analysis with a strict 1%/1 mm acceptance criterion, lower gamma passing rates (%GP) between our dose calculations and benchmarks (treatment planning system calculations, film dosimetry), which can be due to submillimeter misalignments in detector positioning or dose averaging in calculations, emphasize the importance of developing array detectors with highspatial resolution.
This study represents a first step in the development of a realtime high-resolution 3D dose reconstruction technique based on TM measurements with the MP512 prototype. We thank Todsaporn Fuangrod at the School of Medicine and Public Health, Chulabhorn Royal Academy, Bangkok, Thailand.

CONF LICTS OF INTEREST
The authors declare they have no conflict of interest.