Dosimetry audit in advanced radiotherapy using in-house developed anthropomorphic head & neck phantom

The treatment of head and neck (H&N) cancer presents formidable challenges due to the involvement of normal tissue and organs at risk (OARs) in the close vicinity. Ensuring the precise administration of the prescribed dose demands prior dose verification. Considering contour irregularity and heterogeneity in the H&N region, an anthropomorphic and heterogeneous H&N phantom was developed and fabricated locally for conducting the dosimetry audit in advanced radiotherapy treatments. This specialized phantom emulates human anatomy and incorporates a removable cylindrical insert housing a C-shaped planning target volume (PTV) alongside key OARs including the spinal cord, oral cavity, and bilateral parotid glands. Acrylonitrile Butadiene Styrene (ABS) was chosen for PTV and parotid fabrication, while Delrin was adopted for spinal cord fabrication. A pivotal feature of this phantom is the incorporation of thermoluminescent dosimeters (TLDs) within the PTV and OARs, enabling the measurement of delivered dose. To execute the dosimetry audit, the phantom, accompanied by dosimeters and comprehensive guidelines, was disseminated to multiple radiotherapy centers. Subsequently, hospital physicists acquired computed tomography (CT) scans to generate treatment plans for phantom irradiation. The treatment planning system (TPS) computed the anticipated dose distribution within the phantom, and post-irradiation TLD readings yielded actual dose measurements. The TPS calculated and TLD measured dose values at most of the locations inside the PTV were found comparable within ± 4%. The outcomes affirm the suitability of the developed anthropomorphic H&N phantom for precise dosimetry audits of advanced radiotherapy treatments.


Introduction
Radiotherapy is one of the major modalities for cancer treatment, along with surgery and chemotherapy.Advanced radiotherapy techniques, such as Intensity Modulated Radiation Therapy (IMRT) and Volumetric Modulated Arc Therapy (VMAT), offer the remarkable ability to precisely deliver concentrated doses to tumors while minimizing radiation exposure to critical organs at risk (OARs) and normal tissue [1].This attribute proves particularly valuable in treating complex clinical sites like the head and neck (H&N), where sensitive structures like parotid glands, spinal cord, and eyes are situated in close proximity to the tumor [2].The fundamental objective in H&N cancer treatment is to deliver the prescribed dose to the tumor while ensuring minimal radiation exposure to OARs [3].Customized radiotherapy plans are devised to achieve this dual aim.It's essential that these patientspecific treatment plans are executed with meticulous care.A small error in the treatment procedure can results in large deviation between planned and delivered dose [4].Hence, for the delivery of precise and accurate dose it is required to perform all the steps in the treatment procedure with great accuracy [5].
The pre-treatment dose verification is an important part to assure accurate delivery of the prescribed dose [6].This verification method provides an instantaneous verification of dose calculations before the start of a patient's treatment and hence, assures the accurate delivery of the planned treatment.In general, every radiotherapy center performs the pre-treatment dose verification right before the treatment delivery.Institutional pre-treatment dose verification is basically a self-evaluation which is prone to miss systematic errors that may be involved in the planning, treatment and dose analysis procedures.The methods that are selected for validating advanced radiotherapy treatment plans are rather dependent upon the decision of each institution.Third party dosimetry quality audit (QAu) program is an independent verification and is free from these limitations [7].The dosimetry QAu procedure is important to assess whether the participating institutes are providing the accurate dose delivery using uniform dosimetry methodology mentioned in the international protocols [8].
To ensure accurate dosimetric verification, an ideal phantom must replicate the tumor site with realism and encompass heterogeneous structures in its vicinity [9].Such phantoms are vital for the commissioning and clinical implementation of IMRT and VMAT for specific clinical scenarios.The materials used for fabrication must closely resemble human tissue in terms of scattering and attenuation.Although air-filled ionization chambers, electronic portal imaging devices, ion chamber arrays, and liquid ion chambers are commonly employed for pre-treatment dose verification [10], remote dosimetry QAu programs typically favor passive dosimeters such as thermoluminescent dosimeters (TLDs) and radiochromic films [11].Thus, the phantom should be designed to accommodate these dosimeters within the planning target volume (PTV) and OARs.In line with these requisites, an indigenous anthropomorphic and heterogeneous H&N phantom was conceived and developed.This study aims to explore the feasibility of utilizing this phantom for dosimetry QAu of advanced radiotherapy techniques, namely IMRT and VMAT.

Design and development of anthropomorphic H&N phantom
An anthropomorphic and heterogeneous head and neck (H&N) phantom, devised for dosimetry QAu purposes, was meticulously designed and developed, taking into account the following salient considerations: a. Representative Patient Size: The phantom's dimensions were tailored to closely resemble the average dimensions of a typical H&N patient.
b. Material Properties: Materials chosen for fabrication were not only non-toxic but also possessed structural integrity for seamless transportation.
c. Anatomical Fidelity: The design was meticulously crafted to mirror anatomical accuracy, ensuring conformity with internal geometries.

d. Material Properties Resembling Human Tissue:
The selected materials were required to closely mimic the radiological and physical characteristics of the actual relevant tissues.
e. Structural Differentiation: Visual distinction of neighbouring structures was incorporated to facilitate accurate delineation in the treatment planning system.
f. Ease of Assembly and Disassembly: User-friendly assembly and disassembly mechanisms were integrated into the design.
g. Geometrical Repositioning Facilitation: The phantom incorporated tools to simplify geometrical repositioning.
h. Provision for Different Dosimeters: The phantom was equipped to accommodate a range of dosimeters, facilitating comparative analyses.
i. Appropriate Weight: The weight was optimized for easy handling and transportation.
The dimensions of the H&N phantom including PTV and OARs were established based on averaged dimensions extracted from ten nasopharyngeal cancer patients.However, the final shapes were simplified in order to ease the manufacturing procedure.The different parts of the phantom such as phantom body, PTV insert and OAR inserts were fabricated from the blocks of requisite material.These blocks were shaped as per the planned design using cutting tools, such as CNC machine.The different parts of phantom were assembled and joined together using ABS-compatible bonding material to form the final structure of H&N phantom.The developed phantom is made up of body structure accurately emulating the general contours of the human H&N region.Within this body, a detachable cylindrical insert measuring 10 cm in diameter and 15 cm in length can be introduced.This removable insert encompasses a C-shaped PTV and assorted OARs, including the spinal cord, oral cavity, and bilateral parotid glands.The PTV assumes a C-shaped configuration with a diameter of 4.5 cm and a length of 7.5 cm.Correspondingly, the various OARs, such as the spinal cord, left parotid, right parotid, and oral cavity, are modeled as cylindrical structures measuring 2 cm in diameter and 7.5 cm in length.For illustrative purposes, figure 1 shows a photograph of the developed H&N phantom alongside the removable insert housing the PTV and OARs.
The selection of materials for the fabrication of distinct phantom components was determined by the radiological properties of the tumor and OARs within the H&N region.Acryline Butadiene Styrene [(C8H8•C4H6•C3H3N)n] (ABS) is a copolymer made by polymerizing acrylonitrile and styrene in the presence of polybutadiene.The composition of ABS is: Acrylonitrile (15 to 35%); Butadiene (5 to 30%) and Styrene (40 to 60%).An ABS plastic being a tissue equivalent material was used in the fabrication of PTV and Parotids.Similarly, Delrin [(CH2O)n], for its bone-equivalent characteristics, was selected for constructing the spinal cord.The phantom outer body was fabricated using Poly Methyl Methylene Acrylate (PMMA)[(C5O2H8)n].Table 1 shows the dosimetryrelated physical properties of the selected materials.
The effective Z (Z eff ) of the phantom materials were supplied by the manufacturer, for e.g.ABS: 5.76, Delrin: 6.95 and PMMA: 6.47.However, to validate manufacturer provided values, Z eff were computed using Mayneord's equation.The computed Z eff for ABS, Delrin, and PMMA were found to be 5.90, 7.01, and 6.63, respectively. ) where a 1 , a 2 , a 3 K an are the fractional contributions of each element to the total number of electrons in the mixture.
The phantom incorporates locations for measuring point doses at various positions.For point dose measurements, TLDs were chosen due to their suitability for remote dosimetry analysis, as they allow dose assessment several days post-irradiation the irradiation [12].Within the PTV, nine holes were drilled, spaced 1 cm apart along three planes.In a similar

Dosimetry quality audit methodology
The fundamental components of advanced radiotherapy treatment encompass three key stages: (i) Radiotherapy Imaging, (ii) Radiotherapy Treatment Planning, and (iii) Treatment Delivery [13].Achieving precise and accurate dose delivery to patients mandates the meticulous execution of each element within the treatment protocol.
For the implementation of dosimetry QAu, a TLD system based on LiF:Mg,Ti (TLD-100) discs (Thermo-Fisher Scientific, Waltham, USA) with a diameter of 4.5 mm and a thickness of 0.8 mm was employed.The dosimetry phantom, together with the TLD-100 discs, was dispatched to the participating centres.Accompanying this, a set of standard operating procedures (SoPs) was provided.These SoPs elucidated the protocol for (i) performing phantom imaging, (ii) delineating the pertinent structures, (iii) generating radiotherapy treatment plans, (iv) positioning dosimeters at reference points, and (v) executing the treatment delivery.This comprehensive package ensured a standardized approach to conducting dosimetry QAu at each participating centre.The selected audit centers are located in metro city.These audit centers/hospitals have been identified as H1 to H4 so that their exact identity could not be disclosed as per the policy of our study.Table 2 provides the details of the participating centers and beam delivery system.

Radiotherapy imaging
The CT scan parameters were selected as per the clinical Head and Neck protocol (Tube Voltage: 120 kVp, mAs: 320 mAs and slice thickness: 1.5 mm).The CT scans were performed by matching external marks on the phantom with machine lasers.This arrangement was useful for reproducible positioning of the phantom during treatment delivery.The photographs of phantom setup during the CT scan are shown in figure 3.
The contours outlining the structures of the PTV, OARs, and the designated positions for the TLDs were delineated on the CT images.These delineated structures are essential for generating the treatment plans and subsequently recording the doses calculated by the TPS at the specified TLD locations.Figure 4 provides photographic representations of the CT images, clearly showcasing the delineation of the structures on these images.This depiction offers insight into the  critical step of contouring that serves as the foundation for precise treatment planning and dose calculation.

Radiotherapy treatment planning
In the radiotherapy planning phase, the participating institutions were directed to formulate their radiotherapy treatment plans in strict adherence to the guidelines outlined in the provided SoP.To ensure consistency and standardization, the dose volume constraints and beam arrangement specifications essential for creating treatment plans were sourced from the Task Group 119 (TG-119) protocol designed for a simulated head and neck tumor scenario [14].
The particulars of dose volume constraints are detailed in table 3, while the specifications for beam arrangements are presented in table 4.These references provided a benchmark against which treatment plans could be optimized and assessed for conformity to established standards.Analytical Anisotropic Algorithm (AAA) dose calculation algorithm was used for IMRT and VMAT.The dose volume histogram (DVH) of PTV and OARs  Upon the completion and approval of the treatment plans, the next step involved the practical implementation of these plans on the phantom.This was achieved by placing TLDs at their designated positions by the hospital physicist of the participating institutions.Figure 6 provides photograph capturing the treatment delivery setup for both IMRT and VMAT.

Dosimetric analysis
The TLD system used for audit work consists of dosimeters in the form of disc (diameter 4.5 mm and  thickness 0.8 mm) of LiF:Mg,Ti (TLD-100; Thermo-Fisher Scientific Waltham, USA) and a Harshaw model 3500 TLD reader (ThermoFisher Scientific, Waltham, USA).Prior to utilization in dose measurements, each TLD disc underwent an annealing process consisting of one hour at 400 °C, followed by rapid cooling for six minutes, and a subsequent two-hour period at 100 °C.
A meticulous assessment of the dosimetric characteristics of the TLD-100 discs was conducted before dispatching them to the participating institutions.One hundred TLDs were procured from the manufacturer, and for the dosimetry audit, TLDs with a coefficient of variation (CoV) within ± 2% were selected.To achieve this, the initial set of 100 TLDs were subjected to irradiation with a radiation dose of 100 cGy, utilizing a 6 MV photon beam with a 10×10 cm 2 field size in a solid water phantom at a depth of 1.5 cm.This irradiation process was performed in triplicates, and the responses of the TLDs were recorded.From the initial pool of 100 samples, 30 TLDs exhibiting a CoV within ± 2% were identified.Additionally, individual dosimeter sensitivity correction factors (SCFs) were determined for each of these selected TLDs.The SCFs were computed as the ratio of the average signal from the group of dosimeters (30 samples) to the signal of an individual dosimeter.
The generation of a dose-response curve (calibration curve) necessitated irradiating the TLDs using a 6 MV x-ray beam at a depth of 5 cm, maintaining a source-to-surface distance of 100 cm.The absorbed dose to water at the reference distance was pre-determined utilizing the methodology outlined in the International Atomic Energy Agency (IAEA) Technical Reports Series (TRS) 398 protocol [15].Given that the prescription dose for most treatment plans was 200 cGy, the calibration curve was established over a dose range spanning from 25 to 400 cGy (i.e., 25, 50, 100, 150, 200, 250, 300, and 400 cGy).
The dosimetry phantom, alongside the TLD-100 discs, was dispatched to the participating centres.All audited centers are situated within a 50 km radius from the audit conducting center.The phantom and detectors were transported by road using a car.Additionally, a control TLD was transported round-trip to assess the radiation dose received by the field TLDs.Upon the completion of the audit procedure, the exposed TLDs, together with TPS-calculated doses at the reference points, were sent back to us for further analysis.By applying the calibration factor, the response of each TLD was converted into dose units.Subsequently, the TLD-measured doses at all locations within the PTV and OARs were compared to the TPS-calculated doses using the methodology outlined in the American Association of Physicists in Medicine (AAPM) Task Group Report 119.
The percentage variation between TPS calculated and TLD measured dose were estimated as; Where, D TPS is the TPS calculated dose, D TLD is the TLD measured dose, and D PRESCRIBED is the prescribed dose.
For the preliminary dosimetry quality audit program, the dosimetry phantom, along with dosimeters and the Standard Operating Procedures were disseminated to four different hospitals that utilize IMRT and VMAT techniques in their treatment practices.

Results
A preliminary dosimetry QAu investigation was conducted across four different hospitals, each equipped with IMRT and VMAT capabilities.The percentage variation between TLD-measured doses and TPScalculated doses at different locations within the PTV and other OARs for VMAT technique is shown table 5. Similarly, table 6 shows the percentage variation between TLD-measured doses and TPS-calculated doses for IMRT technique.
The analysis of percentage variation between TLDmeasured and TPS-calculated doses for VMAT and IMRT techniques across various locations within the PTV yields a range of −4.45% to 4.93% for VMAT and −4.35% to 5.66% for IMRT.At most of the locations, these variations remain within the limits of ± 4% for VMAT and ± 3.5% for IMRT.The percentage variation within OARs during VMAT remains below the threshold of ± 4%.With respect to the IMRT technique, the left parotid and the spine exhibit percentage variations under ± 3%.However, the right parotid displays a broader range, spanning from −7.63% to 5.71%.These findings slightly exceed the passing criteria of ± 7% set forth by the Radiological Physics Center, particularly for the high dose gradient region situated between PTV and OAR [16].

Discussion
The dosimetric verification for IMRT and VMAT radiotherapy plans was investigated by comparing the TPS-calculated and TLD-measured dose at different locations of PTV and OARs.The measured doses at most of the location within PTV are in good agreement with the TPS calculated dose.However, the higher percentage variation between measured and calculated dose ranging from −7.63 to 5.71% was observed in the case of right parotid.Imaging and Radiation Oncology Core Houston Quality Assurance Center (IROC-H) used H&N phantom for credentialing of IMRT, requires that dose difference between an institution's treatment plan agrees within ± 7% of measured thermoluminescent dosimeter (TLD) doses [17,18].Our results for right parotid are slightly higher, while dose to other points are well below to this criteria.IROC-H's H&N phantom does not have the parotids.Higher percentage dose variation is expected and it may be attributed to slight positioning error of dosimeter in high dose gradient region.Kowalik et al have also observed significant differences between the planned and measured dose in the area of high dose gradients, i.e. on the OAR-target border [19].This indicate that in the verification of complex plans involved in the H&N treatment, the measurement points should be selected in low dose gradient regions to reduce the potential impact of setup errors in the measurement.
Countries like the UK and USA have established regional audit group networks for conducting Quality Assurance (QA) programs [20].In India, where over 500 radiotherapy centres administer cancer treatment using radiation, many employ advanced techniques like IMRT and VMAT.However, there is currently no dosimetry audit network for such advanced techniques.In this pilot study, we explored the feasibility of conducting a dosimetry audit using a developed phantom and methodology.By extending the applicability of our established phantom and methodology, we aim to foster a more comprehensive and standardized approach to dosimetry audits.This, in turn, will contribute to the establishment of a robust network that ensures the uniformity and accuracy of dosimetry practices across diverse radiotherapy centres in the country.Further, there is a necessity to create site-specific phantoms to facilitate dosimetry audits on a larger scale.The ongoing and future development of such site-specific phantoms and corresponding methodologies holds substantial promise for advancing the dosimetry quality audit network in India.The insights gained from this quality audit will be implemented to streamline the audit structure effectively.The widespread implementation of robust QA procedures on a larger scale has the potential to significantly reduce inaccuracies in treatment delivery, thereby improving clinical outcomes [21].Given the rapid advancements in radiotherapy technology, the complexity of treatment dose delivery has increased substantially.Consequently, the establishment of comprehensive dosimetry QA methods becomes paramount to address the challenges presented by these advanced treatment procedures.A comprehensive QA program aids in defining acceptable and unacceptable variations, facilitating proactive measures to rectify root causes and document the frequency of variations.
Dosimetry QA programs are generally conducted through either on-site audits or remote audits by sending dosimetry systems and related materials via postal services [22][23][24][25].In this context, the developed phantom has been purposefully designed to facilitate both remote and on-site dosimetry audits.The chosen materials for the phantom prioritize non-toxicity, durability, and structural integrity.Additionally, these materials approximate tissue and bone densities, enabling the differentiation of structures in CT images and treatment planning.The phantom's assembly and disassembly process is efficient and reproducible, with its manageable weight and size making it suitable for manual handling and remote transportation.In this trial study, phantom has been primarily utilized for QA using TLDs.The developed phantom has a provision of measurement using Gafchromic EBT3 film.Hence, in the next phase, dosimetry audit will be conducted including the gamma analysis of TPS calculated and film measured fluence.Before deploying the phantom and dosimeters for QA programs at various radiotherapy hospitals, the phantom's performance and methodology were rigorously tested by our team at one of the hospitals.Following a satisfactory evaluation, a comprehensive Standard Operating Procedure (SoP) detailing imaging, planning, and treatment delivery procedures was finalized and distributed to participating institutes to ensure uniformity and consistency in the dosimetry QAu process.

Conclusion
The development of an anthropomorphic and heterogeneous H&N phantom was undertaken with the specific aim of integrating it into the dosimetry QAu process for advanced radiotherapy methodologies.This methodical approach to dosimetry QAu was meticulously formulated and subsequently put into practice for IMRT and VMAT techniques.The outcomes yielded a good agreement with the prescribed acceptance criteria set forth by the IROC-H across all participating medical institutions.
this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Figure 1 .
Figure 1.In-house developed anthropomorphic head and neck phantom.

Figure 2 .
Figure 2. Photographs showing holes for the placement of TLDs inside PTV and OARs.

Figure 3 .
Figure 3. Photographs showing the phantom setup during the CT scan.

Figure 4 .
Figure 4. Photographs showing axial computed tomography image and the different structures delineated on it.

Figure 5 .
Figure 5. Photographs showing the dose distribution and dose volume histograms for VMAT.

Figure 6 .
Figure 6.Photograph of the treatment delivery setup.

Table 1 .
Dosimetry related physical properties of the materials.
manner, the parotid glands and spinal cord structures are capable of accommodating three TLDs each: one at the center and two additional TLDs situated 1 cm away from the central one.A photograph depicting the distinct structures with TLD grooves is displayed in figure 2. To facilitate consistent positioning during computed tomography (CT) scanning and treatment administration, the phantom body features three fiducial marks.

Table 2 .
Details of the participating centers and beam delivery system.

Table 5 .
Percentage variation of TLD-measured and TPS-calculated dose for VMAT.

Table 6 .
Percentage variation of TLD-measured and TPS-calculated dose for IMRT.