The impact of corrected field output factors based on IAEA/AAPM code of practice on small‐field dosimetry to the calculated monitor unit in eclipse™ treatment planning system

Abstract The objective of this study was to investigate the effect of field output factors (FOFs) according to the current protocol for small‐field dosimetry in conjunction to treatment planning system (TPS) commissioning. The calculated monitor unit (MU) for intensity‐modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT) plans in Eclipse™ TPS were observed. Micro ion chamber (0.01 CC) (CC01), photon field diode (shielded diode) (PFD), and electron field diode (unshielded diode) (EFD) were used to measure percentage depth doses, beam profiles, and FOFs from 1 × 1 cm2 to 10 × 10 cm2 field sizes of 6 MV photon beams. CC01 illustrated the highest percentage depth doses at 10 cm depth while EFD exhibited the lowest with the difference of 1.6% at 1 × 1 cm2. CC01 also produced slightly broader penumbra, the difference with other detectors was within 1 mm. For uncorrected FOF of three detectors, the maximum percent standard deviation (%SD) was 5.4% at 1 × 1 cm2 field size. When the correction factors were applied, this value dropped to 2.7%. For the calculated MU in symmetric field sizes, beam commissioning group from uncorrected FOF demonstrated maximum %SD of 6.0% at 1 × 1 cm2 field size. This value decreased to 2.2% when the corrected FOF was integrated. For the calculated MU in IMRT‐SRS plans, the impact of corrected FOF reduced the maximum %SD from 6.0% to 2.5% in planning target volume (PTV) less than 0.5 cm3. Beam commissioning using corrected FOF also decreased %SD for VMAT‐SRS plans, although it was less pronounced in comparison to other treatment planning techniques, since the %SD remained less than 2%. The use of FOFs based on IAEA/AAPM TRS 483 has been proven in this research to reduce the discrepancy of calculated MU among three beam commissioning datasets in Eclipse™ TPS. The dose measurement of both symmetric field and clinical cases comparing to the calculation illustrated the dependence of the types of detector commissioning and the algorithm of the treatment planning for small field size.

and FOFs from 1 × 1 cm 2 to 10 × 10 cm 2 field sizes of 6 MV photon beams. CC01 illustrated the highest percentage depth doses at 10 cm depth while EFD exhibited the lowest with the difference of 1.6% at 1 × 1 cm 2 . CC01 also produced slightly broader penumbra, the difference with other detectors was within 1 mm. For uncorrected FOF of three detectors, the maximum percent standard deviation (%SD) was 5.4% at 1 × 1 cm 2 field size. When the correction factors were applied, this value dropped to 2.7%. For the calculated MU in symmetric field sizes, beam commissioning group from uncorrected FOF demonstrated maximum %SD of 6.0% at 1 × 1 cm 2 field size. This value decreased to 2.2% when the corrected FOF was integrated. For the calculated MU in IMRT-SRS plans, the impact of corrected FOF reduced the maximum %SD from 6.0% to 2.5% in planning target volume (PTV) less than 0.5 cm 3 . Beam commissioning using corrected FOF also decreased %SD for VMAT-SRS plans, although it was less pronounced in comparison to other treatment planning techniques, since the %SD remained less than 2%. The use of FOFs based on IAEA/AAPM TRS 483 has been proven in this research to reduce the discrepancy of calculated MU among three beam commissioning datasets in Eclipse™ TPS.
The dose measurement of both symmetric field and clinical cases comparing to the calculation illustrated the dependence of the types of detector commissioning and the algorithm of the treatment planning for small field size.

| INTRODUCTION
The trend of utilizing the concept of small field in radiation therapy has been tremendously increasing over the past decades. 1,2 However, smallfield dosimetry presents several challenges due to three major problems.
The first is loss of lateral charged particle equilibrium (LCPE). This problem occurs when the beam half width or beam radius is getting smaller than the range of lateral charged particle equilibrium (r LCPE ). The second problem is partial source occlusion from collimating devices which yields to the larger full width at half maximum (FWHM) of dose profiles than the actual field size setting. 2 Both problems are beam related conditions, while the third obstacle is associated to the choice of appropriate detectors. Detectors with sensitive volume smaller than the beam radius are required as they are able to measure the dosimetric characteristics in small field at high spatial resolution. 3,4 Previous work reported disagreement within 14% among various types of detectors to determine the small field output factors (FOFs). 5 Selection of suitable detectors for small field output measurement becomes clearly cumbersome. 5,6 Several detectors such as ionization chambers and diode detectors have been developed specifically in accordance to the demand of smallfield dosimetry. The ionization chamber, which is commonly used in external beam radiotherapy provides the dose rate and energy independence. Nevertheless, ion chamber is found to underestimate the beam output, as a consequence of volume averaging effect. Volume averaging effect is an effect related to the corresponding signal from detector relative to mean absorbed dose over its sensitive volume. When the active volume of detector is larger than the beam radius, detector signal will be averaged incorrectly over its sensitive volume and eventually leads to the underestimation of measured dose. 7,8 Besides volume averaging, the perturbation of charged particle fluence due to the presence of a detector is an important issue and it must be noted that both effects are always entangled. 2 In contrast, diode detectors have been reported as the promising detectors for small field. 9 Diode detectors possess small active volume, excellent spatial resolution, and high sensitivity. However, the angular and energy dependence become the pitfalls when using diode. 1,9 The existence of encapsulating material with high atomic number and density also introduces another problem with respect to perturbation effect (backscattering from metallic electrode). [8][9][10][11] Prior to the establishment of new Code of Practice on small-field dosimetry, Alfonso et al. 12 The quantity k fclin;fmsr Qclin;Qmsr takes into account the difference between detector reading in clinical field and machine specific reference field.
This factor accounts any influences from volume averaging effect, as well as perturbation effect. 11 By using eqs. (1) and (2) Another attention related to the small-field external beam radiotherapy is the accuracy of output factors where the dose computation algorithm holds pivotal role on that. Two types of dose computation algorithms are available in radiotherapy TPS, namely type "a" and type "b" algorithms. 14 The algorithm of type "a" is correction based algorithm, where it strongly depends on the attenuation process. For type "b" algorithm, the calculation algorithm is model based algorithm to estimate the scatter and secondary electrons, which is essential to obtain an accurate dose estimation in small field. 15 The examples of type "b" algorithm are Anisotropic Analytical Algorithm (AAA) and Acuros XB (AXB). Recently, the first publication dedicated for small-field dosimetry used in external beam radiotherapy entitled Technical Reports Series (TRS) number 483 has been issued by IAEA in cooperation with AAPM Therapy Physics Committee. 2 The current protocol introduces and classifies the field output correction factor (k fclinfmsr Q clin Qmsr ) according to types of recommended detectors, the equivalent square field sizes (S clin ), energy of the beam, as well as types of the equipments (Cyber Knife, Gamma Knife, Linear Accelerator, and TomoTherapy).
In response to the establishment of current protocol for small-field dosimetry, the present study aims to address the effort for exploring the calculated MU in Eclipse™ TPS where the field output are collected from various types of detectors with the use of field output correction factors (k fclinfmsr Q clin Qmsr ) according to the current protocol. The comparison of calculated and measured output factors were presented for the commissioning of each detector, both corrected and uncorrected FOFs.
The measurement of isodose distribution in the small SRS cases compared with the calculations were also included.  Tables 1 and 2. All measurements were undertaken in IBA Blue water phantom. The field sizes were segmented using jaw setting.
Source to surface distance (SSD) was 100 cm. CC01 was set in T A B L E 1 Characteristics of ionization chambers.

2.C | Equivalent square field size
The IAEA/AAPM TRS 483 recommended that the radiation field size is defined by FWHM of the lateral beam profile measured at 10 cm depth. The measurement of radiation field was performed using EDGE detector with 100 cm SSD at 10 cm depth. The beam profiles were scanned in IBA blue water phantom from 1 × 1 cm 2 , 2 × 2 cm 2 , 3 × 3 cm 2 , 4 × 4 cm 2 , 6 × 6 cm 2 , and 10 × 10 cm 2 field sizes in both cross and in-plane directions. The dosimetric field width at 50% of relative dose (FWHM) from the cross (A) and in-plane (B) direction was recorded. Then, the equivalent square field size (S clin ) was calculated following equation 4 as follows: S clin was recommended to determine the field output correction factors according to TRS 483 Determination of FOF was divided into two groups: uncorrected FOF (ratio of detector reading at any field sizes to reference field size) and corrected FOF. Measurements were conducted in various collimator field sizes: 1 × 1 cm 2 , 2 × 2 cm 2 , 3 × 3 cm 2 , 4 × 4 cm 2 , 6 × 6 cm 2 , and 10 × 10 cm 2 . The SSD and reference depth were 100 cm and 10 cm, respectively. The output reading from each field size was normalized to the output acquired at 10 × 10 cm 2 machine specific reference field size (f msr ). IBA DOSE-1 electrometer was connected to each detector (CC01, PFD, and EFD) to measure the collected charge. For the first group, the use of field output correction factors (k f clin fmsr QclinQmsr ) based on IAEA/AAPM TRS 483 were omitted. For the second group, the field output correction factors from IAEA/ AAPM TRS 483 were applied using equivalent square field size S clin ð Þ and calculated and calculated following [eq. (2)].

2.E | Dose computation in Eclipse™ treatment planning system and verification
The second step in this research was computing the corrected FOF and uncorrected FOF for each commissioning detector type of the symmetric field in the TPS. Acuros XB algorithm was chosen to compute the measured data. For percentage depth doses, beam profiles, and output factors: the measured data from each detector mentioned above were entered into the TPS from 2 × 2 cm 2 until the field size of 6 × 6 cm 2 . The heterogeneity correction was turned on and the dose reporting mode was set into the dose to medium (D m ).

3.A.1 | Percentage depth doses and beam profiles
The comparison of measured percentage depth doses at 10 cm depth for three detectors is demonstrated in Fig. 1. All detectors showed agreeable percentage depth doses at 10 × 10 cm 2 and started to differ at 6 × 6 cm 2 , increasing more from 2 × 2 to 1 × 1 cm 2 field sizes. CC01 demonstrated slightly higher percentage depth dose, and EFD exhibited the lowest outcome comparable to PFD. The difference between PFD and EFD slightly increased more than 1% at 2 × 2 cm 2 and 1 × 1 cm 2 field size. EDGE was comparable to PFD with the maximum difference of 0.3%. CC13 was comparable within 0.5% to all detectors for field size down to 3 × 3 cm 2 , except CC01 at 6 × 6 cm 2 and EFD at 3 × 3 cm 2 that showed difference of about 1%.
For beam profiles, the lateral distance between 20% and 80% isodose line (penumbra width) at d max was analyzed as displayed in Fig. 2. Measurement using CC01 yielded broader penumbra within 1 mm compared to measurement using diode detectors, and more pronounced as the field size decreased. In contrast, penumbra width obtained from PFD and EFD agreed well. This was due to the equal size of both diode detectors.

3.B | Equivalent square field size
The equivalent square field S clin in 10 cm under the surface of water phantom is shown in Table 4, it is used to determine the FOF correction factor according to Table 26 in TRS 483.

3.C | Field output factors (Ω f clin f msr Q clin Q msr )
The FOFs for symmetric fields in this study were observed between the uncorrected and the corrected groups. Table 5 shows the uncorrected FOF of PFD, EFD, CC01, CC13, and EDGE. The uncorrected FOF from PFD were the highest among other detectors, followed by CC13 and EDGE detector which the values were comparable until 2 × 2 cm 2 field size. EFD exhibited the lowest field output factors.
The results from CC01 were between EDGE and EFD diode detectors. The %SD was considered for only three detectors of CC01, PFD, and EFD. An increase of %SD was discovered by decreasing the field size to 1 × 1 cm 2 , where the maximum %SD was 5.4%. The second measurement with the use of field output correction factors according to S clin from IAEA/AAPM TRS 483 2 is displayed in Table 6.
T A B L E 3 Summary of 10 SRS cases of brain tumors for MU comparison in IMRT and VMAT plans. F I G . 1. Percentage depth doses at 10 cm depth measured using CC01, PFD, and EFD in various geometric field sizes from 1 × 1 cm 2 to 10 × 10 cm 2 .
F I G . 2. Penumbra width between 20% and 80% isodose line at depth of maximum dose (d max ) measured using CC01, PFD, and EFD in various geometric field sizes from 1 × 1 cm 2 to 10 × 10 cm 2 .

| 69
The deviation among detectors significantly decreased after implementing the correction factors to each detector. In the smallest field size, the maximum %SD drastically reduced to 2.7%.

3.D | Verification of dose computation in Eclipse™ treatment planning system and measurement
The difference between TPS calculation and the measurement of symmetric fields for corrected and uncorrected FOF in three detectors and uncorrected FOF of CC13 and EDGE detectors used in clinical situation are plotted in Fig. 3. Both corrected and uncorrected FOF for three detectors and uncorrected FOF for EDGE and CC13 in clinical agreed well with the TPS calculation within 1%, for 2 × 2 cm 2 and 3 × 3 cm 2 field sizes, but deviated more for the larger fields of 4 × 4 cm 2 and 6 × 6 cm 2 . When comparing the difference with the detectors in clinical use, the three detectors exhibited close agreement for 2 × 2 cm 2 and 3 × 3 cm 2 field sizes than the larger fields of 4 × 4 cm 2 , and 6 × 6 cm 2 . For 1 × 1 cm 2 field size which was calculated by extrapolation data from the field of 2 × 2 cm 2 and more, the difference between calculation and measurement was larger in the uncorrected CC01 and smaller in PFD for both corrected and uncorrected FOF.
Other analysis was made by comparing the average uncorrected FOF to the average corrected FOF as shown in Fig. 4, where a strong agreement between both groups within 0.5% mean difference was discovered. It should be noted that only three detectors were involved (CC01, PFD, and EFD).

3.E | Comparison of calculated MU among commissioning datasets
For the calculated MU in symmetric field sizes, the average values for three detectors from 2 × 2 cm 2 to 6 × 6 cm 2 field sizes in both uncorrected and corrected FOF were very close, except at 1 × 1 cm 2 field size. It was apparent that commissioning from uncorrected FOF yielded %SD less than 2.0% from 2 × 2 cm 2 to 6 × 6 cm 2 field sizes as displayed in Fig. 5. The %SD eventually reached the largest of 6.0% (183 ± 11.0 MU) at 1 × 1 cm 2 field size.
The geometric field size along with the corresponding dosimetric field width in the cross-plane and in-plane direction measured at 10 cm depth of FWHM.  Fig. 6. When commissioning using corrected FOF was employed, the maximum %SD in case number 10 decreased to 2.5% (3981 ± 100 MU). The last observation was conducted in 10 VMAT-SRS plans. Following the plot in Fig. 7, the %SD from both commissioning groups were less than 2.0%. The %SD from commissioning using uncorrected FOF ranged from 1.1% to 1.9%, and slightly declined to the range from 1.0% to 1.7% for commissioning using corrected FOF.

| DISCUSSION
The focus of this study was to observe the effect of integrating cor- F I G . 5. Percent SD of calculated MU from 1 × 1 cm 2 to 6 × 6 cm 2 field sizes between commissioning using uncorrected FOF and corrected FOF. The number of MU (average ± SD) is labelled in each commissioning dataset for each field size.
CC01 represents the recommended ionization chamber, while the PFD and EFD designate the recommended diode detectors. CC01 illustrated the highest percentage depth doses at 10 cm depth while EFD exhibited the lowest. The maximum difference of 1.6% was observed at 1 × 1 cm 2 as previously shown in Fig. 1. For measured penumbra, CC01 produced slightly broader penumbra in regards to the larger size of detector. Nevertheless, the difference was only within 1 mm.
For the uncorrected FOF (ratio of reading), higher results from measurement using PFD, EDGE, and CC13 detectors were observed than the measurement using CC01 and EFD detectors. The uncorrected FOF of CC13 was lowest at 1 × 1 cm 2 field size due to the limited of large volume. Meanwhile, the unwanted scatter from high density of encapsulating material of PFD and EDGE was the main reason of higher outcome. However, smaller size of EDGE detector makes less effect of scatter than PFD. [8][9][10][11] Unlike PFD, EFD detector contains no high density material of encapsulating component. This enables the EFD to minimize an over-response from backscattering effect of shielding material. 8,9 When the FOF correction factors were applied, the three detectors come closer with the reduction of %SD. Once the measured output factors were configured to the TPS, the AXB algorithm did calculation. The tuning parameter to optimize the calculation of output factors was made by adjusting the effective spot size (ESS). ESS is the parameter associated to the geometric penumbra and partial source occlusion related to the small field condition. 27 Following the recommendation, ESS in this study was tuned to 1 mm in X and Y direction. 15 The tuning was performed with the same parameters for all the models.
Consider the detectors that were used for previous commissioning in the routine clinic practice without FOF correction: EDGE F I G . 6. Percent SD of calculated MU between commissioning using uncorrected FOF and corrected FOF in 10 brain SRS cases treated with IMRT technique. The number of MU (average ± SD) is labelled in each commissioning dataset for each case.
F I G . 7. Percent SD of calculated MU between commissioning using uncorrected FOF and corrected FOF in 10 brain SRS cases treated with VMAT technique. The number of MU (average ± SD) is labelled in each commissioning dataset for each case.
detector for 2 × 2 cm 2 and 3 × 3 cm 2 field sizes and CC13 for 4 × 4 cm 2 up to 40 × 40 cm 2 field sizes. The FOF correction factors for EDGE detector were close to one (0.999, 0.994 for 2 × 2 cm 2 and 3 × 3 cm 2 field sizes, respectively). Therefore, the corrected FOF of EDGE detector did not change much as seen in Table 6 compared to The equation above clearly demonstrates that two output correction factors are required, one for each detector.
The data for all detectors by IFM are shown in The comparison of measurement and TPS calculation of symmetric field for all detectors (Fig. 3) yielded a good agreement for 2 × 2 cm 2 and 3 × 3 cm 2 field sizes, but more deviated for 4 × 4 cm 2 and 6 × 6 cm 2 field sizes for both the uncorrected and the corrected FOF especially for CC01. The clinical data yielded better agreement with calculation for field sizes of more than 1 × 1 cm 2 field size. The set of small field data should match to the large field data. In this study, the % difference between measurement and calculation of clinical data was less than the three detectors where the commissioning data were 2 × 2 cm 2 to 6 × 6 cm 2 field sizes. The larger fields of more than 3 × 3 cm 2 commissioning data using CC13 illustrated good result. The poor signal noise ratio of small size ion chamber makes it inferior to the large volume ion chamber for large field measurement. The unshielded diode, in contrast, tends to over response in large field measurement due to its energy dependency. The increasing contribution of lower energy scattering photon from larger field will soften the beam.  The agreement of calculated MU among all commissioning datasets was characterized in terms of %SD. Overall, our research presented a significant reduction of %SD when the corrected FOF was incorporated to commissioning procedure. For symmetric field, %SD between both commissioning groups matched well from 6 × 6 cm 2 to 2 × 2 cm 2 field sizes. When the correction factors were applied to each detector, reduction of %SD was discovered until the smallest field size, which clearly demonstrates that the implementation of corrected FOF is able to reduce the discrepancy of calculated MU.
For the field size smaller than 2 × 2 cm 2 field size, the Eclipse™ TPS will calculate and extrapolate the dose according to the field output factors data set. If the corrected field output factors data are implemented, the calculated dose for smaller field sizes would be followed.
Similar trend was observed in IMRT-SRS plans where the implementation of corrected FOF notably reduced %SD. IMRT technique employs variable intensity across multiple composite fields with the use of MLC to achieve a highly conformal dose distribution to the small and irregular tumor. 29 Our results revealed that the reduction of %SD was more pronounced in case numbers 9 and 10 where multiple composite fields less than 2 × 2 cm 2 field size were predomi- The comparison of dose distribution between the calculation and measurement produced more deviation for IMRT plans with field sizes less than 3 × 3 cm 2 due to the field output factors difference.
The modulation factor seems to be less pronounced because of nonirregular tumor shape. The jaw tracking was implemented to both the IMRT and VMAT plans. The MLC for round shape tumors were not significantly modulated in the fluence map, so field output factors of the MLC can follow the field output factor defined by jaws.
Therefore, no significant differences in MU were observed for all groups of datasets.
The case numbers 9 and 10 were selected to compare the dose distribution between calculated and measured of IMRT plans using MapCHECK2. The result of case number 9 (3 × 4 cm 2 ) with gamma index of 2% dose difference and 2 mm distance to agreement (DTA) demonstrated that the calculated doses for corrected, uncorrected, and clinical data were comparable with 96.6% pass for all detectors.
However, in case number 10 (2 × 2 cm 2 ), the outcome showed datasets of corrected and uncorrected FOF by PFD and EFD passed 100%, which were different from CC01 for both corrected and uncorrected, as well as the clinical data set. They yielded the passing score of 88.9%. With the low resolution patient-specific QA tool, no difference was observed for different data set of commissioning. The high-resolution tools such as film should be studied for the next phase.
The limitation of this work was related to the commissioning procedure in Eclipse™ TPS. We were unable to integrate the field output factors less than 2 × 2 cm 2 . For smaller fields, it was left for Eclipse™ TPS to did extrapolation itself.

CONFLI CT OF INTEREST
No potential conflict of interest to disclose in this study.