Research Article  Onkologia i Radioterapia ( 2022) Volume 16, Issue 5
Analysis of dosimetric characteristics of energy 6 MV with and without flattening filter photon beam generated by the varian true beam linac using kruskal wallis H test
Achir Sara^{1}^{*}, Mohammed El Adnani Krabch^{2} and Mourad Trihi^{1}^{2}Department of Physics, FSR, Rabat, Morocco
Achir Sara, Department of Physics, FSAC, Casablanca, Morocco , Faculty of Sciences Aïn Chock, Casablanca, Morocco, Email: achir.sara1@gmail.com
Received: 25Jan2022 Accepted: 18Feb2022 Published: 25Feb2022
Abstract
The aim of this study is to assess the capacity of the beammatching of the characteristics of the Flattened (FF) and Unflattened (FFF) photon beam of the 6 MV energy of the Varian TrueBeam linac located at the Achifaa Specialized Center in Casablanca (Morocco). The dosimetric parameters in this research such as Percentage Depth Dose (PDD), surface dose, beam profile, Output Factor (OF), Transmission Factor (TF) and Dosimetric Leaf Gap (DLG) are analyzed by a nonparametric Kruskalwallis test in order to study their degree of independence and a coefficient of variation a descriptive statistical analysis to measure the difference between these measurements. We observed a similarity between the FFF and FF beam data for surface dose and the beam profile for large irradiation field sizes; the output factor for small to medium field sizes, it was noted that the maximum depth dose (dmax) and the DLG factor were less than 1 mm. But a slight disparity existed between the FFF beams compared to the FF beams for the surface dose which was higher for small fields, the dmax was 7% lower and closer to the surface, the average energy of the PDD was 5% lower; the TPR (20/10) phantom tissue ratio was 6% lower for the reference field size (10* 10 cm2). The energy spectra of the photons was smoother generally, profiles are sharper for medium to large field sizes, the penumbra width was smaller and 7% lower, the output factor was 2% lower for the selected fields, 5% lower for medium to large field sizes, transmission factor was reduced by 17%. Finally, the parameters were reproducible between both types of beam energy ×6 MV (FFFFF) since the pvalue> 5% and V.R <20%; except the DLG of the MLC was 24% lower for the same X6 MV energy.
Keywords
Kruskalwallis test, flattening filterfree, dosimetric characteristics, true beam
Introduction
A linear accelerator is a system that uses highfrequency electromagnetic waves (around 3000 MHz) to accelerate electrons to very high energies (around 25 MeV) through a linear tube. The beam that comes from this machine is typically used to treat surface lesions, strike metal targets (tungsten), or produce xrays that can give treatment to deeper tumours [1]. The TrueBeam linear accelerator (TBLINAC) equipped with a Millenium 120 leaf MLC (Varian Medical System, USA) was installed at the Achifaa specialist Center in Casablanca, Morocco.
The FF beam "flattening filter" was one of the basic components of the treatment head of the medical accelerator, located between the primary collimator and the ion chamber [2]. The principal function of the Flattening Filter (FF) in the Xray beam path of the linear accelerator is to provide practically uniform fluence over a collimated field. This characteristic was the most commonly employed in traditional radiotherapy [3]. Studies by [47] mentioned that the flattened filter was one of the components that contribute to the dispersion of the number of particles. This is precisely the reason why many authors such as [813], have been interested in the removal of the flattened filter for its high performance compared to the flattened filter [2].
They concluded that this mode produces a softening of the spectrum, a reduction in scattered radiation from the head of the linear accelerator, a nonuniform beam profile (forward peaked dose profile), the maximum dose as close to the surface as possible, a higher surface dose, reduced MLC leakage and less variation in outoffield dose [7,5].
In addition, there were many publications on the subject. Although FFF beams were installed in many centres, FF mode is stilled in use. In this study, energy 6 MV between FF and FFF photons of True Beam linear accelerator was investigated. The research is based on a descriptive and nonparametric statistical analysis for dosimetric characteristics.
Material and Methods
Nonparametric and descriptive statistical analysis
For nonparametric statistical analysis, the Kruskalwallis test evaluates to test the null hypothesis that all independent samples K come from groups with equal means and the alternative hypothesis that at least one group is different. The hypotheses for the test are: H0: population medians are equal H1: population medians are not equal (Significance pvalue <0.05) and descriptive statistical analysis, the coefficient of variation assess the uniformity of the defined distribution (V.R<20%) [1419].
Commissioning photon beams
The determination and comparison of parameters measured for commissioning Eclipse Treatment Planning System (TPS) involve Percentage Depth Dose (PDD), surface dose, beam profile, Output Factor (OF), Dosimetric Leaf Gaps (DLG) and Multileaf Collimator (MLC) transmission factor [20, 21]. All these parameters were measured by a 0.125 cc cylindrical chamber; model 31010 Semiflex (PTW German) was used for ionization field and reference. The chamber specifications include an effective length of 6.5 mm and an inner diameter of 5.5 mm. The output factor was measured with a PTW pinpoint chamber; model 31016 (PTW German), with the nominal sensitive volume, is 0.016 cc, an active length of 2.9 mm and an internal diameter of 2.9 mm [1].
Percentage Depth Dose (PDD): PDD refers to the percentage depth dose curve. The energy performance index of the beam is determined by the following formula of the depth of the measuring point for a constant SSD. This parameter depends on the energy and field size [2,13]. PDD was studied for FF and FFF beam at a depth of 10 cm with respect to the maximum absorbed dose position [13].
A maximum depth dose Dmax, PDD at 10 cm quality index defined as the absorbed dose at any depth relative to the position of the maximum absorbed dose determined by the empirical formula TPR20/10=1.2661 D20/10 0.0595 expressed as a percentage [13,22], PDD (20, 10) is ratio of the percentage depth dose at 20 cm and 10 cm depth. This factor gives a notion of stability of the photon beam [1]. A value of an additional energy parameter was obtained for comparison using a TPR (20, 10) a quotient of dose for 10 × 10 fields at the depths of 20 cm and 10 cm. This analysis of Percent Depth Dose (PDD) data was performed to evaluate the energy match between the FF and FFF beams that were compared for selective field sizes (6 × 6, 10 × 10, 20 × 20, 30 × 30 and 40 × 40) cm2 [15].
Surface dose: The surface dose is the dose calculated at the entrance of the phantom [2]. The value of the surface dose of any calculated field size is obtained by dividing the dose for the first millimetre (1 mm) of the homogeneous water model by the dose D for the corresponding field [13].
In this work, for open field size (6 × 6, 10 × 10, 20 × 20, 30 × 30 and 40 × 40) cm2, estimated surface doses was evaluated between the 6 MV FF and FFF configurations [2].
Profiles: Beam profiles for FF and FFF modes were compared, measured at selective field sizes (6 × 6, 10 × 10, 20 × 20, 30 × 30 and 40 × 40) cm2 and evaluated at a depth of 10 cm and an SSD=100 cm [23]. The water phantom cannot measure the full profile for field sizes greater than 20 cm. In addition, the TPS treatment planning system accepts a half profile. In the present study, we cannot expose the data to these fields [1].
Output factor: In this study, the Output Factor (OF) was obtained based on the dose ratio of a given field size to a reference field size at the same depth in water. This factor includes the phantom and collimator scatter and was normalized to a (10 × 10) cm2 reference field at a depth of 5 cm (95 cm SSD) [8]. The OF was measured for squared and rectangular field sizes ranging from (3 × 3) cm2 to (40 × 40) cm2 for 6 MV and 6 MV FFF photon beams. This factor takes it into consideration the variation of the total scatter factor. This data can determine variations in the configuration of the beam filter and other characteristics of the head structure of the linear accelerator [23].
Transmission factor: The multileaf collimator does not completely block a small part of the radiation, but transmits directly between the leaves [23]. For all MLCs closed behind the jaws, using the same size, the transmission can be estimated as the ratio of the measured dose in an open field, and the dose measured by all MLCs closed behind the jaws with the same size [1].
Dosimetric Leaf Gap (DLG): The round leaf tip is a typical Varian MLC system. The physical difference between the light and the irradiation field formed by the MLC is defined as DLG [20]. Therefore, DLG should be measured as an important factor to quantify the contribution of this transmission to the calculation of the dose to the patient [24, 25].
Results and Discussion
Percentage Depth Dose (PDD) and TPR _{(20,10)}
The PDD curves were measured from water surface (0 mm) to a depth of 300 mm for energy 6 MV without flattening filter and with a filter. For the (6 × 6) cm2 and (40 × 40) cm2 irradiation fields, the mean dose values at 10 cm depth for the FFF and FF beams showed a p value> 0.05, which proves that there is a statistically significant difference between these two beams. This is due to the hardening of the beam through the flat filter [14,26], the exponential region of the FFF profile has a steeper slope than the flattening filter, as shown in figure 1, except for the field ( 10 × 10) cm2, (20 × 20) cm2, (30 × 30) cm2 the curves were similar and showed a pvalue <0.05 (Table 1), but to quantify this effect, a comparison of the characteristics of the quality index between 6X FF and 6X FFF of the linear accelerator was defined in Table 2.
At PDD_{10} (%), the rate of change of the unflattened beam of this energy was 5% lower compared to the flattened beam, this is due to the decrease in the quantity of contaminating electrons and to the absence of a beam hardening effect [7], but the difference was not statistically significant. The result corresponds to that reported by [5, 12, 21] and seem convergent with the results of [9,18,22,27] with a slight difference (about 1%; 4%).
A significant parameter of the PDD of a beam is the depth of dose maximum; it’s to note that the rate of change of dmax values of FFF beam was 7% lower and closer to the surface than FF beam for the selected field sizes, because of the lowenergetic photons decrease [5], but this variation remains at a minimum. Indeed, it has been reported that the difference in the value of dmax between the filter and without a filter was less than 1 mm [28]. The result is slightly different from [3, 9] with a difference of (1%; 3%).
In Figure 1, the distance between both curves increases linearly with the irradiation field. The dose ratio between 20 cm and 10 cm depth (D20/D10) was used to quantify this effect. The FFF ratio was 7% lower than FF beam. Table 2, the difference between both curves is nonsignificant. This dose ratio shows a slight disparity with [2], (approximately 1%). The TPR20/10 beam quality index had no relation to the electron contamination in the incident beam [18]. This index is presented and compared for the voltage 6 MV FFF, FF in Table 3.
The absence of the flattening filter affects the quality index; hence the rate of change of the quality index TPR20/10 of the FFF was 6% lower than for the FF beam respectively for a reference field size (10 × 10) cm2. The elasticity curve in depth of the MV FFF beam beyond dmax decreased slightly compared to that of the MV FF beam. In fact, the MV FFF beam favours a more pronounced dose reduction FF [8]. This downward movement is due to the reduction in scattering and electron contamination of the flattening filter which affects the dose in depth after the maximum depth dose (D_{max}).
This indicator of the True beam machine is compatible with the reports [5, 9, 21] and showed minimum variation (around 1%; 2%; 3%). Indeed these quality indexes have demonstrated the reproducibility of the measurements of the PDD parameter of the energy X6 MV between FFFFF.
Surface dose
The surface dose of 6MV FF and FFF configurations was determined for open field sizes (6 × 6, 10 × 10, 20 × 20, 30 × 30 and 40 × 40) cm2. Table 4 indicates the surface dose between FF and FFF for these irradiation fields producing a pvalue>0.05 overall, consequently this statistical test checks the reproducibility of the data. However, the coefficient of variation of surface dose values of FFF beams was 5% greater than FF for all field sizes. We observed that in general FFF beams generated more doses to the skin for small until medium fields; it was 9% less than FF and a similar or even less dose for large field sizes (Table 5).
Tab.1. Observed results from kruskalwallis test for the percentage depth dose of energy 6 mv
H0 
Data 
Statistical tests 
F.S (cm2) 
Sig. 
Pvalue 
Decision 



6 × 6 
0,467 
>0,05 
Retain H0 



10 × 10 
0 


H0: (FF)=(FFF) 
PDD 
kruskalwallis 
20 × 20 
0 
>0,05 
Retain H0 



30 × 30 
0 





40 × 40 
0,467 
>0,05 
Retain H0 
Tab.2. Depth dose analyses parameters of FF and FFF photon beams performed at SSD=100 cm from the varian true beam accelerator
Energy 
F.S (cm2) 
PDD10 (%) 
Dmax (mm) 
PDD (20,10) 
6 MV FF 

66,1 
13,5 
0,57 

10*10 



6 MV FFF 

63 
12 
0,54 
Tab.3. Difference of TPR (20/10) calculated beam quality index with FF and FFF beam energy
Index of quality 
F.S ( cm2) 
Energy (MV) 

TPR20/10 
6 MV FF 
6 MV FFF 

0,663 
0,624 
The surface dose increased linearly with field size for FF and FFF photon beams, due to the removal of the flattening filter promoting the elimination of the primary electrons entering by the thin high Z targets used to generate bremsstrahlung photons [7,2830]. The difference between them is clearly visible in Figure 2. The results presented in Table 5 were compared to the study [18, 21] with a difference of (2%; 3%).
Beam profile characteristics
As presented in Table 6, the FFF and FF beam profiles were compared using the Kruskalwallis test. For large irradiation fields (30 × 30) cm2 and (40 × 40) cm2 the dose values at 10 cm depth showed a pvalue> 0.05 but for the field (6 × 6) cm2, (10 × 10) cm2 and (20 × 20) cm2 the curves were similar and showed a pvalue <0.05. Overall this implies that there was no statistical difference between FFF and FF of the energy X6 (Table 6).
Figure 3 and Figure 4 show the dose profile and halfdose profiles of FF and FFF beams for selective field sizes (6 × 6, 10 × 10, 20 × 20, 30 × 30 and 40 × 40) cm2 evaluated at a depth of 10 cm. From a field size, (6 × 6) cm2 to (10 × 10) cm2, the shape of the profile was very slightly affected for ×6 FFF. There was no significant difference; indeed the rates of change of FFF were 1% and 4% lower than FF (Figure 3 and Table 7). The average energy of photon spectra of FFF was 11% lower compared to FF, because the removal of FF elicits a rapid distribution of the low fluence [5,9,30].
In Figure 4, the FFF beam for large irradiation fields (30 × 30 cm2 and 40 × 40 cm2) appeared with a pronounced forward peak. Obviously this beam lacked flatness and that its maximum dose was on the central axis and gradually decreased towards the edge of field [12]. In order to quantify the peak amplitude of nonflat profiles, the relative dose at 80% for FFF beams can be used, profile measurements tend to decrease, a 5% reduction in dose for FFF compared to FF beams at the same depth.
Tab. 4. Observed results from kruskalwallis test, for the surface dose of 6MV between FF and FFF photon beam
H0 
Data 
Tests 
F.S (cm2) 
Sig. 
Pvalue 
Decision 
6 × 6 

10 × 10 

H0: (FF)=(FFF) 
Surface_Dose 
kruskalwallis 
20 × 20 
0,406 
>0,05 
Retain H0 
30 × 30 

40 × 40 
Tab.5. The variation rates of surface dose of the 6MV FF and FFF at SSD=100 CM
Energy 
Statistical tests 
Data 
Fieldsize (cm2) 
6 MV FF 
6MV FFF 
Variation rates (%) 



6 × 6 
0,592 
0,669 
13% 



10 × 10 
0,625 
0,688 
10% 
6 MV 
Variation rates (%) 
Surface_Dose 
20 × 20 
0,699 
0,732 
5% 



30 × 30 
0,756 
0,759 
0% 



40 × 40 
0,78 
0,774 
1% 





Average of V.R (%) 
5% 
Tab. 6. Observed results from kruskalwallis, for the mean of relative doses of 6MV between FF and FFF photon beam
H0 
Data 
Test 
Fieldsize (cm2) 
Sig. 
Pvalue 
Decision 



6 × 6 
0,02 





10 × 10 
0,01 
< 0,05 
Reject H0 
H0: (FF)=(FFF) 
Beam_profile 
kruskalwallis 
20 × 20 
0,01 





30 × 30 
0,154 
> 0,05 
Reject H0 



40 × 40 
0,362 





Average 
0,111 
> 0,05 
Reject H0 
Tab. 7. The variation rates of mean of relative doses of the 6MV FF and FFF at SSD=100 CM
Energy 
statistical tests 
Data 
F.S (cm2) 
6 MV FF 
6 MV FFF 
Variation rates (%) 



6 × 6 
25, 24 
24, 88 
1% 



10 × 10 
41, 73 
40, 12 
4% 
FFFFF 
Variation rates (%) 
Beamprofile 
20 × 20 
50 
44, 41 
11% 



30 × 30 
28,41 
23, 31 
18% 



40 × 40 
73, 11 
56, 46 
23% 





Average of V.R (%) 
11% 
Tab.8. Values of relative doses for field sizes equal to 80% for FF and FFF photon beams for depth 100 MM
Relative dose 80 % of Energy=6 MV 

Depth (mm) 
Linear accelerator 
F.S (cm2) 
FF 
FFF 
Variation rates (%) 


6 × 6 
97,3 
96,3 
1% 


10 × 10 
98,2 
95,2 
3% 


20 × 20 
100,5 
95,5 
5% 


30 × 30 
100,8 
95 
6% 


40 × 40 
101,8 
93,2 
8% 




Average of V.R (%) 
5% 
Tab.9. Characteristics of beam profiles calculated with and without FF for energy 6 MV photon beams for the selected field sizes at a depth of 10 CM and a SSD of 100 CM
P80_20% (mm) of energy=6 MV FFFFF 

Depth (mm) 
Linac 
F.S (cm2) 
FF 
FFF 
Variation rates (%) 


6 × 6 
0,648 
0,638 
2% 


10 × 10 
1,631 
1,544 
5% 
100 
True Beam 
20 × 20 
2,78 
2,515 
10% 


30 × 30 
2,807 
2,542 
9% 


40 × 40 
2,815 
2,55 
9% 




Average of V.R (%) 
7% 
Tab.10. Observed results from kruskalwallis test, for the output factor of 6mv FF and FFF photon beam
H0 
Data 
Tests 
F.S (cm2) 
Sig 
Pvalue 
Decision 
H0: (FF)=(FFF) 
FOC 
kruskalwallis 
3 × 3 ≥ 40 × 40 
0,443 
>0,05 
Retain H0 
Tab.11. summarized of outfieldscatterfactor for FFF and FF beams
F.S(cm2) 
6 MV FF 
6 MV FFF 
Variation rates (%) 
3 × 3 
0,845 
0,846 
0% 
4 × 4 
0,87 
0,868 
0% 
5 × 5 
0,892 
0,888 
0% 
7 × 7 
0,926 
0,916 
1% 
10 × 10 
0,962 
0,946 
2% 
12 × 12 
0,98 
0,958 
2% 
15 × 15 
1 
0,974 
3% 
20 × 20 
1,027 
0,993 
3% 
25 × 25 
1,046 
1,004 
4% 
30 × 30 
1,061 
1,014 
4% 
35 × 35 
1,071 
1,017 
5% 
40 × 40 
1,078 
1,023 
5% 



2% 
The width of the penumbra decreased with increasing field size, for FFF case was 7% lesser than FF case. These differences between FF and FFF in the penumbra width were due to its lower average energy, resulting in a shorter span of secondary electrons [5].
Relative photon output factor
Relative photon output factor measurements for FF (6 MV) and FFF (6 MVFFF) were performed for field sizes ranging from (3 × 3) cm^{2} to (40 × 40) cm^{2}. This measured output factor varies from 0.845 to 1.078 for 6 MV, while from 0.846 to 1.023 for 6 MVFFF is presented. The increase of the output factor is in line with the increasing field size. These values are analysed statistically as can be seen in Table 10. This statistical analysis demonstrates that the distributions were homogeneous in nature since the pvalue> 0.05.
As indicated in Figure 5 and Table 11, the outoffield dose of the FFF beam decreases with increasing field size in relation to the FF beam, for a small field sizes, indeed there was no significant difference between the values of this parameter until that they converge to unity at a field size of (20 × 20) cm^{2}, but at (25 × 25) cm^{2} to a larger field size, the output factor values of 6 MV FFF were 5% lower than 6 MV FF. This reduction is caused by less secondary radiation emanating from the jaw of the accelerator head, therefore less electronic contamination [31, 32].
However, the outfielddose for the FFF beams was slightly low with a rate of change of 2% less than the FF beam for all irradiation field sizes. It was consistent with that reported by [2], the difference is not substantial with (approximately 1%2%).
Transmission factor/dosimetric leaf gap
The transmission factor MLC and DLG are 17%, 24% inferior to the flattening filter of the same energy X6 MV (Table 12). The softening effect of the Xray spectra and the reduction in the average beam energy 6 MV induce a decrease in the lateral dose and a change in the shape of the beam profile FFF that can lead to an optimization of the leakage radiation at the edge of the field could cause a reduction of MLC transmission [31].
In Table 12, transmission causes a decrease in the DLG value; these parameters are essential during the setting of the MLC treatment planning system [13]. The results of this work were compatible with those of with a nonsignificant difference (approximately ranging from 1% up to 2%).
Conclusion
In this present work, dosimetric parameter analysis of which KruskalWallis and the variation rates have been applied to evaluate the photon beams with and without the 6 MV energy flattening filter of a Varian True Beam TM medical linear accelerator.
The results showed that the surface dose, the beam profile for large field sizes and the outoffield doses for small to medium field sizes are almost similar between FFF and FF as well as a maximum depth dose less than 1 mm for these two modes.
Despite this homogeneity, there was a slight difference in:
• Depth dose curves of FFF beam which have a greater depth compared to the FF beam
• D_{max} FFF closer to the surface
• Increase in surface dose for FFF for small field size.
• Nonuniform beam profile (direct peak dose profile) for medium to large field size
• Beam penumbra width for FFF less than FF.
• Output factor of FFF at large field size greater than FF
Especially since the coefficient of variation of these parameters of FFF beam is less than 20% and the pvalue of the Kruskal Wallis test was greater than 5%. We can therefore deduce that these measurement data were homogeneous and reproducible between both types of X6 energy beams of the Varian True Beam TM except for the dosimetric leaf gap, its coefficient of variation was slightly greater than 20%.
Certainly in terms of quality of FFF beam, as the softening of the photon energy spectra, the shift of the maximum dose to the surface, the peak forward and the smaller penumbra width resulting in reduced dispersion coming from the head have been very beneficial for treatment.
Furthermore, for better comfort during treatment, surface dose optimization to avoid acute skin reactions, for a faster dose calculation and a more optimized radiotherapy treatment for cancer patients, we have found that is necessary to optimize the output factor for less radiation exposure to healthy tissue in order to eliminate any risk of secondary malignancies and many cancer complications. A reduction in leaf transmission affecting the DLG to prevents dose measurement errors in the PTV and OAR (organs at risk).
Acknowledgment
I would like to thank Pr. Trihi Mourad the thesis supervisor and professor of Faculty of Sciences Ain Chock Casablanca, University Hassan II of Casablanca. In addition, I would like to express my sincere thanks to Dr. Mohammed El Adnani Krabch, a medical physicist at 16 november oncology clinic in rabat for his support of this work.