|Year : 2022 | Volume
| Issue : 4 | Page : 336-343
Does fluence smoothing reduce the complexity of the intensity-modulated radiation therapy treatment plan? A dosimetric analysis
Dinesh Kumar Saroj1, Suresh Yadav2, Neetu Paliwal3
1 Department of Radiotherapy, Alexis Multispecialty Hospital, Nagpur, Maharashtra; Department of Physics, Rabindranath Tagore University, Bhopal, Madhya Pradesh, India
2 Department of Radiation Oncology, Gandhi Medical College, Bhopal, Madhya Pradesh, India
3 Department of Physics, Rabindranath Tagore University, Bhopal, Madhya Pradesh, India
|Date of Submission||06-Sep-2022|
|Date of Decision||17-Oct-2022|
|Date of Acceptance||20-Nov-2022|
|Date of Web Publication||10-Jan-2023|
Dr. Suresh Yadav
Department of Radiation Oncology, Gandhi Medical College, Bhopal - 462 001, Madhya Pradesh
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background: Intensity-modulated radiation therapy (IMRT) may have too many peaks and valleys, making the treatment plan undeliverable. When there are too many fluency differences between adjacent pixels in the X or Y directions, the X and Y smoothing factors are utilized as weighting factors to penalize this behavior. Generally, a high degree of complexity is accompanied by many monitor units (MUs), large number of segments, small sized segments, and complex segment shapes. The degree of plan delivery uncertainty can all increase with a higher detailed fluence map. Aim: This study aims to evaluate the dosimetric effects of various smoothing levels on the planning target volume (PTV) and organs at risk (OARs) for cervix cancer. Materials and Methods: IMRT treatment plans were re-optimized by combining several values of the X and Y penalty between 0 and 100. The dose–volume histogram assessed various dosimetric indicators for PTV and OARs. Additionally, gamma passing rates were evaluated and noted as an indicator of the complex treatment plan. Results: At X = 60, Y = 60 fluence map penalty, the conformity index (CI) value reached its highest value of 0.996 ± 0.004. At X = 0, Y = 0, the homogeneity index (HI) was determined to have a maximum value of 0.0628 ± 0.0235. The highest and lowest MU values were 2424.30 ± 471.12 and 1087.80 ± 91.57, respectively, with X = 0, Y = 0 and X = 100, Y = 100. At X = 100, Y = 100, the gamma passing rate reaches its highest value of 99.28% ± 0.44% and minimum value of 85.93% ± 3.87% at X = 0, Y = 0. Conclusion: The CI and HI values showed no discernible fluctuation, and the OAR doses were barely affected as smoothing was increased. When the smoothing factor was raised, the number of MUs sharply dropped, and a decrease in the number of segments and higher gamma passing rates were also seen.
Keywords: Complexity, fluence, gamma passing rate, monitor units, radiotherapy
|How to cite this article:|
Saroj DK, Yadav S, Paliwal N. Does fluence smoothing reduce the complexity of the intensity-modulated radiation therapy treatment plan? A dosimetric analysis. J Med Phys 2022;47:336-43
|How to cite this URL:|
Saroj DK, Yadav S, Paliwal N. Does fluence smoothing reduce the complexity of the intensity-modulated radiation therapy treatment plan? A dosimetric analysis. J Med Phys [serial online] 2022 [cited 2023 Jan 27];47:336-43. Available from: https://www.jmp.org.in/text.asp?2022/47/4/336/367430
| Introduction|| |
The most well-known fatal disease in the world is cancer. Uncontrolled cell proliferation is one of its key characteristics. According to published research, external beam radiation therapy is effective for 52% of all cancer patients. The employment of symmetric and asymmetric rectangular beams to treat tumors during the development of radiotherapy led to high-dosage areas within and around the tumor. High doses of radiation are also applied to nearby healthy tissue. This will violate the radiotherapy's goal to provide high radiation doses to the tumor and low or no doses to organs at risk (OARs). Much effort has been made to precise dose delivery to the tumor and blocking OARs as novel radiation treatment methods are being established and developed. For a great degree of conformance to be achieved, the tumor must be tightly fitted. To gain a more conformal treatment plan around the concave target structure, three-dimensional target shaping is needed, which can be attained with the help of the intensity-modulated radiation therapy (IMRT) technique. There are two common methods by which IMRT treatment plans are delivered: step-and-shoot IMRT and sliding window IMRT. The step-and-shoot IMRT technique uses a multi-leaf collimator (MLC) to form numerous static fields, which are then added together to produce variable fluence across the field. The MLC leaves are set (step) and a predetermined number of monitor units (MUs) are delivered for each static field (shoot). While the beam is on, MLC leaves continue to move when using the sliding window method. The gap, or window, between each opposing leaf pair, controls the dose.
To achieve the desired dose distribution inside the planning target volume (PTV) and a minimal dosage to the neighboring OARs, IMRT frequently uses inverse planning procedures with an optimization algorithm. In inverse planning procedure, first the planner decides the desired number of beams and beam angles and after that enters the clinical goal in the optimization window of the treatment planning system (TPS) for optimization. An inverse planning algorithm is used to create a large number of beamlets. The fluences of each beamlet are optimized in successive iterations to achieve the required dose distribution for PTVs and OARs. Mainly, IMRT treatment planning utilizes two different approaches for the optimization process. The first approach is the two-step optimization process where firstly, optimizations are performed on beamlet fluences to produce an “optimal fluence map” with the help of iterative reconstruction. After the fluence optimization, the leaf motion calculator (LMC) generates the MLC positions and accounts for physical and mechanical constraints on the MLC. Because of dosimetric and mechanical limitations on the MLC, such as transmission, maximum leaf speed, and leaf edge shape, LMC calculates an actual fluence map somewhat different from the “optimal fluence map.” The second approach is direct machine parameter optimization. It is the single-step optimization where MLC constraints and fluence optimization are taken care of together., The planner has a better command of the complexity of the treatment plan after optimization as no conversion of fluence takes place and no such criteria exist in two-step optimization. The intensity and magnitude of frequency changes in the beam's fluence distribution can be used to define the “complexity” of the fluence.,, The complexity of the fluence depends upon the shape of PTVs and OARs. More conformal plans and demand to achieve low doses of OARs can generate a complex fluence map. Reduction in complexity attributes to the degradation of treatment plan quality in terms of less conformal and higher doses to OARs. There will be more practical difficulty in delivering the greater complex treatment plan. A higher complex fluence map can result in a more number of MUs. Higher MUs will increase the risk of secondary cancer induction, longer treatment time, increased skin dose, patient alignment, and uncertainty in plan delivery. The optimization is statistical in nature and hence is more prone to noise and artifacts producing sharp peaks and valleys., The above problem can be solved by smoothing the fluence map. Intensity-modulated beam smoothing filters and including smoothness terms into the objective function of the optimization algorithm are the two most often used techniques for fluence map smoothing., The above-suggested methods generally exclude unwanted and random noise in the fluence map. Both strategies aim to preserve aspects of the entire fluence distribution while minimizing fluence disparities between adjacent beamlets. In the present study, we examined the effect of fluence smoothing on the treatment plan quality parameter, doses to OARs, number of MUs, beam on time (BOT), and the gamma index.
| Materials and Methods|| |
We randomly selected 15 patients with carcinoma cervix (Ca-Cx) from our hospital database for this study. All selected patients were previously treated in our institute with IMRT sliding window treatment technique using a linear accelerator (LINAC) Truebeam (M/s Varian Medical Systems, Palo Alto, CA, USA). The Truebeam LINAC consists of 6 MV, 10 MV, and 15 MV photon beam energy with a flattening filter device, while 6 MV and 10 MV photon beam energy with a flattening filter-free(FFF) device. The IMRT treatment plans were created using Eclipse TPS versus 13.7.23 (Varian Medical Systems, Palo Alto, California, USA) and a photon optimizer.
Image acquisition and target delineation
All patients were immobilized using a VacLok bag (Orfit Industry NV, Belgium) with the help of an all-in-one board (AIO, Orfit Industry NV, Belgium). All patients were simulated in the supine position with full bladder protocol. Knee rest was used depending on the patient comfort zone. The computed tomography (CT) scans were acquired for all patients with 3.0 mm slice thickness on the Siemens SOMATOM (Siemens Medical Systems, Germany) CT simulator. Two sets of axial CT scans were obtained for a patient with and without contrast. After image acquisition, Digital Imaging and Communications in Medicine images were transferred and loaded into TPS.
The gross tumor volume (GTV) was drawn around the noticeable tumor growth, including the cervix and corpus uteri. The clinical target volume (CTV) was created with a 5 mm margin to the GTV. CTV encompasses the external, internal, and common iliac and presacral lymph nodes. Another 5 mm of uniform expansion around the CTV will create the PTV. The OARs such as the bladder, rectum, femoral heads, and bowel were also contoured. The radiation oncologist performed target and OAR contouring.
All the patients taken for the study were planned for 50 Gy in 25 fractions with one fraction daily for 5 days per week. All plans were developed with a 6 MV photon beam and millennium 120 MLCs (MLC-120). MLC leaves consist of 60 leaf pairs; the width of the inner 40-leaf pair and outer 20-leaf pair at the isocenter is 0.5 cm and 1.0 cm, respectively. These treatment plans were created using nine equally spaced static beam angles from 0° to 360°. Co-planar treatment fields were used during treatment planning. In the Varian Eclipse TPS, optimal fluence smoothing is integrated into the optimizer. Smoothing is applied across the X and Y direction of MLC movement, which is parallel and orthogonal to the law of MLC movement. Significant fluence differences between neighboring pixels in the X and Y directions are penalized by assigning weighing factors.
All the original treatment plans were developed using smoothing parameters of X = 90 and Y = 90 (institutional protocol), and each of these plans was verified and approved for patient treatment. Additional plans were created by varying the smoothing parameter from X = 0, Y = 0 to X = 100, Y = 100 by keeping other optimization parameters the same as the reference plan. Smoothing can be varied from 0 to 999, but for this study, we concentrated on the range of up to 100. Generally, the vendor provides default asymmetric smoothing fluence penalty X=40 and Y=30 in the x and y-directions respectively because more smoothing is required in the x-direction to minimize MUs as compared to y-directions. For each patient, eight plans with different smoothing penalties in X and Y directions like X = 0, Y = 0 (minimum level); X = 20, Y = 20; X = 40, Y = 30 (vendor-default); X = 40, Y = 40; X = 60, Y = 60; X = 80, Y = 80; X = 90, Y = 90; and X = 100, Y = 100 (maximum level) were created and optimized for 200 iterations. An anisotropic analytical algorithm was used for the final dose calculation with a grid size of 0.25 cm. To avoid utilizing diverse planning methodologies, only one medical physicist generated each treatment plan in this study.
Plan evaluation and statistical methods
Plan evaluation and comparison were performed based on the dose–volume histogram (DVH). All plans were created to deliver 95% of the prescription dose (PD) to 95% of PTV volume. The volume receiving 95% of the prescribed dose (V95%) and the dose received by 98% of the volume (D98%) for PTV were recorded. The mean doses for the bladder, rectum, and bowel were recorded. Plans were optimized to obtain the dose constraint of V40%>45 Gy, V40%>55 Gy, and V45 Gy >180 cm3 for the bladder, rectum, and bowel, respectively, where V40% is the volume receiving the 40% of the prescribed dose and V45 Gy is the volume of the isodose of 45 Gy. Integral doses were calculated for the OARs, and gamma passing rates were calculated and compared.
Treatment plan quality indices
According to the International Commission on Radiation Units and Measurements report 83, the following dosimetric plan quality indices were calculated.
The conformity index (CI) is defined as the prescription isodose volume (VRI) divided by the total PTV volume. The recommended value is unity, but it is usually < 1.
CI = VRI/TV
The homogeneity index (HI) was calculated as the difference between the delivered dose to 2% (D2%) and 98% (D98%) of PTV volume divided by dose to 50% of PTV volume (D50%). It is defined as follows:
HI = (D2%–D98%)/D50%
HI = 0 implies a completely homogeneous dose distribution.
A ratio evaluates the PD coverage in treatment plans. It is defined as follows:
Conformation number (CN) = 1 indicates a perfect plan.
CN = VRI/TV × VRI/VP
where VRI (reference isodose volume) = the volume encompassed by 95% of prescription isodose lines.
Here in our study, we have chosen 95% isodose lines as isodose reference lines.
TV (target volume) = PTV volume, VP = Volume of the body covered by the prescription isodose.
Additionally, MUs for each treatment plan were also recorded for comparison.
The dose delivery is affected by fluence complexity. In our study, we also studied the effect of fluence smoothing on gamma passing rates. The electronic portal image detection (EPID) technology in Eclipse TPS was used to calculate the two-dimensional (2D) dose distributions for each plan, and the associated measured dose distributions were compared.
The high-resolution a-Si1000 EPID, which was used in this investigation for patient setup verification, features arrays of light-sensitive amorphous-Si photodiodes arrayed in a 40 × 30 cm2 active detector area with 1024 × 768 pixels and a resolution of 0.390 mm. The a-Si 1000 portal imaging system was employed in this study, which is an indirect detection system consisting of a 1 mm copper plate above a scintillating layer of phosphor (Gd2O2S: Tb, 70 mg/cm2) that converts incident radiation into optical photons. The measured plan with EPID is matched with the calculated 2D fluence map. The source-to-detector distance was 100 cm. The gamma index analysis was performed to compare the calculated and delivered fluence. For all dynamic IMRT plans, a quantitative evaluation of the measured against the TPS calculated doses was done using the gamma index (percentage dose difference and distance to agreement [DTA]). For dose difference analysis, the standard passing threshold is set at 3%, while for DTA analysis; it is set at 3 mm. We computed and compared the gamma values mean and standard deviation. This will assist in identifying the smoothing values that indicate a higher level of deliverability.
Comparisons among different levels of fluence smoothing on PTV and normal organ doses were analyzed with a one-way analysis of variance test. The post hoc Turkey test was utilized to identify which pair-wise comparisons varied when an overall significant difference was observed. All statistical analysis was conducted with the Statistical Package for Social Sciences (SPSS) software (release 20.0, SPSS Inc., Chicago, IL, USA). The difference was considered statistically significant when P < 0.05.
| Results|| |
All treatment plans achieved our dosimetric and clinical plan acceptability criteria. The patient's characteristics are listed in [Table 1]. The mean age of the patients was 59.53 ± 11.39 years (31–71 years). The structure volume was measured in cubic centimeters (cm3). The mean PTV volume was 849.18 ± 151.15.56 cm3 (ranging from 519.5 cm3 to 1136.6 cm3). The mean bladder and rectum volumes were 244.66 ± 96.12 cm3 (ranges from 124.5 cm3 to 466.1 cm3) and 61.45 ± 19.59 cm3 (ranges from 39.2 cm3 to 105.9 cm3), respectively. The mean bowel volume was 1693.73 ± 426.40 cm3 (ranging from 1025.7 cm3 to 2288.7 cm3).
Target coverage versus various smoothing levels
All dose–volume parameters for PTV at different smoothing levels are tabulated in [Table 2]. The V95% of PTV has a maximum value of 868.08 ± 179.80 cm3 and minimum value of 838.22 ± 183.04 cm3 at smoothing levels of X = 40, Y = 30 and X = 100, Y = 100, respectively. D95% of PTV had a maximum of 4871.21 ± 20.99 Gy on varying X = 0, Y = 0 to X = 100, Y = 100 penalty to fluence map. D98% of PTV has a maximum and minimum value of 4833.93 ± 26.99cGy and 4782.69 ± 44.51 Gy, respectively (P = 0.010). The maximum average dose to PTV was found to be 5009.46 ± 19.88 cGy at smoothing levels of X = 100, Y = 100 (P = 0.028).
|Table 2: The dosimetric parameter for planning target volume for various smoothing levels|
Click here to view
The treatment plan quality indices for different smoothing levels are described in [Table 3]. The CI value achieved the highest value of 0.996 ± 0.0038 at X = 60, Y = 60 fluence map penalty. HI was found maximum at X = 0, Y = 0 with the value of 0.0628 ± 0.0235. Significant variation was seen in CN value by varying smoothing levels across the fluence map from no penalty (X = 0, Y = 0) to maximum penalty (X = 100, Y = 100). CN attained the highest value of 0.870 ± 0.027 at X = 0, Y = 0 and a minimum value of 0.734 ± 0.0043 at X = 80, Y = 80. On increasing the smoothing value across the fluence map, MUs significantly reduced (P = 0.0). Maximum and minimum MU values were 2424.30 ± 471.12 and 1087.80 ± 91.57 X = 0, Y = 0 and X = 100, Y = 100, respectively. The gamma passing rate has its maximum value of 99.28 ± 0.44 at X = 100, Y = 100. The minimum value of gamma passing rates is observed at X = 0, Y = 0.
|Table 3: The variation of conformity index, homogeneity index, conformation number, and monitor units with smoothing levels|
Click here to view
[Table 4] shows the variation of smoothing levels on various dosimetric parameters of the OARs. No significant dose variation was seen in the mean dose to the bladder on the varying smoothing levels. The lowest mean dose to the bladder was 34.02 ± 2.70 Gy. V40 Gy <40% for bladder was found the least value of 38.25% at X = 0, Y = 40 and a maximum value of 46.04% at X = 100, Y = 100. The lowest mean dose to the rectum was 37.23 ± 2.91 Gy achieved at smoothing levels of X = 40, Y = 30. The highest mean dose to the rectum was 34.27 ± 2.90 Gy achieved at smoothing levels of X = 100, Y = 100. V40 Gy of the rectum had a low value of 41.90% at X = 0, Y = 0 and a peak value of 55.17% at X = 100, Y = 100. Bowel achieved a minimum mean dose of 20.08 ± 2.71 Gy and a maximum mean dose of 23.25 ± 3.02 Gy at X = 40, Y = 30 and X = 100, Y = 100, respectively. V45 Gy of bowel gets a lower volume of 118.38 ± 50.84 cm3 at X = 0, Y = 0 and has a maximum volume of 174.55 ± 74.81 cm3 at X = 100, Y = 100. At the smoothing level of X = 100 and Y = 100, the maximum point dose was observed 47.81± 1.56 Gy and 47.96 ±1.78 Gy for the right and left femoral heads respectively.
|Table 4: Effect of smoothing levels on dose-volume parameters for organs at risks|
Click here to view
| Discussion|| |
The goal of our study is to identify the dosimetric effect of different smoothing levels on PTV and OARs and to compare it with the vendor-specified default smoothing levels. The entire plan met our dosimetric goal and achieved the desired clinical aspects of the planning. [Figure 1] shows the axial view of color dose wash of 95% of the prescribed dose for the PTV with different smoothing levels. Referring to the DVH shown in [Figure 1], there was a little variation on V95%, and D95% on varying the smoothing levels. Dmean for the PTV was 0.6% higher at X = 60 and Y = 60 than the vendor's default smoothing levels of X = 40 and Y = 30. A small increment of D98% for PTV was observed on increasing the smoothing penalty till the smoothing levels of X = 60 and Y = 60, after that the values go down and achieved their lowest at X = 100 and Y = 100. There was an unexpected decrease of 0.7% in D98% of PTV as compared to X = 40 and Y = 30 than X = 100 and Y = 100. Armoogum has performed similar studies for prostate and head-and-neck sites, where he concluded that increasing the smoothing levels shows a little variation in the HI and CI of the target. In our study, maximum CI values were achieved at smoothing levels of X = 60 and Y = 60, which was only 0.08% higher than vendor-recommended settings. Furthermore, there was only a 0.77% increase in CI value with no smoothing levels to maximum smoothing levels. [Figure 2]a shows the variation of CI with different levels of fluence smoothing. Treatment plans were found to be more homogeneous at no smoothing levels, and plans were found to be 19% less homogeneous at maximum smoothing levels than at no smoothing. Compared to the vendor-recommended smoothing levels, no smoothing levels were 1.5% more homogeneous. [Figure 2]b shows the HI and CN variation with different smoothing levels. The CN value deteriorated with increasing the smoothing levels with little variation until the vendor-recommended smoothing levels. After that, it changed rapidly and achieved the lowest value at maximum smoothing levels of X = 100, Y = 100. CN value changes to 15% less than the value at no smoothing levels, indicating that there will be more and more isodose lines spilling out of the PTV contours as the smoothing levels increase.
|Figure 1: The different levels of smoothing penalty applied to the fluences: (a) 0 × 0, (b) 20 × 20, (c) 40 × 30, (d) 40 × 40, (e) 60 × 60, (f) 80 × 80, (g) 100 × 100, and (h) the c-DVH of the PTV for one of the patients. Figures A to G also show the contour of bowel and rectum in orange and purple color. c-DVH: Cumulative dose–volume histogram, PTV: Planning target volume|
Click here to view
|Figure 2: (a) The variation of CI with different levels of fluence smoothing. (b) The variation of HI and CN. CI: Conformity index, HI: Homogeneity index, CN: Conformation number|
Click here to view
[Figure 3] shows the variation of DVH for the bladder and the rectum for different levels of smoothing applied. The mean dose to the bladder was at its lowest value at a vendor-recommended setting of smoothing levels. On increasing the smoothing levels, there was an increase in the mean dose of the bladder and a 6.3% of mean dose difference was observed for the bladder between smoothing levels (X=40 and Y=30) and (X=100 and Y=100).
|Figure 3: The c-DVH for bladder and rectum and its variation with various smoothing levels. c-DVH: Cumulative dose–volume histogram|
Click here to view
The V40Gy of the bladder received 46.04% of the prescribed dose at a smoothing level (X=100 and Y=100) which comparatively gets a lower value of 40.01% at a smoothing level (X=40 and Y=30).
The mean dose and V40 Gy of the rectum show an increase of 8.48% and 26.97% on the increment of smoothing level from the default setting of smoothing level to maximum smoothing level of X = 100 and Y = 100, respectively. Doses to bowel like other OARs show similar trends of increased in dose on varying the smoothing level from X = 0, Y = 0 to X = 100, Y = 100. The mean dose and V45 Gy of bowel show 15.8% and 47% higher doses at maximum smoothing level compared to the default smoothing setting, respectively. Dmax of the femoral head shows marginal increment (2.6% and 0.9%) in dose for the left and right femoral head on comparing default smoothing setting and X = 100 and Y = 100. Though the doses to OARs increased marginally with the increases in smoothing levels, their dose constraints were not violated. The possible reason for an increase in OARs doses is the combination of dose penalty and smoothing level used in the inverse treatment planning. To avoid any kind of biasing all plans were re-optimized for similar dose penalty combinations in the inverse optimizer. It is anticipated that different dosage penalty combinations for varying levels of smoothing will be employed to attain the desired dose volume restrictions.
The analysis of 120 individually optimized plans showed that the number of maximum, minimum, and average MUs for both patient groups reduced as smoothing was increased. The concept “MUs factor,” which is connected to the complexity of the plan, is used to calculate MUs. To attain a similar dose distribution as that for large fields, a small field needs more MUs. Several small and asymmetric-shaped beam segments are necessary to attain optimal conformity for large-scale modulations in complex IMRT schemes. Thus, large MUs indirectly point to the high level of complexity in the IMRT plan. Smoothing and plan complexity are mutually exclusive, and any fall in fluence complexity is strongly connected with an increase in MUs. [Figure 4] shows the variation in number of variation in MUs with different levels of smoothing. Our finding for MUs with increasing smoothing levels for Ca-Cx was similar to previous studies by various authors for the different sites.,,, The mean MUs for all the plans decreased on increasing the smoothing levels from 2424.30 ± 471.12 at X = 0, Y = 0 to 1087.80 ± 91.57 at X = 100, Y = 100. It was noted that 2.2 times fewer MUs were delivered at high smoothing levels compared to no smoothing. Furthermore, the MUs were 1.6 times higher at vendor-recommended smoothing levels compared to high smoothing levels.
|Figure 4: The MU variation with different smoothing levels. MUs: Monitor units|
Click here to view
Another important aspect of the IMRT treatment plan is patient-specific quality assurance. The fluence generated from the treatment plan must be evaluated and analyzed before delivering to the patient. The 2D fluence analysis is a good tool to check the deliverability of the treatment plan generated from TPS. The findings of a 2D dose comparison reveal a significant increase in the percentage of points meeting the passing standard about the increasing fluence smoothing. In Eclipse, low smoothing parameters make the fluences appear more complicated, and as the complexity of the plan increased, the gamma passing rate reduced. Various studies have reported on the effect of fluence smoothing factors in inverse IMRT treatment planning. In a study by Armoogum, the effects of fluence smoothing were examined for a group of prostate and head-and-neck cancer patients. They have concluded in their studies that because of the decreased complexity of the plan, increasing smoothing causes a considerable decrease in MUs and a clear gain in average leaf pair opening (LPO). We also observed similar results of increased gamma passing rate with increased fluence smoothing level. This is also illustrated by [Figure 1], where the effect of increasing smoothing levels on the fluence can be seen. Increment in smoothing levels from low levels to high levels smoothened the fluence by increasing the average LPO and reducing the MUs.
[Figure 5] shows the variation of gamma passing rates with different fluence smoothing settings. Anker et al. studied the behavior of three inverse planning systems when smoothing parameters were varied and concluded that depending on the inverse TPS being used, the potential benefits of optimizing fluence smoothing levels can be significant, allowing for increases in either delivery efficiency or plan conformity. Our finding showed similar trends, where below the X = 40 and Y = 40 levels, it was observed that the gamma passing rates were <95%. The gamma passing rates shows a significant increase in the gamma value above the smoothing level of X=60 and Y=60. This increase in the gamma passing rates was expected due to a decrease in MUs, and a reduction in MUs is related to the decline in the treatment plan complexity. Additionally, the comparison of treatment techniques, LINACs, TPS, and operators are made possible by complexity measures. They have proven useful in standardizing and rationalizing the optimization process in conjunction with plan quality metrics.
|Figure 5: The gamma passing rate variation with different smoothing levels|
Click here to view
It is believed that the less complex plan is the requirement for a good IMRT treatment plan. Our investigation into the impact of smoothing levels on IMRT plans led us to the conclusion that, in order to achieve the best IMRT treatment plan, different smoothing levels beyond the default setting must be utilized. Optimum structure dose priority along with smoothing level must be used to achieve the conformal treatment plan. Increasing the smoothing levels reduces the MUs and smoothened the fluence and makes the treatment plan easily deliverable on the LINAC.
| Conclusion|| |
We could thoroughly comprehend the impact of fluence smoothing with the Eclipse TPS through the analysis of Ca-Cx IMRT treatment plans with various fluence-level scenarios. This study showed an insignificant variation of the dosimetric parameter on PTV and OARs. Increasing fluence smoothing still gives the dosimetrically and clinically acceptable treatment plan. Lesser MUs will help reduce the scattered dose and indirectly reduce the probability of secondary malignancy, less BOT, and more patient comfort. Lesser MUs are related to less complex treatment plans, reflecting the higher gamma passing rates. Although there was an insignificant increase in the OAR doses with increasing smoothing levels, those are well below their tolerance level. Our study recommends keeping smoothing levels of at least X = 80 and Y = 80 for the Ca-Cx IMRT treatment plan to have optimum treatment plan doses to PTV and OARs and fewer MUs.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Delaney G, Jacob S, Featherstone C, Barton M. The role of radiotherapy in cancer treatment: Estimating optimal utilization from a review of evidence-based clinical guidelines. Cancer 2005;104:1129-37.
Bakiu E, Telhaj E, Kozma E, Ruçi F, Malkaj P. Comparison of 3D CRT and IMRT tratment plans. Acta Inform Med 2013;21:211-2.
Nutting C, Dearnaley DP, Webb S. Intensity modulated radiation therapy: A clinical review. Br J Radiol 2000;73:459-69.
Webb S. The physical basis of IMRT and inverse planning. Br J Radiol 2003;76:678-89.
Varian Medical Systems. External beam planning reference guide eclipse. Palo Alto, CA: Varian Medical Systems; 2007.
Dobler B, Koelbl O, Bogner L, Pohl F. Direct machine parameter optimization for intensity modulated radiation therapy (IMRT) of oropharyngeal cancer – A planning study. J Appl Clin Med Phys 2009;10:4-15.
Broderick M, Leech M, Coffey M. Direct aperture optimization as a means of reducing the complexity of Intensity Modulated Radiation Therapy plans. Radiat Oncol 2009;4:8.
Mohan R, Arnfield M, Tong S, Wu Q, Siebers J. The impact of fluctuations in intensity patterns on the number of monitor units and the quality and accuracy of intensity modulated radiotherapy. Med Phys 2000;27:1226-37.
Xing L, Li JG. Computer verification of fluence map for intensity modulated radiation therapy. Med Phys 2000;27:2084-92.
Llacer J, Solberg TD, Promberger C. Comparative behaviour of the dynamically penalized likelihood algorithm in inverse radiation therapy planning. Phys Med Biol 2001;46:2637-63.
Hall EJ, Wuu CS. Radiation-induced second cancers: The impact of 3D-CRT and IMRT. Int J Radiat Oncol Biol Phys 2003;56:83-8.
Giorgia N, Antonella F, Eugenio V, Alessandro C, Filippo A, Luca C. What is an acceptably smoothed fluence? Dosimetric and delivery considerations for dynamic sliding window IMRT. Radiat Oncol 2007;2:42.
Du W, Cho SH, Zhang X, Hoffman KE, Kudchadker RJ. Quantification of beam complexity in intensity-modulated radiation therapy treatment plans. Med Phys 2014;41:021716.
Sun X, Xia P. A new smoothing procedure to reduce delivery segments for static MLC-based IMRT planning. Med Phys 2004;31:1158-65.
Matuszak MM, Larsen EW, Fraass BA. Reductions of IMRT beam complexity through the use of beam modulation penalties in the objective function. Med Phys 2007;34:507-20.
Hodapp N. Der ICRU-Report 83: Verordnung, Dokumentation und Kommunikation der fluenzmodulierten Photonenstrahlentherapie (IMRT) [The ICRU Report 83: prescribing, recording and reporting photon-beam intensity-modulated radiation therapy (IMRT)]. Strahlenther Onkol. 2012;188:97-9. German. doi: 10.1007/s00066-011-0015-x.
van't Riet A, Mak AC, Moerland MA, Elders LH, van der Zee W. A conformation number to quantify the degree of conformality in brachytherapy and external beam irradiation: Application to the prostate. Int J Radiat Oncol Biol Phys 1997;37:731-6.
Niyas P, Abdullah KK, Noufal MP, Sankaran Nair T. Effect of fluence smoothing on the quality of intensity-modulated radiation treatment plans. Radiol Phys Technol 2016;9:202-13.
Stasi M, Giordanengo S, Cirio R, Boriano A, Bourhaleb F, Cornelius I, et al.
D-IMRT verification with a 2D pixel ionization chamber: Dosimetric and clinical results in head and neck cancer. Phys Med Biol 2005;50:4681-94.
Low DA, Harms WB, Mutic S, Purdy JA. A technique for the quantitative evaluation of dose distributions. Med Phys 1998;25:656-61.
Nelms BE, Simon JA. A survey on planar IMRT QA analysis. J Appl Clin Med Phys 2007;8:76-90.
Armoogum KS. Effect of smoothing on treatment plan efficiency in IMRT: Eclipse HeliosTM
dose optimisation. J Radiother Pract 2012;11:229-38.
Spirou SV, Fournier-Bidoz N, Yang J, Chui CS, Ling CC. Smoothing intensity-modulated beam profiles to improve the efficiency of delivery. Med Phys 2001;28:2105-12.
Anker CJ, Wang B, Tobler M, Chapek J, Shrieve DC, Hitchcock YJ, et al.
Evaluation of fluence-smoothing feature for three IMRT planning systems. J Appl Clin Med Phys 2010;11:3035.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
[Table 1], [Table 2], [Table 3], [Table 4]