Multileaf Collimator Research Articles

Dosimetric Analysis of Intra-Fraction Motion Detected by Surface-Guided Radiation Therapy During Linac Stereotactic Radiosurgery

Stereotactic radiosurgery (SRS) immobilization with an open face mask is more comfortable and less invasive than frame based, but concerns about intrafraction motion must be addressed. Surface-guided radiation therapy (SGRT) is an attractive option for intrafraction patient monitoring because it is continuous, has submillimeter accuracy, and uses no ionizing radiation. The purpose of this study was to investigate the dosimetric consequences of uncorrected intrafraction patient motion detected during frameless linac-based SRS. Fifty-five SRS patients were monitored during treatment using SGRT between January 1, 2017, and September 30, 2020. If SGRT detected motion >1 mm, imaging was repeated and the necessary shifts were made before continuing treatment. For the 25 patients with intrafraction 3-dimensional vector shifts of ≥1 mm, we moved the isocenter in the planning system using the translational shifts from the repeat imaging and recalculated the plans to determine the dosimetric effect of the shifts. Planning target volume (PTV) coverage, minimum gross tumor volume (GTV) dose (relative and absolute), and normal brain V12 were evaluated. Wilcoxon signed rank tests were used to compare planned and simulated dosimetric parameters and median 2 sample tests were used to investigate these differences between cone and multileaf collimator (MLC) plans. For simulated plans, V12 increased by a median of 0.01 cc (P=.006) and relative GTV minimum dose and PTV coverage decreased by a median of 15.8% (P < .001) and 10.2 % (P < .001), respectively. Absolute minimum GTV dose was found to be significantly lower in the simulated plans (P < .001). PTV coverage decreased more for simulated cone plans than for simulated MLC plans (11.6% vs 4.7%, P=.011) but median V12 differences were found to be significantly larger for MLC plans (-0.34 cc vs -0.01 cc, P=.011). Differences in GTV minimum dose between cone and MLC plans were not statistically significant. SGRT detected clinically meaningful intrafraction motion during frameless SRS, which could lead to large underdoses and increased normal brain dose if uncorrected.

Concurrent definitive chemoradiation incorporating intensity-modulated radiotherapy followed by adjuvant chemotherapy in high risk locally advanced cervical squamous cancer: a phase II study

BackgroundAlthough the prognosis of locally advanced cervical cancer has improved dramatically, survival for those with stage IIIB-IVA disease or lymph nodes metastasis remains poor. It is believed that the incorporation of intensity-modulated radiotherapy into the treatment of cervical cancer might yield an improved loco-regional control, whereas more cycles of more potent chemotherapy after the completion of concurrent chemotherapy was associated with a diminished distant metastasis. We therefore initiated a non-randomized prospective phaseII study to evaluate the feasibility of incorporating both these two treatment modality into the treatment of high risk locally advanced cervical cancer.ObjectivesTo determine whether the incorporation of intensity-modulated radiotherapy and the addition of adjuvant paclitaxel plus cisplatin regimen into the treatment policy for patients with high risk locally advanced cervical cancer might improve their oncologic outcomes.Study designPatients were enrolled if they had biopsy proven stage IIIA-IVA squamous cervical cancer or stage IIB disease with metastatic regional nodes. Intensity-modulated radiotherapy was delivered with dynamic multi-leaf collimators using 6MV photon beams. Prescription for PTV ranged from 45.0 ~ 50.0 Gy at 1.8 Gy ~ 2.0 Gy/fraction in 25 fractions. Enlarged nodes were contoured separately and PTV-nodes were boosted simultaneously to a total dose of 50.0–65 Gy at 2.0- 2.6 Gy/fraction in 25 fractions. A total dose of 28 ~ 35 Gy high-dose- rate brachytherapy was prescribed to point A in 4 ~ 5 weekly fractions using an iridium- 192 source. Concurrent weekly intravenous cisplatin at 30 mg/m2 was initiated on the first day of radiotherapy for over 1-h during external-beam radiotherapy. Adjuvant chemotherapy was scheduled within 4 weeks after the completion of concurrent chemo-radiotherapy and repeated 3 weeks later. Paclitaxel 150 mg/m2 was given as a 3-h infusion on day1, followed by cisplatin 35 mg/m2 with 1-h infusion on day1-2 (70 mg/m2 in total).ResultsFifty patients achieved complete response 4 weeks after the completion of the treatment protocol, whereas 2 patients had persistent disease. After a median follow-up period of 66 months, loco-regional (including 2 persistent disease), distant, and synchronous treatment failure occurred in 4,5, and 1, respectively. The 5-year disease-free survival, loco-regional recurrence-free survival, distant-metastasis recurrence-free survival was 80.5%, 90.3%, and 88.0%, respectively. Four of the patients died of the disease, and the 5-year overall survival was 92.1%. Most of the toxicities reported during concurrent chemo-radiotherapy were mild and transient. The occurrence of hematological toxicities elevated mildly during adjuvant chemotherapy, as 32% (16/50) and 4% (2/50) patients experienced grade 3–4 leukopenia and thrombocytopenia, respectively. Grade 3–4 late toxicities were reported in 3 patients.ConclusionsThe incorporation of intensity-modulated radiotherapy and adjuvant paclitaxel plus cisplatin chemotherapy were highly effective and well-tolerated in the treatment of high-risk locally advanced cervical cancer. The former yields an improved loco-regional control, whereas distant metastases could be effectively eradicated with mild toxicities when adjuvant regimen was prescribed.

Predicted Inferior Outcomes for Lung SBRT With Treatment Planning Systems That Fail Independent Phantom-Based Audits

More than 15% of radiotherapy clinics fail external audits with anthropomorphic phantoms conducted by Imaging and Radiation Oncology Core-Houston (IROC-H) while passing other industry-standard quality assurance (QA) tests. We seek to evaluate the predicted impact of such failed plans on outcomes for patients treated with stereotactic body radiation therapy (SBRT) for lung tumors. We conducted a retrospective study of 55 patients treated with SBRT for lung tumors with prescription BED10 ≥100Gy using a treatment planning system (TPS) that passed IROC-H phantom audits. Sample linac beam models with introduced errors were commissioned by varying the multi-leaf collimator leaf tip offset parameter (i.e., 'dosimetric leaf gap') over the range +/-1.0 mm relative to the validated model. These models mimic TPS that pass internal QA measures but fail IROC-H tests. Patient plans were recalculated on sample beam models. Predicted tumor control probability (TCP) and normal tissue complication probability (NTCP) were calculated based on published dose-response models. A leaf tip offset value of -1.0 mm decreased the fraction of plans receiving a PTV BED10 ≥100 Gy from 95% to 27%. This translated to a significant decrease in 2-yr TCP of 4.8% (95% CI: 2.0-5.5%) with a decrease in TCP up to 21%. Conversely, leaf tip offset of +1.0 mm resulted in 36% of patients exceeding previously met OAR constraints, including two instances of spinal cord and brachial plexus overdoses and a small increase in chest wall NTCP by 0.7%, (95% CI 0.5%-0.8%). Simulated treatment plans with modest MLC leaf offsets result in lung SBRT plans that significantly under-dose tumor or exceed OAR constraints. These dosimetric endpoints translate to significant detriments in TCP. These simulated plans mimic planning systems that pass internal QA measures but fail independent phantom-based tests, underscoring the need for enhanced quality assurance including external audits of TPS.

Evaluation of measures related to dosimetric uncertainty of VMAT plans.

Dosimetric uncertainty is most often not included in the process of creating and selecting plans for treatment. Treatment planning and the physician's choice of treatment plan is instead often based only on evaluation of clinical goals of the calculated absorbed dose distribution. Estimation of the dosimetric uncertainty could potentially have impact in the process of comparing and selecting volumetric modulated arc therapy (VMAT) plans. In this study, different measures for estimation of dosimetric uncertainty based on treatment plan parameters for plans with similar dose distributions were evaluated. VMAT plans with similar dose distributions but with different treatment plan designs were created using three different optimization methods for each of ten patient cases (tonsil and prostate cancer). Two plans were optimized in Eclipse, one with and one without the use of aperture shape controller, and one plan was optimized in RayStation. The studied measures related to dosimetric uncertainty of treatment plans were aperture-based complexity metric analysis, investigation of modulation level of multi leaf collimator leaves, gantry speed and dose rate, quasi-3D measurements and evaluation of simulations of realistic delivery variations. The results showed that there can be variations in dosimetric uncertainty for treatment plans with similar dose distributions. Dosimetric uncertainty assessment could therefore have impact on the choice of plan to be used for treatment and lead to a decrease in the uncertainty level of the delivered absorbed dose distribution. This study showed that aperture shape complexity had a larger impact on dosimetric uncertainty compared to modulation level of MLC, gantry speed and dose rate.

Machine learning-based predictions of gamma passing rates for virtual specific-plan verification based on modulation maps, monitor unit profiles, and composite dose images

Machine learning (ML) methods have been implemented in radiotherapy to aid virtual specific-plan verification protocols, predicting gamma passing rates (GPR) based on calculated modulation complexity metrics because of their direct relation to dose deliverability. Nevertheless, these metrics might not comprehensively represent the modulation complexity, and automatically extracted features from alternative predictors associated with modulation complexity are needed. For this reason, three convolutional neural networks (CNN) based models were trained to predict GPR values (regression and classification), using respectively three predictors: (1) the modulation maps (MM) from the multi-leaf collimator, (2) the relative monitor units per control point profile (MUcp), and (3) the composite dose image (CDI) used for portal dosimetry, from 1024 anonymized prostate plans. The models’ performance was assessed for classification and regression by the area under the receiver operator characteristic curve (AUC_ROC) and Spearman’s correlation coefficient (r). Finally, four hybrid models were designed using all possible combinations of the three predictors. The prediction performance for the CNN-models using single predictors (MM, MUcp, and CDI) were AUC_ROC = 0.84 ± 0.03, 0.77 ± 0.07, 0.75 ± 0.04, and r = 0.6, 0.5, 0.7. Contrastingly, the hybrid models (MM + MUcp, MM + CDI, MUcp+CDI, MM + MUcp+CDI) performance were AUC_ROC = 0.94 ± 0.03, 0.85 ± 0.06, 0.89 ± 0.06, 0.91 ± 0.03, and r = 0.7, 0.5, 0.6, 0.7. The MP, MUcp, and CDI are suitable predictors for dose deliverability models implementing ML methods. Additionally, hybrid models are susceptible to improving their prediction performance, including two or more input predictors.

Craniospinal Irradiation: A Dosimetric Comparison Between O-Ring Linac and Conventional C-arm Linac

PurposeThe aim of this study was to perform a dosimetric evaluation between craniospinal irradiation volumetric modulated arc therapy plans designed for an O-Ring and a conventional C-arm Linac. Methods and MaterialsTwo adult patients were selected for this study. Two plans were designed one for a TrueBeam Edge and one for Halcyon O-ring Linac for each patient. The evaluation of the plans was conducted in terms of dose volume histogram analysis of the target volume and organs at risk (OARs) along with total plan monitor units and beam-on time. Paired sample t test was performed to compare Dmax and Dmean of OARs for each plan's comparison. The delivery of volumetric modulated arc therapy plans was evaluated using Octavius 4D phantom. ResultsAll plans demonstrated dose distributions with sufficient planned target volume conformity and homogeneity. The Homogeneity Index and Conformity Index for all plans were found comparable. The paired sample t test did not demonstrate significant difference in terms of Dmax and Dmean of OARs between plans for both patients. All plans met the threshold of 90%, with Halcyon plans having higher gamma passing rates. ConclusionsCraniospinal irradiation plans generated for Halcyon and Edge are equivalent in terms of plan quality and dose sparing at OARs. The small variations may have originated from the differences in beam profile or mean energy, the degree of the modulation for each plan and characteristics of multi leaf collimators for each treatment unit. Halcyon is able to deliver a distinctly faster treatment, but Edge provides an automatic rotational correction for patient positioning.

Intensity-modulated radiation therapy - you can have your cake and eat it too!

Arguably, intensity-modulated radiotherapy (IMRT) is one of the most important contributions by physicists in medicine to the treatment of cancer. It enabled the achievement of dose distributions that allowed treatment of targets to high doses and yet spared normal tissues to degrees previously considered virtually impossible. The concept underlying IMRT for photons, originally called "inverse planning," was introduced by Anders Brahme in the early 1980s. Since then, many physicists have been conducting research to advance the state of the art, to overcome obstacles discovered and to develop tools and techniques to translate it clinically. IMRT was first implemented clinically the early 1990's. Soon thereafter, it was also incorporated into intensity modulated particle therapy (IMPT), initially for protons and eventually for ions heavier than protons. In IMRT, intensities of small segments or beamlets of multiple incident beams are optimized to produce the best approximation of desired dose distributions that deliver the requisite tumor dose and maximally spare normal tissues. Such dose distributions are delivered using dynamic multi-leaf collimators whose leaves move continuously while the radiation is on or in a step-and-shoot fashion. For IMPT, there is an extra degree of freedom, that of energies of particle beamlets, which are part of the optimization process. IMPT dose distributions allow substantial additional sparing of normal tissues. They are delivered using scanning beams. IMRT and IMPT represents a true paradigm shift in the way cancer patients are treated. Their evolution, especially that of IMPT, is still ongoing, with new advancements being added continually. With the improving understanding of the immunomodulatory effects of radiotherapy, it is quite plausible that a combination of IMPT and immunotherapy will mature into one of the most effective ways of treating not just localized cancers but also systemic disease.

Towards 2D dosimetry using monolithic active pixel sensors and a copper grating

Higher energy and intensity X-ray radiotherapy treatments are coming into wider use, having the benefit of requiring fewer treatment fractions and fewer hospital visits per patient. However, small percentage errors in multileaf collimator positioning and dose become bigger problems with higher doses per fraction. Hence, real-time treatment verification becomes essential. Where devices downstream from the patient suffer from scattering in the patient, upstream devices can disturb the therapeutic beam. Here, a method is proposed for performing dosimetry upstream using monolithic active pixel sensors, which can be made thin enough to disturb the beam by <1%. In order to calculate the dose to the tumour, a verification device needs to make a measurement of the photon field. Some photons will Compton scatter an electron in the silicon and generate a signal. However, this signal is obscured by energy deposits from contamination electrons, originating from Compton scattering in the accelerator head and air. Often extensive build-up material is added to verification devices to reduce the electron contamination and enhance the photon signal. However, this leads to degradation of the beam intensity to the patient. Instead we propose using thin strips of 50 µm thick copper in a grating pattern and measuring the difference in the signal with and without it. The contamination electrons are relatively undisturbed by the presence of the thin copper strips and the photon signal generated via Compton scattering is enhanced under the strips. Hence the difference in the two signals mostly consists of energy deposits originating from the therapeutic photons. From this the dose to the patient can be derived. The 50 μm of copper is thin enough to keep the beam attenuation below 1%, but in itself would give a total signal response which consists of 38% contamination electrons. Using the grating technique, we show that the electron contamination can be reduced to 2.6% of the total signal. This allows the photon signal only to be extracted from the data and thus the dose to patient with a very thin upstream detector can be calculated using a Monte Carlo model to extrapolate the photon flux into the tumour.

Patient-specific quality assurance prediction models based on machine learning for novel dual-layered MLC linac.

Patient-specific quality assurance (PSQA) is an indispensable and essential procedure in radiotherapy workflow, and several studies have been done to develop prediction models based on the conventional C-arm linac of single-layered multileaf collimator (MLC) with machine learning (ML) and deep learning techniques to predict PSQA results and improve efficiency. Recently, a newly designed O-ring gantry linac "Halcyon" equipped with unique jawless stacked-and-staggered dual-layered MLC was released. However, few studies have focused on developing PSQA prediction models for this novel dual-layered MLC system. To evaluate the performance of ML to predict PSQA results of fixed field intensity-modulated radiation therapy (FF-IMRT) plans for linac equipped with dual-layered MLC. A total of 213 FF-IMRT treatment plans, including 1383 beams from various treatment sites, were selected and delivered with portal dosimetry verification on Halcyon linac. Gamma analysis was performed using 1%/1, 2%/2, and 3%/2mm criteria with a 10% threshold. The training set (TS) of ML models consisted of 1106 beams, and an independent evaluation set (ES) consisted of 277 beams. For each beam, 33 complexity metrics were extracted as input data for training models. Three ML algorithms (gradient boosting decision tree/GBDT, random forest/RF, and Poisson Lasso/PL) were utilized to build the models and predict gamma passing rates (GPRs). To improve the prediction accuracy in the rare region, a method of reweighting for TS has been performed and compared to the unweighted results. The importance of complexity metrics was studied by permuted interesting features. The GBDT model had the best performance in this study. In ES, the minimal mean prediction error for unweighted results was 1.93%, 1.16%, 0.78% under 1%/1, 2%/2, and 3%/2mm criteria, respectively, from GBDT model. Comparing to the unweighted results, the models after reweighting gained up to 30% improvement in the rare region, whereas the overall prediction error was slightly worse depending on the kind of models. For feature importance, 2 tree-based models (GBDT and RF) had in common the top 10 most important metrics as well as the same metric with the largest impact. For linac equipped with novel dual-layered MLC, the ML model based on GBDT algorithm shows a certain degree of accuracy for GPRs prediction. The specific ML model for dual-layered MLC configuration could be a useful tool for physicists detecting PSQA measurement failures.

Validation of pencil beam scanning proton therapy with multi-leaf collimator calculated by a commercial Monte Carlo dose engine.

This study aimed to evaluate the clinical beam commissioning results and lateral penumbra characteristics of our new pencil beam scanning (PBS) proton therapy using a multi-leaf collimator (MLC) calculated by use of a commercial Monte Carlo dose engine. Eighteen collimated uniform dose plans for cubic targets were optimized by the RayStation 9A treatment planning system (TPS), varying scan area, modulation widths, measurement depths, and collimator angles. To test the patient-specific measurements, we also created and verified five clinically realistic PBS plans with the MLC, such as the liver, prostate, base-of-skull, C-shape, and head-and-neck. The verification measurements consist of the depth dose (DD), lateral profile (LP), and absolute dose (AD). We compared the LPs and ADs between the calculation and measurements. For the cubic plans, the gamma index pass rates (γ-passing) were on average 96.5%±4.0% at 3%/3mm for the DD and 95.2%±7.6% at 2%/2mm for the LP. In several LP measurements less than 75mm depths, the γ-passing deteriorated (increased the measured doses) by less than 90% with the scattering such as the MLC edge and range shifter. The deteriorated γ-passing was satisfied by more than 90% at 2%/2mm using uncollimated beams instead of collimated beams except for three planes. The AD differences and the lateral penumbra width (80%-20% distance) were within±1.9% and±1.1mm, respectively. For the clinical plan measurements, the γ-passing of LP at 2%/2mm and the AD differences were 97.7%±4.2% on average and within±1.8%, respectively. The measurements were in good agreement with the calculations of both the cubic and clinical plans inserted in the MLC except for LPs less than 75mm regions of some cubic and clinical plans. The calculation errors in collimated beams can be mitigated by substituting uncollimated beams.

Monte Carlo simulation of linac using PRIMO

BackgroundMonte Carlo simulation is considered as the most accurate method for dose calculation in radiotherapy. PRIMO is a Monte-Carlo program with a user-friendly graphical interface.Material and methodA VitalBeam with 6MV and 6MV flattening filter free (FFF), equipped with the 120 Millennium multileaf collimator was simulated by PRIMO. We adjusted initial energy, energy full width at half maximum (FWHM), focal spot FWHM, and beam divergence to match the measurements. The water tank and ion-chamber were used in the measurement. Percentage depth dose (PDD) and off axis ratio (OAR) were evaluated with gamma passing rates (GPRs) implemented in PRIMO. PDDs were matched at different widths of standard square fields. OARs were matched at five depths. Transmission factor and dose leaf gap (DLG) were simulated. DLG was measured by electronic portal imaging device using a sweeping gap method.ResultFor the criterion of 2%/2 mm, 1%/2 mm and 1%/1 mm, the GPRs of 6MV PDD were 99.33–100%, 99–100%, and 99–100%, respectively; the GPRs of 6MV FFF PDD were 99.33–100%, 98.99–99.66%, and 97.64–98.99%, respectively; the GPRs of 6MV OAR were 96.4–100%, 90.99–100%, and 85.12–98.62%, respectively; the GPRs of 6MV FFF OAR were 95.15–100%, 89.32–100%, and 87.02–99.74%, respectively. The calculated DLG matched well with the measurement (6MV: 1.36 mm vs. 1.41 mm; 6MV FFF: 1.07 mm vs. 1.03 mm, simulation vs measurement). The transmission factors were similar (6MV: 1.25% vs. 1.32%; 6MV FFF: 0.8% vs. 1.12%, simulation vs measurement).ConclusionThe calculated PDD, OAR, DLG and transmission factor were all in good agreement with measurements. PRIMO is an independent (with respect to analytical dose calculation algorithm) and accurate Monte Carlo tool.

Elekta Unity MR-linac commissioning: mechanical and dosimetry tests

Abstract We report the commissioning results of Elekta Unity for the dosimetric performance and mechanical quality assurance (QA), and propose additional commissioning procedures. Mechanical tests included multi-leaf collimator (MLC) positional accuracy, radiation isocenter diameter at the center and off-center position, and coincidence between the magnetic resonance (MR) image center and radiation isocenter. Comparisons between the measurements and calculations of the simple irradiated field, intensity modulated radiation therapy (IMRT) commissioning, MLC output factor ratio, validation of independent dose calculation software and end-to-end testing were performed to evaluate dosimetric performance. The average values of the MLC positional accuracy for film- and imaging device-based analysis were −0.1 and 0.3 mm, respectively. The measured radiation isocenter size was 0.41 mm, and the off-center results were within 1 mm. The coincidence was −0.21, −1.19 and 0.49 mm along the x-, y- and z-axes, respectively. The calculated percent depth doses (PDD) and profiles agreed with the measurements. The results of independent dose calculation were within the action level recommended by American Associations of Physicist in Medicine. The gamma passing rate (GPR) for IMRT commissioning was 98.6 ± 0.9%, and end-to-end testing of adapted plans showed agreement within 2% between the measurement and calculation. We reported the results of mechanical and dosimetric performances of Elekta Unity, and proposed novel commissioning procedures. Our results should provide knowledge to the physics community for enhancing the QA programs.

Anomaly detection in time and spectral domain for the driving motors of multileaf collimator of a medical linear accelerator

Anomaly detection in time and spectral domain for the driving motors of multileaf collimator of a medical linear accelerator

Measurement of Absorbed Dose in Water Using Co-60 gamma source Teletherapy Units According to IAEA Dosimetry Protocols

Modern radiotherapy facilities like 3-Dimensional conformal radio therapy (3DCRT), Intensity Modulated Radiotherapy (IMRT), were recently suggested in Co-60 machine with Multi-leaf Collimator (MLC). In this study, two reference chambers NE-2571#1205 and NE2581#537 were used for absolute dose measurement in Equinox accelerator. A comparison of dose measurement by two different IAEA protocols TRS-277 and TRS-398 has been studied. Analyzing TRS-398, a common shaped empirical formula was developed for the four Co-60 units of four Medical Colleges in Bangladesh with fitting parameters. It was found that an average discrepancy in the determination of absorbed dose in water among the two different protocols TRS-277 and TRS-398 were 1.33 % for the chamber NE-2571#1205 and 0.65 % for the chamber NE2581#537 with combined uncertainty ±1.59 % (k=1). A good convergence has been obtained in the concepts and methods in this study.

Challenges in modeling the Agility multileaf collimator in treatment planning systems and current needs for improvement.

The Agility multileaf collimator (MLC) mounted in Elekta linear accelerators features some unique design characteristics, such as large leaf thickness, eccentric curvature at the leaf tip, and defocused leaf sides ('tilting'). These characteristics offer several advantages but modeling them in treatment planning systems (TPSs) ischallenging. The goals of this study were to investigate the challenges faced when modeling the Agility in two commercial TPSs (Monaco and RayStation) and to explore how the implemented MLC models could be improved in thefuture. Four linear accelerators equipped with the Agility, located at different centers, were used for the study. Three centers use the RayStation TPS and the other one uses Monaco. For comparison purposes, data from four Varian linear accelerators with the Millennium 120 MLC were also included. Average doses measured with asynchronous sweeping gap tests were used to characterize and compare the characteristics of the Millennium and the Agility MLCs and to assess the MLC model in the TPSs. The FOURL test included in the ExpressQA package, provided by Elekta, was also used to evaluate the tongue-and-groove with radiochromic films. Finally, raytracing was used to investigate the impact of the MLC geometry and to understand the results obtained for eachMLC. The geometry of the Agility produces dosimetric effects associated with the rounded leaf end up to a distance 20mm away from the leaf tip end measured at the isocenter plane. This affects the tongue-and-groove shadowing, which progressively increases along the distance to the tip end. The RayStation and Monaco TPSs did not account for this effect, which made trade-offs in the MLC parameters necessary and greatly varied the final MLC parameters used by different centers. Raytracing showed that these challenging leaf tip effects were directly related to the MLC geometry and that the characteristics mainly responsible for the large leaf tip effects of the Agility were its tilting design and its small source-to-collimatordistance. The MLC models implemented in RayStation and Monaco could not accurately reproduce the leaf tip effects for the Agility. Therefore, trade-offs are needed and the optimal MLC parameters are dependent on the specific characteristics of treatment plans. Refining the MLC models for the Agility to better approximate the measured leaf tip and tongue-and-groove effects would extend the validity of the MLC model, reduce the variability in the MLC parameters used by the community, and facilitate the standardization of the MLC configurationprocess.

Statistical process control to monitor use of a web-based autoplanning tool.

To investigate the use of statistical process control (SPC) for quality assurance of an integrated web-based autoplanning tool, Radiation Planning Assistant (RPA). Automatically generated plans were downloaded and imported into two treatment planning systems (TPSs), RayStation and Eclipse, in which they were recalculated using fixed monitor units. The recalculated plans were then uploaded back to the RPA, and the mean dose differences for each contour between the original RPA and the TPSs plans were calculated. SPC was used to characterize the RPA plans in terms of two comparisons: RayStation TPS versus RPA and Eclipse TPS versus RPA for three anatomical sites, and variations in the machine parameters dosimetric leaf gap (DLG) and multileaf collimator transmission factor (MLC-TF) for two algorithms (Analytical Anisotropic Algorithm [AAA]) and Acuros in the Eclipse TPS. Overall, SPC was used to monitor the process of the RPA, while clinics would still perform their routine patient-specific QA. For RayStation, the average mean percent dose differences across all contours were 0.65%±1.05%, -2.09%±0.56%, and 0.28%±0.98% and average control limit ranges were 1.89%±1.32%, 2.16%±1.31%, and 2.65%±1.89% for the head and neck, cervix, and chest wall, respectively. In contrast, Eclipse's average mean percent dose differences across all contours were -0.62%±0.34%, 0.32%±0.23%, and -0.91%±0.98%, while average control limit ranges were 1.09%±0.77%, 3.69%±2.67%, 2.73%±1.86%, respectively. Averaging all contours and removing outliers, a 0% dose difference corresponded with a DLG value of 0.202±0.019cm and MLC-TF value of 0.020±0.001 for Acuros and a DLG value of 0.135±0.031cm and MLC-TF value of 0.015±0.001 for AAA. Differences in mean dose and control limits between RPA and two separately commissioned TPSs were determined. With varying control limits and means, SPC provides a flexible and useful process quality assurance tool for monitoring a complex automated system such as the RPA.

MLC performance prognosis using a degradation model based on trajectory log data from a daily test.

This paper investigates the feasibility of implementing a predictive maintenance program for a multileaf collimator (MLC) based on data collected in trajectory logs (TLs) obtained by conducting a simple daily test, with the aim of minimizing unscheduled downtime. A dynamic field test was designed, and the TLs generated in the course of daily administration in a linear accelerator were collected to evaluate trajectory deviations of the MLC leaves as well as interlocks (COL 420219/20, COL 420207/08) reported by the machine. During this evaluation, we observed that the trajectory deviations of some leaves increased up to a threshold value beyond which certain interlocks began to appear in treatment fields in those leaves. An exponential degradation model was therefore developed to predict this drift and determine each leaf's remaining useful life (RUL). Once the applicability of the model was confirmed, we added a second accelerator equipped with an MLC with the same configuration to validate the model. The model was able to predict certain COL 420219/20 interlocks resulting from primary readout/expected position discrepancies and to estimate each leaf's RUL. In total, 11 cases (8 interlocks + 3 potential interlocks avoided due to service interventions [27.3% of the total]) were detected over 7 days in advance, with no false positive results. Scheduling of service interventions several days prior to MLC failure would therefore be possible. When these types of interlocks were not predicted by the model, they were always generated by leaf motor failure. Consequently, intervention time could also be optimized by directly replacing the motor. During the study period, for these types of interlocks, our approach would have reduced downtime from 35.25 to 4.00h (88.7%) and from 34.75 to 22.83h (34.3%) for each accelerator, respectively. For COL 420207/08 interlocks, which are generated by primary/secondary readout discrepancies, no correlation with leaf trajectory deviation increases recorded in the TLs was found. Throughout the study period, these types of interlocks requiring service intervention, also mainly for motor replacement, represented a downtime of 9.50h for the first accelerator (21.2% of total downtime) and by 4.33h (11.1% of total downtime) for the second accelerator. This study demonstrates that by applying a predictive MLC maintenance program based on information collected in TLs, it is possible to predict certain interlocks and therefore schedule preemptive interventions to avoid their occurrence. This could optimize health-care delivery performance and minimize the loss of treatment sessions.

Small-field dosimetry with detector-specific output correction factor for single-isocenter stereotactic radiotherapy of single and multiple brain metastases.

Recently, the International Atomic Energy Agency and the American Association of Physicists in Medicine reported correction factors (CFs) for detector-response variation considering the uncertainty in detector readings in small-field dosimetry. In this study, the effect of CFs on small-field dosimetry measurements was evaluated for single-isocenter stereotactic radiotherapy for brain metastases. The output factors (OPFs) were measured with and without CFs in a water-equivalent sphere phantom using TrueBeam with a flattening-filter-free energy of 10 MV. Five detectors were used in a perpendicular orientation: CC01, 3D pinpoint ionization chambers, unshielded SFD detector, shielded EDGE detector, and microDiamond detector. First, the square-field sizes were set to 5-100mm using a multi-leaf collimator (MLC) field. The OPFs were evaluated in the presence and absence of CFs. Second, single-isocenter stereotactic irradiation was performed on 22 brain metastases in 15 patients following dynamic conformal arc (DCA) treatment. The equivalent field size was calculated using the MLC aperture for each planning target volume. For the OPFs, the mean deviations from the median of the doses measured with detectors other than the CC01 for square-field sizes larger than 10mm were within ± 4.3% of the median without CFs, and ± 3.3% with CFs. For DCA plans, the deviations without and with CFs were -2.3 ± 1.9% and -4.8 ± 2.4% for CC01, -1.1 ± 3.0% and 1.0 ± 1.6% for 3D pinpoint, 8.8 ± 3.0% and 2.9 ± 2.8% for SFD, -3.1 ± 3.0% and -13.5 ± 4.0% for EDGE, and 8.9 ± 2.1% and 0.8 ± 1.9% for microDiamond. This feasibility study confirmed that the deviation of the detectors can be reduced using an appropriate detector with CFs.

HyperArc Planning for Benign Intracranial Pathologies: A Prospective Analysis on Plan Quality, Delivery, Control and Toxicity

<h3>Purpose/Objective(s)</h3> This study reviews a treatment planning system (TPS) that automates both treatment planning and delivery of single-isocenter VMAT radiosurgery (SRS). The TPS was intended for complex multi-metastasis cases, for which it generates high quality, rapidly-deliverable plans. The effectiveness of treating benign intracranial pathologies (BIP) with TPS was unknown. Under an IRB-approved prospective registry, we collected data on the treatment planning and delivery as well as clinical outcomes of BIP managed with SRS since deployment of TPS. <h3>Materials/Methods</h3> Patients included received SRS between October 2017 and November 2021 using TPS. Patients with less than 3 months follow-up (FU) were censored. For all targets, full prescription dose was normalized to ≥ 99% of target volume without additional PTV expansion. All treatments were delivered on a linear accelerator with 10MV flattening-filter free beam at 2400 MU/min with high-definition (2.5mm) multi-leaf collimator. Descriptive statistics related to SRS treatment, pathology, and prior therapy determined at the time of or prior to treatment were analyzed. Post-treatment imaging, toxicities, and standard pathology-specific outcomes were assessed during FU visits. Significant toxicity was defined as ≥ Grade 3 by CTCAE. <h3>Results</h3> 245 targets (min=0.1cc, max=58.9cc) in 235 patients (53.5% female, median age=59) were treated with TPS. The most common pathologies were meningiomas (134), AVMs (43), pituitary adenomas (30), and acoustic schwannomas (23). 45% of patients were treated due to recurrent, residual, or persistent disease after prior treatment with 15% having prior radiation. A majority (53.5%) were treated in a single fraction (12-22Gy), and 46.5% were treated with fractionated SRS (24-35Gy). A majority (74%) were treated with 3 arcs with median (MED) total treatment and beam-on time lasting 10.4 and 2.3 minutes, respectively. MED RTOG CI and Paddick GI were 1.12 and 3.28, respectively. MED FU was 1.1 years. Excluding AVMs, 11 of 202 (5.4%) benign tumors progressed on FU imaging. Among those that progressed, MED time to failure was 1.14 years. Of those, 9 were WHO grade 2+ meningiomas with prior surgery. Of all patients, significant CNS toxicity was reported in 13 (5.6%, G3=9, G4=4), most of whom had persistent cerebral edema after steroids. Of those with BIP near cochlea (> 3 Gy dose max) without severe hearing loss prior to SRS (n=65), 14% subjectively reported decreased hearing after SRS and 3.5% did not preserve hearing. Of those with BIP near the optics (> 3 Gy dose max, n=81), visual preservation was 100%. <h3>Conclusion</h3> Although developed for multi-metastasis SRS, TPS is capable of generating and efficiently delivering high quality plans characterized by excellent conformality and rapid fall-off. Early clinical outcomes appear congruent with historical controls.

3D Dose-Driven, Automatic VMAT Machine Parameter Generation with Deep Learning

<h3>Purpose/Objective(s)</h3> Recent research efforts utilizing knowledge-based treatment planning for the prediction of 3D radiation dose distributions from planning structure sets have achieved positive results. Most ongoing efforts to generate deliverable plans from the predicted doses rely on full inverse optimizations using dose-volume histogram (DVH) objectives derived from these doses. In this study, we aim to leverage deep learning (DL) to rapidly generate machine delivery parameters for volumetric modulated arc therapy (VMAT) from predicted desired doses. <h3>Materials/Methods</h3> Data of 50 previously treated patients at our institution with prostate adenocarcinoma who received definitive radiotherapy were retrospectively obtained. All plans were generated with a one-arc VMAT technique, with conventional fractionation (78 Gy in 39 fx or 79.2 Gy in 44 fx to the prostate gland +/- seminal vesicles). A multi-task U-Net was constructed: it takes the 2D projections of the 3D dose and planning structures as inputs, and it predicts the numerical multi-leaf collimator (MLC) sequence and weights for the 178 control points. Five cases were randomly selected for testing only, and the remaining 45 formed the training set. The algorithm was implemented in Python 3.8 with PyTorch 1.7 as the DL framework. Model training was performed on a GPU. The DL-predicted plans underwent further inverse optimization with the 3D-dose-derived DVH objectives, utilizing only the last step of the Photon Optimizer (PO) in a treatment planning system. The optimization time and plan quality were compared to plans generated with one full PO optimization with the same objectives and clinical plans (all normalized to D95%=100% Rx dose). <h3>Results</h3> The DL model was trained for 200 epochs. On average, DL-predicted plans could be optimized in 22% (range, 18-26%) of the time required for full optimization plans. Dosimetric comparison (Table 1) demonstrated that the quality of the DL-predicted plans was comparable with clinical plans and full optimization plans, but the DL-predicted plans tended to have increased dose inhomogeneity within the PTVs. <h3>Conclusion</h3> We demonstrated the feasibility of rapidly generating deliverable VMAT plans from desired 3D doses with deep learning. Further work is needed to improve PTV dose homogeneity and generalize the method to multi-arc VMAT delivery.