Evaluation of MLC performance in VMAT and dynamic IMRT by log file analysis

Olasolo-Alonso, José; Vázquez-Galiñanes, Alejandro; Pellejero-Pellejero, Santiago; Pérez-Azorín, José Fernando

doi:10.1016/j.ejmp.2016.12.013

Cited by 35 publications

(43 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A linear relationship between the MLC leaf velocity and positional error was reported by some investigators. 7,15,16 Hence, constraining the leaf speed and millimeters traveled per leaf per monitor unit (MU) can help improve the accuracy of the IMRT treatment plan delivery. 15,17 Advanced approaches using machine learning (ML) methods could be used for predicting the MLC leaf positional deviations during the IMRT/VMAT delivery priori.…”

Section: Introductionmentioning

confidence: 99%

“…Even though several studies 8,15,16,[27][28][29][30][31][32][33][34][35] were conducted on the MLC leaf positional errors for dynamic IMRT delivery through the analysis of the machine log file data 15,16,[27][28][29] or electronic portal imaging device, 27,30-33 they may be considered passive and do not provide predictions ahead of time. ML methods can provide the active prediction of the MLC positional errors during the IMRT/VMAT delivery.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of the individual multileaf collimator positional deviations during dynamic IMRT delivery priori with artificial neural network

2020

View full text Add to dashboard Cite

Purposes: Multileaf collimator (MLC) positional accuracy during dynamic intensity modulation radiotherapy (IMRT) delivery is crucial for safe and accurate patient treatment. The deviations of individual leaf positions from its intended positions can lead to errors in the dose delivered to the patient and hence may adversely affect the treatment outcome. In this study, we propose a state-ofthe-art machine learning (ML) method based on an artificial neural network (ANN) for accurately predicting the MLC leaf positional deviations during the dynamic IMRT treatment delivery priori using log file data. Methods: Data of ten patients treated with sliding window dynamic IMRT delivery were retrospectively retrieved from a single-institution database. The patients' plans were redelivered with no patient on the couch using a Varian linear accelerator equipped with a Millennium 120 HD MLC system. Then the machine recorded log files data, a total of over 400 files containing 360 800 control points, were collected. A total of 14 parameters were extracted from the planning data in the log files such as leaf planned positions, dose fraction, leaf velocity, leaf moving status, leaf gap, and others. Next, we developed a feed-forward ANN architecture mapping the input parameters with the output to predict the MLC leaf positional deviations during the delivery priori. The proposed model was trained on 70% of the total data using the delivered leaf positional data as a target response. The trained model was then validated and tested on 30% of the available data. The model accuracy was evaluated using the mean squared error (MSE), regression plot, and error histogram. Results: The deviations between the individual MLC planned and delivered positions can reach up to a few millimeters, with a maximum deviation of 1.2 mm. The predicted leaf positions at control points closely matched the delivered positions for all MLC leaves during the treatment delivery. The ANN model achieved a maximum MSE of 0.0001 mm 2 (root MSE of 0.0097 mm) in predicting the leaf positions at control points of test data for each leaf. The correlation coefficient, that measures the goodness of fit, was perfect (R = 0.999) in all plots indicating an excellent agreement between the predicted and delivered MLC positions for the training, validation, and test data. Conclusions: We successfully demonstrated a proposed ANN-based method capable of accurately predicting the individual MLC leaf positional deviations during the dynamic IMRT delivery priori. Our ML model based on ANN outperformed the reported accuracy in the literature of various ML models. The results of this study could be extended to actual application in the dose calculation/optimization, hence enhancing the gamma passing rate for patient-specific IMRT quality assurance.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Prediction of the individual multileaf collimator positional deviations during dynamic IMRT delivery priori with artificial neural network

2020

View full text Add to dashboard Cite

show abstract

“…Scaggion et al reported that the mean mechanical errors in clinical VMAT plans were less than 1 mm, 0.5°, and 0.1 MU for MLC position, gantry angle, and MU, respectively, during RapidArc treatments delivered with a 6 MV UNIQUE linac (Varian Medical Systems, Palo Alto, CA, USA) over 2 years [ 9 ]. José Olasolo-Alonso et al also reported that the average RMSE values of the MLC were 0.3 mm for Clinac linac (Varian Medical Systems) in intensity-modulated radiation therapy (IMRT) treatments, and 0.04 mm for TrueBeam linacs (Varian Medical Systems) in VMAT treatments, according to 3000 logfiles obtained from four institutions [ 10 ]. The mechanical accuracy results observed in this study for VMDWAT, which is a more complex delivery technique compared with VMAT, were similar to those results .…”

Section: Discussionmentioning

confidence: 99%

“…Even if the dosimetric QA testing results meet institutional criteria, there is no guarantee that treatment machines will continue to perform correctly throughout the course of treatment. Several studies have reported that daily machine errors were small throughout the course of VMAT treatment [ 9 , 10 ]; however, it is generally difficult to predict dosimetric impacts on patient anatomy due to mechanical errors. Thus far, no studies have reported monitoring of the dosimetric impact of mechanical errors throughout the course of treatment.…”

Section: Introductionmentioning

confidence: 99%

Monitoring of mechanical errors and their dosimetric impact throughout the course of non-coplanar continuous volumetric-modulated arc therapy

et al. 2018

View full text Add to dashboard Cite

BackgroundVolumetric-modulated Dynamic WaveArc therapy (VMDWAT) is a non-coplanar continuous volumetric modulated radiation therapy (VMAT) delivery technique. Here, we monitored mechanical errors and their impact on dose distributions in VMDWAT using logfiles throughout the course of treatment.MethodsFifteen patients were enrolled (2 skull base tumor patients and 13 prostate cancer patients). VMDWAT plans were created for the enrolled patients. The prescribed dose for the skull base tumor was set as 54 Gy at 1.8 Gy per fraction, and that for the prostate cancer was set as 72 to 78 Gy at 2 Gy per fraction. We acquired logfiles to monitor mechanical errors and their impact on dose distribution in each fraction. The root mean square error (RMSE) in the multi-leaf collimator (MLC), gantry angle, O-ring angle and monitor unit (MU) were calculated using logfiles throughout the course of VMDWAT for each patient. The dosimetric impact of mechanical errors throughout the course of VMDWAT was verified using a logfile-based dose reconstruction method. Dosimetric errors between the reconstructed plans and the original plans were assessed.ResultsA total of 517 datasets, including 55 datasets for the 2 skull base tumor patients and 462 datasets for the 13 prostate cancer patients, were acquired. The RMSE values were less than 0.1 mm, 0.2°, 0.1°, and 0.4 MU for MLC position, gantry angle, O-ring angle, and MU, respectively. For the skull base tumors, the absolute mean dosimetric errors and two standard deviations throughout the course of treatment were less than 1.4% and 1.1%, respectively. For prostate cancer, these absolute values were less than 0.3% and 0.5%, respectively. The largest dosimetric error of 2.5% was observed in a skull base tumor patient. The resultant dosimetric error in the accumulated daily delivered dose distribution, in the patient with the largest error, was up to 1.6% for all dose-volumetric parameters relative to the planned dose distribution.ConclusionsMLC position, gantry rotation, O-ring rotation and MU were highly accurate and stable throughout the course of treatment. The daily dosimetric errors due to mechanical errors were small. VMDWAT provided high delivery accuracy and stability throughout the course of treatment.Trial registrationUMIN000023870. Registered: 1 October 2016.Electronic supplementary materialThe online version of this article (10.1186/s13014-018-0972-7) contains supplementary material, which is available to authorized users.

show abstract

“…Therefore, there is a need for a new method for verifying the accuracy of IMRT and VMAT deliveries in clinically relevant terms, such that the actual patient dose can be investigated; several authors 9-14 have described methods for doing this. Patient specific quality assurance using the linear accelerator (linac) log file to generate a dosevolume histogram (DVH) of the patient during the actual delivery allows the possibility of monitoring the average MLC positional errors, [15][16][17][18][19][20] as well as providing information on the accuracy of the dose delivery to the patient's target and critical organ structures.…”

Section: Introductionmentioning

confidence: 99%