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.

In this study the dosimetric characteristics of 120-leaf multileaf collimators (MLCs) were evaluated for 6-MV and 18-MV photon beams. The dose rate, percentage depth dose, surface dose, dose in the build-up region, beam profile, flatness, symmetry, and penumbra width were measured using three field-defining methods: (i) ‘Jaw only’, (ii) ‘MLC only’, and (iii) ‘MLC+Jaw’. Analysis of dose rate shows that the dose rate for ‘MLC only’ field was higher than that for ‘Jaw only” and ‘MLC+Jaw’ fields in both the energies. The ‘percentage of difference’ of dose rates between ‘MLC only’ and ‘MLC+Jaw’ was (0.9% to 4.4%) and (1.14% to 7%) for 6 MV and 18 MV respectively. The surface dose and dose in the build-up region were more pronounced for ‘MLC only’ fields for both energies, and no significant difference was found in percentage depth dose beyond dmax for both energies. Beam profiles show that flatness and symmetry for both the energies were less than the 3%. The penumbra width for ‘MLC only’ field was more than that for the other two field-defining methods by (1 to 2 mm) and (0.8 to 1.3 mm) for 6-MV and 18-MV photon beams respectively. Analysis of ‘width of 50% dose level’ of the beam profiles at dmax to reflect the field size shows 1 to 2 mm more for 6-MV photons and 2.2 to 2.4 mm morefor 18-MV photons for ‘MLC only’ fields. The results of this study suggest that the characteristics of 120-leaf MLC system with 6 MV and 18 MV are same in all aspects except the surface dose, penumbra, dose in the build-up region, and width of 50% dose levels.

In this study, the dosimetric characteristics of multileaf collimators (MLCs) with 120 leaves and 80 leaves were evaluated. The dose rate, percentage depth dose, surface dose, dose in the build-up region, beam profile, flatness, symmetry, and penumbra width were measured by three field-defining methods: (1) "Jaw only", (2) "MLC only", and (3) "MLC + Jaw", for a 6-MV photon beam with the two MLC systems. Analysis of the dose rate showed that the dose rate for the "MLC only" field was higher than that for the other two fields. The surface dose was more pronounced for the "MLC only" fields. The dose in the build-up region was higher for the "MLC only" fields, and no significant difference was found in the percentage depth dose (PDD) beyond the dose maximum point (d(max)) for both MLC systems. Beam profiles showed that the differences in flatness and symmetry for both systems were less than 3%. The penumbra width between 80 and 20% isodose levels for the "MLC only" field was more than that for the other two field defining methods. The widths of the 50% dose levels of the beam profiles were analyzed. The dosimetric characteristics of the two MLC systems were the same in all aspects except the surface dose, penumbra, the dose in the build-up region, and the width of the 50% dose levels.