Improvement of deep learning prediction model in patient‐specific QA for VMAT with MLC leaf position map and patient's dose distribution

Abstract

PurposeDeep learning‐based virtual patient‐specific quality assurance (QA) is a novel technique that enables patient QA without measurement. However, this method could be improved by further evaluating the optimal data to be used as input. Therefore, a deep learning‐based model that uses multileaf collimator (MLC) information per control point and dose distribution in patient's CT as inputs was developed.MethodsOverall, 96 volumetric‐modulated arc therapy plans generated for prostate cancer treatment were used. We developed a model (Model 1) that can predict measurement‐based gamma passing rate (GPR) for a treatment plan using data stored as a map reflecting the MLC leaf position at each control point (MLPM) and data of the dose distribution in patient's CT as inputs. The evaluation of the model was based on the mean absolute error (MAE) and Pearson's correlation coefficient (r) between the measured and predicted GPR. For comparison, we also analyzed models trained with the dose distribution in patient's CT alone (Model 2) and with dose distributions recalculated on a virtual phantom CT (Model 3).ResultsAt the 2%/2 mm criterion, MAE[%] and r for Model 1, Model 2, and Model 3 were 2.32% ± 0.43% and 0.54 ± 0.03, 2.70% ± 0.26%, and 0.32 ± 0.08, and 2.96% ± 0.23% and 0.24 ± 0.22, respectively; at the 3%/3 mm criterion, these values were 1.25% ± 0.05% and 0.36 ± 0.18, 1.57% ± 0.35% and 0.19 ± 0.20, and 1.39% ± 0.32% and 0.17 ± 0.22, respectively. This result showed that Model 1 exhibited the lowest MAE and highest r at both criteria of 2%/2 mm and 3%3 mm.ConclusionsThese findings showed that a model that combines the MLPM and dose distribution in patient's CT exhibited a better GPR prediction performance compared with the other two studied models.

Applications of artificial intelligence for machine- and patient-specific quality assurance in radiation therapy: current status and future directions

“…In terms of input data for dose distribution, Tozuka et al. [ 39 ] predicted the GPR from MLC position maps and dose distribution using an ANN. They compared three models, namely, dose distribution in the patient geometry combined with MLC positions (model 1), dose distribution in the patient geometry (model 2), and dose distribution on a helical diode array (model 3).…”

Unbiased evaluation of predicted gamma passing rate by an event‐mixing technique

“…Many authors have been developing methods to predict the GPR from a model developed by learning characteristics of the treatment plan, such as the complexity metric, [3][4][5][6][7] dose uncertainty potential (DUP), 8,9 machine learning, [10][11][12][13][14][15][16] and deep learning techniques. [17][18][19][20][21] The achievement of the predicting model is evaluated using such as a standard deviation (SD) of the difference between the measured GPR [m] and the predicted GPR [p], [8][9][10][11][12]18 and Pearson's correlation coefficient (CC) of the (m, p) pairs. 3,5,[10][11][12]17,18,21 However, these studies used data from different radiotherapy systems comprising TPS, linear accelerator, and detector array.…”

“…[17][18][19][20][21] The achievement of the predicting model is evaluated using such as a standard deviation (SD) of the difference between the measured GPR [m] and the predicted GPR [p], [8][9][10][11][12]18 and Pearson's correlation coefficient (CC) of the (m, p) pairs. 3,5,[10][11][12]17,18,21 However, these studies used data from different radiotherapy systems comprising TPS, linear accelerator, and detector array. TPS has different types of optimization algorithms for intensity modulation and dose calculation algorithms.…”