A machine learning approach to the accurate prediction of monitor units for a compact proton machine.

PURPOSE Clinical treatment planning systems for proton therapy currently do not calculate monitor units (MUs) in passive scatter proton therapy due to the complexity of the beam delivery systems. Physical phantom measurements are commonly employed to determine the field-specific output factors (OFs) but are often subject to limited machine time, measurement uncertainties and intensive labor. In this study, a machine learning-based approach was developed to predict output (cGy/MU) and derive MUs, incorporating the dependencies on gantry angle and field size for a single-room proton therapy system. The goal of this study was to develop a secondary check tool for OF measurements and eventually eliminate patient-specific OF measurements. METHOD The OFs of 1754 fields previously measured in a water phantom with calibrated ionization chambers and electrometers for patient-specific fields with various range and modulation width combinations for 23 options were included in this study. The training data sets for machine learning models in three different methods (Random Forest, XGBoost and Cubist) included 1431 (~81%) OFs. Ten-fold cross-validation was used to prevent "overfitting" and to validate each model. The remaining 323 (~19%) OFs were used to test the trained models. The difference between the measured and predicted values from machine learning models was analyzed. Model prediction accuracy was also compared with that of the semi-empirical model developed by Kooy (Phys. Med. Biol. 50, 2005). Additionally, gantry angle dependence of OFs was measured for three groups of options categorized on the selection of the second scatters. Field size dependence of OFs was investigated for the measurements with and without patient-specific apertures. RESULTS All three machine learning methods showed higher accuracy than the semi-empirical model which shows considerably large discrepancy of up to 7.7% for the treatment fields with full range and full modulation width. The Cubist-based solution outperformed all other models (P < 0.001) with the mean absolute discrepancy of 0.62% and maximum discrepancy of 3.17% between the measured and predicted OFs. The OFs showed a small dependence on gantry angle for small and deep options while they were constant for large options. The OF decreased by 3%-4% as the field radius was reduced to 2.5 cm. CONCLUSION Machine learning methods can be used to predict OF for double-scatter proton machines with greater prediction accuracy than the most popular semi-empirical prediction model. By incorporating the gantry angle dependence and field size dependence, the machine learning-based methods can be used for a sanity check of OF measurements and bears the potential to eliminate the time-consuming patient-specific OF measurements.

[1]  Dong Wook Kim,et al.  Prediction of output factor, range, and spread-out bragg peak for proton therapy. , 2011, Medical dosimetry : official journal of the American Association of Medical Dosimetrists.

[2]  Yuanshui Zheng,et al.  Commissioning of output factors for uniform scanning proton beams. , 2011, Medical physics.

[3]  Tianqi Chen,et al.  XGBoost: A Scalable Tree Boosting System , 2016, KDD.

[5]  T D Solberg,et al.  A mathematical framework for virtual IMRT QA using machine learning. , 2016, Medical physics.

[6]  Thomas Bortfeld,et al.  Monitor unit calculations for range-modulated spread-out Bragg peak fields. , 2003, Physics in medicine and biology.

[8]  L. Ungar,et al.  MediBoost: a Patient Stratification Tool for Interpretable Decision Making in the Era of Precision Medicine , 2016, Scientific Reports.

[9]  E. Klein,et al.  Optimizing field patching in passively scattered proton therapy with the use of beam current modulation , 2013, Physics in medicine and biology.

[10]  M Goitein,et al.  A pencil beam algorithm for proton dose calculations. , 1996, Physics in medicine and biology.

[11]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[12]  Eros Pedroni,et al.  Treating Cancer with Protons , 2002 .

[13]  Timothy D. Solberg,et al.  IMRT QA using machine learning: A multi‐institutional validation , 2017, Journal of applied clinical medical physics.

[14]  Commissioning and initial experience with the first clinical gantry‐mounted proton therapy system , 2016, Journal of applied clinical medical physics.

[15]  David Gozal,et al.  Neurobehavioral implications of habitual snoring in children. , 2004, Pediatrics.

[16]  So-Yeon Park,et al.  A machine learning approach to the accurate prediction of multi-leaf collimator positional errors , 2016, Physics in medicine and biology.

[17]  J. Flickinger,et al.  Machine Learning Approaches for Predicting Radiation Therapy Outcomes: A Clinician's Perspective. , 2015, International journal of radiation oncology, biology, physics.

[18]  Harald Paganetti,et al.  The prediction of output factors for spread-out proton Bragg peak fields in clinical practice , 2005, Physics in medicine and biology.

[19]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[20]  Harald Paganetti,et al.  Field size dependence of the output factor in passively scattered proton therapy: influence of range, modulation, air gap, and machine settings. , 2009, Medical physics.

[21]  Vasant Dhar,et al.  Data science and prediction , 2012, CACM.

[22]  Rich Caruana,et al.  An empirical comparison of supervised learning algorithms , 2006, ICML.

[23]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.