Virtual Patient-Specific Quality Assurance of IMRT Using UNet++: Classification, Gamma Passing Rates Prediction, and Dose Difference Prediction.

Ying Huang,Sichao Fu,Jiyong Wang,Zhiyong Xu,Hao Wang,Yan Shao,Yanhua Duan,Yifei Pi,Kui Ma,Aihui Feng,Xiaojuan Miao,Weihai Zhuo,Hengle Gu,Ruxin Cai,Hua Chen

doi:10.3389/fonc.2021.700343

Abstract

The dose verification in radiotherapy quality assurance (QA) is time-consuming and places a heavy workload on medical physicists. To provide a clinical tool to perform patient specific QA accurately, the UNet++ is investigated to classify failed or pass fields (the GPR lower than 85% is considered “failed” while the GPR higher than 85% is considered “pass”), predict gamma passing rates (GPR) for different gamma criteria, and predict dose difference from virtual patient-specific quality assurance in radiotherapy. UNet++ was trained and validated with 473 fields and tested with 95 fields. All plans used Portal Dosimetry for dose verification pre-treatment. Planar dose distribution of each field was used as the input for UNet++, with QA classification results, gamma passing rates of different gamma criteria, and dose difference were used as the output. In the test set, the accuracy of the classification model was 95.79%. The mean absolute error (MAE) were 0.82, 0.88, 2.11, 2.52, and the root mean squared error (RMSE) were 1.38, 1.57, 3.33, 3.72 for 3%/3mm, 3%/2 mm, 2%/3 mm, 2%/2 mm, respectively. The trend and position of the predicted dose difference were consistent with the measured dose difference. In conclusion, the Virtual QA based on UNet++ can be used to classify the field passed or not, predict gamma pass rate for different gamma criteria, and predict dose difference. The results show that UNet++ based Virtual QA is promising in quality assurance for radiotherapy.

Highlights

Intensity-modulated radiation therapy (IMRT), a widely used treatment modality for cancer patients, provides highly conformal dose distribution to the target while sparing surrounding healthy tissues [1]
We proposed a novel quality assurance (QA) prediction model based on UNet + + using the planar dose distribution as input
The accuracy of the classification model was 95.79%; there were small root mean squared error (RMSE) (1.38-3.72) and mean absolute error (MAE) (0.82-2.52) between the measured and the predicted gamma passing rates (GPR) in the test set, and the trend and position of predicted dose difference were consistent with the measured dose difference

Summary

INTRODUCTION

Intensity-modulated radiation therapy (IMRT), a widely used treatment modality for cancer patients, provides highly conformal dose distribution to the target while sparing surrounding healthy tissues [1]. Previous prediction models based on machine learning or deep learning were only the results of dose verification but could not provide detailed information of dose difference [9,10,11,12,13,14,15,16,17,18]. The prediction model allows physicists to pre-mark potentially failed fields in a proactive way, analyze dose difference simultaneously and reduce patient delays associated with unqualified measurements. It could reduce patient-specific QA measurements to verify data transmission and delivery accuracy combination with other tools. The model is expected to be a practical clinical tool to perform patient-specific QA accurately and provide new ideas for the development of virtual QA and process optimization

MATERIALS AND METHODS

RESULTS

CONCLUSION

ETHICS STATEMENT