Improvement Using Planomics Features on Prediction and Classification of Patient-Specific Quality Assurance Using Head and Neck Volumetric Modulated Arc Therapy Plan

Purpose: To evaluate the utility of a new plan feature (planomics feature) for predicting the results of patient-specific quality assurance using the head and neck (H&N) volumetric modulated arc therapy (VMAT) plan. Methods: 131 H&N VMAT plans in our institution from 2019 to 2021 were retrospectively collected. Dosimetric verification for all plans was carried out using the portal dosimetry system integrated into the Eclipse treatment planning system based on the electronic portal imaging devices. Gamma passing rates (GPR) were analyzed using three gamma indices of 3%/3 mm, 3%/2 mm, and 2%/2 mm with a 10% dose threshold. 48 conventional features affecting the dose delivery accuracy were used in the study. 2476 planomics features were extracted based on the radiotherapy plan file. Three prediction and classification models using conventional features (CF), planomics features (PF), and hybrid features (HF) combining two sets of features were constructed by the gradient boosting regressor (GBR) and Ridge classifier for each GPR of 3%/3 mm, 3%/2 mm, and 2%/2 mm, respectively. The absolute prediction error (APE) and the area under the curve (AUC) were adopted for assessing the performance of prediction and classification models. Results: In the GPR prediction, the average APE of the models using CF, PF and HF is 1.3±1.2%/3.6±3.0%, 1.7±1.5%/3.8±3.5%, and 1.1±1.0%/4.1±3.1% for 2%/2 mm, and 0.7±0.6%/2.0±2.0%, 1.0±1.1%/2.2±1.8%, and 0.6±0.6%/2.2±1.9% for 3%/2 mm, and 0.4±0.3%/1.2±1.2%, 0.4±0.5%/1.3±1.0%, and 0.3±0.3%/1.2±1.1% for 3%/3 mm, respectively. In the regression prediction, three models give a similar modeling performance for predicting the GPR. The classification results are 0.67±0.03/0.66±0.07, 0.77±0.03/0.73±0.06, and 0.78±0.02/0.75±0.04 for 3%/3 mm respectively. For 3%/2 mm, the AUC of training and testing cohorts are 0.64±0.03/0.62±0.07, 0.70±0.03/0.67±0.06, and 0.75±0.03/0.71±0.07 respectively. And for 2%/2 mm, the average AUC of the training and testing cohorts are 0.72±0.03/0.72±0.06, 0.78±0.04/0.73±0.07, and 0.81±0.03/0.75±0.06 respectively. In the classification, the PF model has a better classification performance than the CF model. And the HF model provides the best result among the three classifications models. Conclusions: The planomics features can be used for predicting and classifying the GPR results, and improve the model performance after combining the conventional features for the GPR classification.