Technical Note: Three-dimensional QA of simultaneous integrated boost radiotherapy treatments by a dose-volume histogram methodology and its comparison with 3D gamma results

PURPOSE

Intensity-modulated radiotherapy with simultaneous integrated boost (SIB) presents several attractive advantages to be employed in clinical practice. Its secure application demands a rigorous quality assurance (QA) procedure, ideal for three-dimensional (3D) dose distribution measurements. Thus, a gel dosimetry methodology to evaluate the dose delivery of SIB treatments is presented and compared to conventional gamma evaluation.

METHODS

MAGIC-f gel dosimeter with magnetic resonance images for dose reading were used following its standard procedures. Four SIB QA plans created in gel dosimeter phantoms were used. The gel measured and treatment planning system (TPS) calculated doses were compared using 3D gamma analyses (3%/3mm/15% threshold). Two structures were artificially on the TPS dose distribution expected on the phantom by converting the 1.7 and 2.0 Gy isodose levels into structures to represent the treatment. The gel and TPS dose-volume histogram (DVH) were compared based on five dose points: D95%, D90%, D50%, D10%, and D5%.

RESULTS

Approvals of 93%, 96%, 98%, and 92% were achieved in the 3D gamma analyses for the plans QA 1, 2, 3, and 4. In the DVH analyses, QA plan 1 measured and expected curves showed a good agreement. QA plan 2 showed deviations in the highest doses for both structures with a maximum deviation (Δ_máx ) of 8.0%. QA plans 3 and 4 showed the highest dose variation between the gel and TPS in the smaller doses of the DVH (Δ_máx of 7.2% and -8.9%, respectively). For QA plan 4, the curves of the 1.7 Gy structure presented a good agreement, but deviations in the smaller dose region of the DVH occurred for the 2 Gy structure (Δ_máx of 7.7%).

CONCLUSIONS

A methodology for 3D dose evaluation of complex SIB treatments was proposed. It provided an important overview of the dose distributions. Their results significantly complemented the usual gamma analysis results.

Patient-specific IMRT QA verification using machine learning and gamma radiomics

Gamma function is the standard methodology for comparing dose distributions. It is calculated in dedicated software, and its results verification is not performed. Thus we developed an automatic tool for patient-specific QA results verification through high accuracy machine learning (ML) models based on the radiomics characteristics extraction from gamma images. We used 158 patient-specific QA tests and extracted 105 radiomics features from each gamma image. Three random forest models were developed (ML I, ML II, and ML III). ML I and ML II verified the gamma image approval using criteria of 2%/2mm/15% threshold and 3%/3mm/15% threshold, respectively. ML III verified if the gamma analyzes software recommended protocol was followed to detect if the TPS grid modification step was done. The models were based on the most important features selected using the mean decreased impurity, and their performances were evaluated. ML I included 25 features. Its accuracy was 0.85 using the test set and 0.84 using dataset B. ML II included 10 features, and its accuracy with the test set was 0.98; the same value was achieved using the never seen data (dataset B). The First-order 10th percentile feature was identified as a feature strongly related to the approved classification. ML III selected 23 features with an accuracy of 0.99 for test set and 0.98 for dataset B. An automatic workflow example for gamma analyses QA results verification could be proposed combining the models to detect grid inconsistencies on software evaluation, followed by the test approval classification.