The Dosimetric Effect of Variations in Bolus Placement Identified via Cherenkov Imaging

Abstract

<h3>Purpose/Objective(s)</h3> The expanding use of Cherenkov imaging in radiation therapy (RT) clinics has presented the opportunity for a variety of applications. With the ability to visualize radiation dose on the surface of patients, Cherenkov imaging has emerged as a tool for detecting incidents in treatment delivery. With the use of transparent bolus, we can detect when a bolus has been misplaced during treatment setup. In this study, it is hypothesized that changes in treatment delivery due to bolus misplacement can be detected with Cherenkov imaging and verified with the treatment planning software (TPS). <h3>Materials/Methods</h3> Radiotherapy patients were imaged with a Cherenkov imaging system consisting of multiple time-gated, intensified CMOS cameras installed in the ceiling of two treatment bunkers at an academic center. All patients were treated according to their prescribed plan, and Cherenkov images were recorded for each fraction of their treatment. We investigated the dosimetric effect of misplaced bolus by mocking up the actual bolus positions for each case in the TPS. The dose distributions were then recalculated using an advanced dose calculation algorithm with 0.1 cm voxels, and compared to the original plan. <h3>Results</h3> Images from 622 patients were manually reviewed and a total of nine treatment incidents were detected. Two of these incidents involved variations in bolus placement from day to day during whole-breast RT, resulting in a lack of coverage for one or more beams during a single fraction of the patients' treatments. In Case 1, for a left breast treatment, the bolus did not cover the medial edge of the exit LPO field, resulting in a decrease in the minimum medial skin normalized dose in that area from 28.1% to 10.7% due to the lack of bolus backscatter into the skin. In Case 2, the lateral edge of a right breast RT treatment was left uncovered during the delivery of both the RPO and LAO fields, each contributing to a decrease in dose at the lateral edge. However, there was a slight increase in dose at the medial edge without the bolus due to a lack of beam attenuation at the lateral edge. Overall, the minimum normalized dose to the entire right breast decreased from 23.2% to 18.9%. <h3>Conclusion</h3> This first investigation into the effect of bolus misplacement detected via Cherenkov imaging demonstrated that these images can inform a mock-up of the actual treatment condition in the TPS for a dosimetric analysis. In comparison to the original plans with full bolus coverage, we confirmed that partial coverage results in a noticeable minimum dose reduction for relevant structures. This result validates that the lower Cherenkov emission seen from these fractions relative to the others can be at least partially attributed to the misaligned boluses, as Cherenkov intensity is proportional to dose. Ultimately, this method of treatment review is a valuable tool for radiation oncology quality assurance programs.