ESTRO 2024 - Abstract Book

S362

Beachytherapy - Physics

ESTRO 2024

A dose calculation tool based on the TG43-approach was developed to calculate the dose delivered by every single source position. Using patient contours, clinical dose metrics of the tumour and OARs can be calculated after each dwell, such as PTV-D90%, PTV-V100 or Urethra-D10%. By analyzing the evolution of these clinical dose metrics, it can be assessed whether a treatment deviation is happening. Moreover, using the planned dose that is anticipated from the remaining dwells, it can be determined immediately after every dwell whether clinical goals are still expected to be met at completion of the treatment. A realistic prostate HDRBT treatment was created using contours of a real patient ultrasound acquisition registered to the CT of an anthropomorphic 3D printed pelvic phantom (see figure 1). The plan consists of 124 dwell positions divided over 17 different catheters, with dwell times ranging from 0.6 to 6 seconds, with an average of 3.4 seconds. The treatment was delivered to the phantom during which the source was tracked using an imaging panel-based method, which has an uncertainty in finding the position of the source of 0.83 ± 0.41 mm. To test the effectiveness of the proposed method, catheter 3 (dwell positions 16-22) was shifted by 3 mm to simulate an error.

Results:

Figure 2 shows the evolution of the PTV-D90% after every dwell position. The top row shows how the source tracking uncertainty of the measurement system causes the measured PTV-D90% to deviate from the PTV-D90% calculated using the planned positions. Because the uncertainty of the source tracking system is random, after all 124 dwell positions are calculated it ends up very close to the planned value. The shift applied to the measured values in catheter 3 (dwell position 16-22) is systematic, and causes the value for PTV-D90% to deviate. In this case the final value is lower than the expected outcome, suggesting the PTV would receive around 0.2 Gy less dose than originally planned due to this catheter shift. In the bottom row can be seen that, by using the planned position of the dwells that have not occurred yet, the final value for PTV-D90% can already be predicted immediately after the last dwell position (dwell 22) of the shifted catheter.