ESTRO 37 Abstract book

ESTRO 37

S466

previously validated using experimental water tank measurements.

Results Fig. 2a displays the acoustic dose image and the overlaid dose distribution. The yellow line indicates the location of the acoustic and dose profiles displayed in Fig. 2b. It is observed that each dose gradient produces a bipolar peak in the acoustic profile. The dashed black boxes in Fig. 2b demonstrate how the negative portion of each bipolar peak aligns with the beginning of the dose gradient and that where the acoustic profile returns to zero the dose profile has reached its maximum value. Hence, the acoustic images can be used to infer the locations of dose gradients in the patient received dose distribution, and therefore verify that dose is being delivered to the desired location. A set-up error of 3 mm was introduced and found to temporally shift the induced acoustic signals by 2.1 µs. This shift was reflected in the reconstructed acoustic dose images, and corresponded within 0.2 mm to the introduced patient set-up error of 3 mm.

Table 1: Gamma comparison of dose distributions between EBT3 film and TPS using 4 algorithms to determine dose from the film: (1) Red channel only, (2) Méndez (truncated normal distribution) algorithm, (3) Pérez-Azorín algorithm and (4) a Hybrid algorithm of both Méndez and Pérez-Azorín. Each gamma comparison was made using 3%/3mm and 2%2mm criteria. Conclusion The red channel based dosimetry is inadequate. The three-channel algorithms give good results, however the Pérez-Azorín’s algorithm seems to be more robust, and is capable of obtaining a very close dose map when compared to the TPS. The effects of noise have been minimised with the use of a median filter and matching of image res olutions PO-0882 Detecting radiation-induced acoustic waves with a transperineal transducer for in vivo dosimetry S. Hickling 1 , M. Hobson 1 , I. El Naqa 2 1 McGill University Health Center- Cedars Cancer Centre, Medical Physics, Montreal, Canada 2 University of Michigan, Radiation Oncology, Ann Arbor, USA Purpose or Objective Acoustic pressure waves with properties related to the dose deposited in an object are induced following a pulse of linac irradiation. It is hypothesized that detecting these acoustic waves during radiotherapy could be used for in vivo dosimetry. This work aims to demonstrate through simulations that useful dosimetric information can be obtained in vivo using a transperineal ultrasound transducer (TPUS) to detect acoustic waves-induced during radiotherapy treatment for a prostate case. Material and Methods The computed tomography (CT) scan and dose treatment plan for a prostate patient undergoing volumetric modulated arc therapy (VMAT) were obtained. Using an in-house simulation workflow combining Monte Carlo and acoustic wave transport methods, the acoustic waves induced at a given control point were modelled to determine the time-varying acoustic signals detected by a TPUS transducer. The simulated signals at each transducer element were then filtered to account for realistic transducer bandwidth characteristics. These signals were back-projected from the transducer location assuming a constant speed of sound to form an acoustic dose image. Fig. 1 shows the set-up geometry of simulations performed in a sagittal slice for a lateral beam. The simulation workflow used in this work was

Conclusion Since acoustic signals are induced by a single pulse of irradiation and reconstruction is rapid, this could be a real-time in vivo dosimetry technique for monitoring treatment delivery. Intrafractional ultrasound imaging is becoming increasingly investigated as a tool to monitor target motion during treatment. A system capable of obtaining both anatomical ultrasound images and detecting radiation-induced acoustic waves with the same transducer could be invaluable as a treatment monitoring technique to allow clinicians to visualize where dose is being delivered during treatment in real time.