Room 202
Purpose: Proton therapy is planned using CT images assuming unchanged anatomy throughout the treatment. This assumption does not always hold causing tumor underdosing and overdosing of neighboring tissue. Therefore, a radiographic modality using the therapeutic beam itself is desired.
Methods: Absorption of short proton pulses in tissue leads to thermoelastic expansion producing protoacoustic waves, sensed by ultrasonic transducers placed around the irradiated tissue. Using acoustic reconstruction algorithms, the map of proton energy deposition (PED) in the tissue can be estimated from the collected protoacoustic signals. We used the TOPAS Monte-Carlo model to generate the PED in water with five Aluminum cylinders with (0.5mm–5mm radii) in the beam-eye-view at the Bragg peak (BP) (~14.9mm for 40MeV protons). The PED map was then exported to the k-wave toolbox which was used to simulate protoacoustic propagation and the signals were collected at a 128-element-ring array in the BP plane. Time-reversal (TR) was then employed to reconstruct the PED map in the BP plane.
Results: While the PED map was reasonably reconstructed, the off-center targets having received the scattered proton fields have poor contrast with respect to the central target which receives the main proton field. This was corrected by normalizing the reconstructed PED with the simulated PED map at the BP in water (without targets). This was computationally achieved using pointwise matrix division conjugated with Laplacian regularization. The normalization significantly improved the contrast of the off-center targets.
Conclusion: The preliminary results demonstrate that the proposed protoacoustic radiography can provide imaging using the therapeutic proton beam during proton therapy. This can assist in correcting for the unexpected changes in anatomy, or misalignments with respect to the proton beam can cause overdosing of healthy tissues surrounding the tumor and/or underdosing of the tumor volume.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the National Institute of Health (R37CA240806), American Cancer Society (133697 RSG 19 110 01 CCE). The authors would like to acknowledge the support from UCI Chao Family Comprehensive Cancer Center (P30CA062203)
Not Applicable / None Entered.
Not Applicable / None Entered.