Real time in-vivo dosimetry for eye, brain and head and neck proton irradiation

Principal Investigators: prof. dr. Coen Rasch (LUMC) and Dennis Schaart (TU Delft)

Funding: KWF 2020


Proton radiotherapy is not only more precise in depth but also more sensitive for changes in e.g. anatomy of the patient during subsequent days of treatment. Current practice in photon radiotherapy is to measure the exit dose and back-calculate the dose given to the patient. This procedure is called in-vivo dosimetry. Such a method does not exist for proton radiotherapy, although several options like Prompt gamma, Iono-acoustics, Cerenkov and PET imaging are under investigation. Prompt gamma visualization lacks currently sizeable detectors for verification.
Iono-acoustics potentially can detect the location of the Bragg peak but has not been tested in clinical setting yet. Cerenkov measurements uses emitted light as derivative for location of the Bragg peak, however in the presence of light absorbing human tissue is less likely to retain enough signal to be measured outside the patient. PET imaging outside the treatment room is feasible but has less desirable accuracy due to the loss of activation in time (half-life of the isotope), change in position of the patient, motion of the isotopes (blood stream) and summation including decay of activation in time after administration of the subsequent portions of the treatment bundle in the treatment room.

This project uses the proven principle of PET imaging but instantaneous inside the treatment room when the actual beam is delivered. The aim is to develop a detector allowing for patients with tumors in the vicinity of the skull and neck allowing to measure the dose on the spot enabling correction of the treatment. The proposed detector utilizes the short-term activation of the irradiated tissue in the patient (activation). By means of two opposite laterally placed PET-detectors the signal will be measured on the spot whilst radiation is given. An important advantage of this bilateral setup is that it allows the detectors to be placed closer to the patient allowing for 1/3 of the total signal to be analysed, as opposed to <10% with normal body-sized PET ring. The disadvantage is that the non-cylindrical system geometry introduces severe artefacts in the reconstructed images.

Thus, a crucial part of the innovation is the development of a PET detectors with sub-100 picosecond time resolution as well as parallax correction, enabling the elimination of such artefacts through the application of time-of-flight (TOF) image reconstruction. The signal measured with these detectors can be compared with the pre-treatment dose calculations allowing the location and the dose to be verified and if needed corrected. Now we can close the loop and verify the dose and location in the patient during radiation.