Single proton LET characterization with the Timepix detector and artificial intelligence for advanced proton therapy treatment planning

Protons have advantageous dose distributions and are increasingly used in cancer therapy. At the depth of the Bragg peak range, protons produce a mixed radiation field with low- and high-linear energy transfer (LET) components, where the high-LET is associated with increased biological effectiveness. LET prediction by Monte Carlo simulations is challenging to verify experimentally. This study presents measurements using the Timepix detector in a mixed radiation field produced by a therapeutic proton beam in water, validated by Monte Carlo simulations. The Timepix detector, with high-resolution particle tracking enhanced by artificial intelligence, identified particle types and measured deposited energy. Biologically relevant LET parameters, including single proton LET and dose-averaged LET, were computed. The neural network achieved over 95% accuracy in proton recognition. The LET spectra from measurements matched simulations with a mean dose-averaged LET difference of 17%. A wide LET range from sub-keV/μm to 10 keV/μm was observed. The proposed method offers accessible, experimental LET data for treatment planning validation and can be implemented in clinical proton therapy facilities.

Text supplied in this site's publication record. Consult the full paper for methods, experimental conditions and limitations.