• dose
• electrons
• flash
• online
• radiation
• radionuclide
• radiotherapy

Ultra-high dose rate irradiations, particularly for inducing the FLASH effect, have gained significant attention due to reduced toxicity in normal tissues compared to conventional rates. Recent advancements suggest using VHEE, delivered in ultra-short pulses at ultra-high dose rates, as a potential method for implementing FLASH radiotherapy (RT) for deep-seated tumors. A major hurdle in translating VHEE-FLASH RT into clinical practice is the development of real-time, online monitoring techniques to optimize treatment delivery.This thesis introduces a novel online monitoring method based on detecting prompt gamma rays emitted during the electromagnetic showers from VHEE interactions with matter. Additionally, the use of β+ emitting radionuclides generated by photonuclear reactions in a target, detectable via PET, is proposed as a complementary offline monitoring technique.Monte Carlo simulations were conducted to evaluate the dose delivered to a PMMA phantom and optimize the monitoring system design. A Normalizing Flow-based generative model neural network was trained to predict dose distributions from detected prompt gamma rays. Experimental measurements at the INFN BTF and the CNR-INO confirmed the feasibility of prompt gamma ray detection for real-time dose measurement and, when combined with the neural network model, for online dose verification. The β+ emitting radionuclides were validated as an effective complementary offline monitoring technique.