Data-Driven Optimization of Nanoparticles for Radiotherapy via Scientific Literature Mining and Machine Learning

This thesis focuses on the application of data mining, statistical analysis, and machine learning techniques to support the design of nanoparticles for radiotherapy applications. The initial phase involved the collection and analysis of scientific literature, from which relevant information was extracted and organized into a structured dataset. Statistical analysis was performed on this dataset to investigate potential relationships between nanoparticle composition, structural features, and their functional properties relevant to radiotherapy. Building on this foundation, a machine learning model was developed to predict the performance of nanoparticles for radiotherapy purposes, and an optimization strategy based on a genetic algorithm was employed to identify formulations with favorable properties under specific conditions. As a complementary experimental component, Au-based nanoparticles were synthesized via pulsed laser ablation in liquid and functionalized, with the aim of providing an empirical reference for the developed computational model. The synthesized nanoparticles were characterized using appropriate experimental techniques to assess their functional properties.