%0 Generic %A Hartmann, Bernadette %D 2013 %F heidok:15690 %R 10.11588/heidok.00015690 %T A Novel Approach to Ion Spectroscopy of Therapeutic Ion Beams Using a Pixelated Semiconductor Detector %U https://archiv.ub.uni-heidelberg.de/volltextserver/15690/ %X In highly precise carbon ion radiotherapy, fragmentation of the primary nuclei in the patient results in a spectrum of lighter ions. Due to their radio-biological effectiveness being different from the primary ions, they need to be considered separately in therapy planning. To determine secondary ion spectra, mainly large apparatus based on scintillation detectors have been used until now, limiting the flexibility of the methods and the amount of available data. In this thesis, a novel method for ion spectroscopy based on a small pixelated semiconductor detector is presented. The used Timepix detector, originally designed for photon beam imaging, offers a high spatial resolution enabling the detection of single particles. At first, an extensive characterization of the detector response to therapeutic ion beams was performed. The detector was found suitable for energy-loss measurements on a single ion basis in proton beams between 0.55 and 221MeV, providing mean energyloss values, which deviate less than only 10% from calculations. For the investigated heavier ions, deviations of up to about 30% were observed. The presented novel experimental approach to fragment distinction with the Timepix detector is based on pattern recognition analysis of the signal created by individual ions. For designated configurations, it enables identification of all ion species in mixed particle fields resulting from 12C-fragmentation. The performance of the method was evaluated using reference data of an established technique. The relative fractions of H-, He-, Be- and B-ions agree within 1.1Δref (uncertainty of the reference). For lithium, the agreement is within 2.3Δref. In addition, applications of the method relevant for benchmarking physical models used in Monte Carlo simulations and treatment planning are presented. Providing the advantages of a small and flexible set-up, together with the further improvements suggested, the method is promising to widely expand the available fragmentation data and to complement large experimental set-ups.