Calibrazione del detector QUAD TimePix3 per la misura di concentrazione e distribuzione spaziale di 10B in campioni biologici

Boron Neutron Capture Therapy (BNCT) is a hadron therapy with a selectivity acting at the single cell level with characteristics that are complementary to proton an carbon-ion therapy. BNCT uses the neutron capture reaction on 10B to produce two energetic ions that release all of their energy approximately within the borders of the single cell where the reaction takes place. To exploit the therapeutic selectivity, the 10B enriched capture agent must be loaded by tumor cells thanks to dedicated vectors. In this scenario, it is important to quantify the selective accumulation of 10B in tissues in order to properly link the delivered dose to the biological effects both on the healthy and carcinogenic tissues. For this purpose, several techniques are currently being used and developed for 10B concentration and spatial distribution measurement. The TimePix3 is a hybrid pixelated detector that allow to simultaneously carry out a measurement of Time of Arrival (ToA) and Time over Threshold (ToT) with high spatial resolution, dictated by the pixel pitch of 55 μm. This allows to detect and identify the particles produced by the 10B neutron capture reaction using PID techniques. Furthermore, it is possible to juxtapose several TimePix3 chips to allow for a larger sensitive area. Such a configuration has been used in this thesis work in the form of a QUAD TimePix3 detector, a 2×2 array of TimePix3 chips. Pilot and promising information about TimePix3 capability of working in a neutron field for BNCT as well as imaging capabilities on 10B enriched samples have been collected in previous measuring campaigns. In this framework, the purpose of this thesis is to explore the possibility of using a QUAD TimePix3 detector to simultaneously measure concentration and visualize spatial distribution of 10B in organic matrices at low concentrations, compatible with those expected in clinics. To do this, two sets of measurements were carried out at the Laboratorio di Energia Nucleare Applicata (LENA) of the university of Pavia. The first in the Prompt Gamma Neutron Activation Analysis (PGNAA) facility with a highly thermalized and collimated neutron beam. In this facility, it was performed a first attempt at the calibration curve, but due to the relatively low neutron flux and the limits to perform a long irradiation for each of the samples, it was not possible to get a satisfactory calibration curve. Nonetheless, this set of measurements was useful as a confirmation of the imaging capability of the detector and as a qualitative assessment of the linear response of the 10B concentration range 0-100 ppm. For the second set of measurements, the experiment was transferred in the thermal column of the reactor. In it, the neutron flux can reach 10^3 times more than in the PGNAA facility, but the beam is not collimated. The measurements taken in the thermal column allowed to get a preliminary, but reliable calibration curve for the analysis of thin tissue samples deposited on Mylar supports and expected to contain an unknown amount of 10B. The raw results of the calibration curve obtained in the LENA thermal column required a careful statistical analysis of the data as well as of the detection system. To support the analysis, Quantitative Neutron Autoradiography was performed on the samples used for the calibration, in particular to confirm their microscopic features as 10B standards. Finally, this calibration was then tested against biological samples containing unknown concentrations of 10B. The 10B concentrations that the TimePix3 technique yields are compatible with the results obtained on twin samples using another consolidated technique such as ICP-OES. The comparison with twin samples measured with Alpha Spectrometry yielded lower concentration values than expected, but this can be imputed to the state of conservation of the sample.