Student Number : 0000713F - PhD thesis - School of Physics - Faculty of Science / The Doppler shift and the Doppler broadening of prompt gamma emissions were measured for some residues formed in the interaction of 33 A MeV 12C ions with a 63Cu target using the AFRODITE detector array at Faure, Cape Town. This is a potentially new technique to carry out nuclear interaction studies. Coincident gamma rays emitted by the residues are used in their identification. Detection at angles other than 90◦ with respect to the beam axis gives the magnitude of the mean Doppler shifts and the average linear momentum transfer. The Doppler broadening of the detected gamma lines at 90 ◦ with respect to the beam axis could give the residue recoil angular distribution. The precise shapes of the Doppler shifted and broadened gamma lines for each of the residues extracted, reveals the distribution, in magnitude and angle, of the momentum transferred in the interaction process. In addition, characteristic gamma energy transitions of each residue populated carry additional information on angular momentum (spin) transfer, production cross-section and nuclear excitation states. The measured residues show a unique distribution of momentum ranging from single nucleon transfer to complete damping of the projectile momentum. The measured observables are consistent with the existing data from other techniques, making the new technique viable option for studying nuclear interaction kinematics. A comparison of the experimental measurements with the predictions of the model developed in Milano 1 and GEANT4 calculations shows that the model developed in Milano model give a much better agreement compared to the GEANT4 calculations, attributed to the assumption of projectile break-up and re-emission process of some of the fragments during the first step of the nuclear interaction process.

L’imagerie TEP repose à l’heure actuelle sur des détecteurs à base de cristaux scintillants ou de semi-conducteurs. Le projet CaLIPSO propose de tirer parti à la fois de l’émission de lumière (par effet Cerenkov) et de l’ionisation du milieu de détection, pour réaliser un détecteur de résolution temporelle et spatiale améliorée. Le milieu de détection, le TriMéthylBismuth liquide à température ambiante, de par sa forte teneur en Bismuth, permet une détection par effet photoélectrique efficace. Cette étude a pour objectif de concevoir et d’optimiser le détecteur optique du projet CaLIPSO, afin de prouver le concept de la détection de photons de 511 keV par effet Cerenkov dans le TMBi, et de caractériser les performances d’un tel détecteur en termes de résolution temporelle et efficacité de détection. Un premier démonstrateur a validé le principe de détection reposant sur l’effet Cerenkov, bien que ses performances soient décevantes. C’est la raison pour laquelle nous avons entrepris un effort d’optimisation en simulation Monte Carlo dans Geant4, afin d’améliorer la collection de la lumière Cerenkov dans le détecteur, et donc son efficacité de détection et sa résolution temporelle. Avant, nous avons mesuré les propriétés optiques du TMBi (indice de réfraction, absorption et diffusion de la lumière), afin d’être capables de modéliser la propagation de la lumière Cerenkov dans le détecteur. Nous avons également optimisé par simulation Monte Carlo l’outil permettant la mesure de résolution en temps, un cristal scintillant de YAP:Ce couplé à un PMT. Cela a permis une mesure plus fine de la résolution temporelle du démonstrateur. A l’issue de ces travaux, nous avons construit un second démonstrateur optique. On mesure alors une efficacité de détection de l’ordre de 32% pour une résolution en temps de 660 ps (FWHM). L’efficacité mesurée prouve que le détecteur est pleinement efficace à détecter les conversions photoélectriques du photon de 511 keV (27% des photons incidents). Plusieurs optimisations technologiques sont proposées pour améliorer la résolution temporelle, et espérer à l’avenir une mesure du temps de vol des photons gamma. / PET detectors are usually based on scintillation crystals or semiconductor materials. The CaLIPSO project aims to build a PET detector working on the double detection of Cerenkov light and pair productions in a novel detection material called TriMethylBismuth. This would allow at the same time an enhanced time resolution (thanks to the Cerenkov signal) and a excellent spatial resolution (thanks to the ionization signal). Liquid TMBi (at room temperature), thanks to its good photo fraction (47%), allows a good detection efficiency, principally by photoelectric effect. In this context, this work aims to design and optimize an optical detector as a proof of concept for the Cerenkov detection of 511 keV gamma photons, and to measure the time resolution and detection efficiency of such a detector. The optical signal based on Cerenkov effect in TMBi has been observed on a first demonstrator, but its performances were clearly inappropriate. So we used a Monte Carlo simulation (Geant4) of the detector in order to model the relevant phenomena and to optimize de detection. It appeared that light collection efficiency in the detector was the most important parameter to optimize so as to improve time resolution and detection efficiency. Before that, we measured TMBi optical properties (refractive index, light absorption and diffusion), in order to model accurately the Cerenkov light propagation in the detector. The tool used for the time resolution measurement is a YAP: Ce scintillator coupled to a PMT. We also needed to optimize this tool in order to allow a more accurate measurement of the detector time resolution. At the end of this work, a second version of the optical demonstrator was built. We measured a detection efficiency of 32%, and a time resolution of 660 ps FWHM. The measured efficiency proved that our detector is fully efficient to detect the photoelectric conversions of the 511 keV photons (27% of the incident photons). Several technological optimizations are proposed to further improve the time resolution, in order to be able to measure the gamma photons’ time-of-flight in the future.

Lungscintigrafi är en av de mest vanligt förekommande nuklearmedicinska undersökningarna. Akut lungemboli är en huvudindikation för lungscintigrafi. Undersökningen består av två delar, först undersöks ventilationsdelen och sedan perfusions-fördelningen i lungan. Vid ventilationen inhalerar patienten 99mTc-Technegas. Vid perfusionen injiceras 99mTc-Pulmocis (MAA) intravenöst. Enligt EANM ligger den administrerade aktiviteten mellan 25 – 30 MBq för Technegas och mellan 140 – 160 MBq för Pulmocis. Studien utfördes på avdelningen klinisk fysiologi vid Centralsjukhuset Kristianstad. Först utfördes en fantomstudie före patientstudien för att testa metoden. Den inkluderade totalt 20 patienter som hade en bokad tid för lungscintigrafi. Syftet med studien var att optimera metoden för lungscintigrafi genom att reducera patientstråldosen genom att uppskatta aktiviteten som administreras vid ventilation. Vidare var syftet att kontrollera om kvoten mellan ventilation- och perfussionsundersökningen överensstämmer. Detta utfördes med hjälp av två metoder: den ena var en gammadetektor som mätte antalet counts som patienten har andats in och den andra var med att kvantifiera bilderna efter en slutförd undersökning. Resultaten över aktiviteten i ventilationsbilderna i Xeleris Software har erhållits med en beräkning. Perfusionen beräknades genom att mäta sprutans aktivitet före och efter undersökningen. Genom dessa värden kunde kvoten mellan perfusion och ventilation erhållas. Medelvärdet för alla patienter visar att kvoten blir fyra gånger så hög i perfusion som i ventilation. Resultaten över mätmetoderna visade att gammadetektorn har samma effektivitet som kamerans detektorer. Det erhölls en positiv korrelationskoefficient mellan metoderna och den låg på 0,79. Med Bland-Altmans diagram var 95% av värden för båda metoderna inom referensramarna förutom ett extremvärde som berodde på en hög inhalerad aktivitet. Slutsatsen blir att mätmetoden kräver en större studie med fler deltagare för att kunna implementera mätningen med gammadetektorn i kliniken. / Pulmonary scintigraphy is one of the most common nuclear medicine examinations. Acute pulmonary embolism is a major indication for pulmonary scintigraphy. The examination consists of two parts, first the ventilation part and then the perfusion distribution in the lung is examined. During ventilation, the patient inhales 99mTc-Technegas. At perfusion, 99m Tc-Pulmocis (MAA) is injected intravenously. According to EANM, the administered activity is between 25 - 30 MBq for Technegas and between 140 - 160 MBq for Pulmocis. The study was performed at the department of Clinical physiology at Central Hospital Kristianstad. A phantom study was first performed before the patient study to test the method. It included a total of 20 patients who had an appointment for pulmonary scintigraphy. The aim of the study was to optimize the method of pulmonary scintigraphy by reducing the patient radiation dose by estimating the activity administered during ventilation and perfusion. Furthermore, the aim was to check whether the ratio between the ventilation and perfusion examination is correct. This was done using two methods: one was a gamma detector that measured the number of counts that the patient has inhaled, and the other was involved in quantifying the images after a completed examination. The results of the activity in the ventilation images in Xeleris Software have been calculated. Perfusion is calculated by measuring the activity of the syringe before and after the examination. Through these values, the ratio between perfusion and ventilation could be obtained. The average value for the ratio and diagrams showed that the ratio is four times as much in perfusion as in ventilation. The results of the measurement methods showed that the gamma detector has the same efficiency as the camera detectors. A correlation coefficient was maintained between methods that indicated a good significance. With Bland-Altman's diagram, 95% of the values ​​for both methods were within the frame of reference except for an extreme value that was due to a high inhaled activity. The conclusion is that the measurement method requires a larger study with more participants to be able to implement the measurement with the gamma detector in clinical practice.

Current positron emission mammography (PEM) depth of interaction (DOI) enabling detectors have low scintillator to photodetector encoding ratios, RE causing high system complexity and cost. The modularized dual-ended readout block (DERB) detector combines the Anger logic block detector with dual-ended readout to increase RE while measuring DOI. To investigate the trade-off between RE and spatial resolution, scalable DERB detectors with varying RE and light guide thickness were modelled with Monte- Carlo. Simulation showed RE can increase up to six-fold compared to the dual-ended readout design without significantly degrading spatial resolution. Experimental characterization of a RE = 9 : 8 DERB detector was found to achieve super-resolution <0.5 mm for resolving crystal indices, DOI resolution of ~5 mm FWHM, and mean energy resolution of 20% without recovering photons lost to neighbouring detector modules. The model was validated by agreement of simulation results adjusted for detector quantum efficiency with experimental results.

Current positron emission mammography (PEM) depth of interaction (DOI) enabling detectors have low scintillator to photodetector encoding ratios, RE causing high system complexity and cost. The modularized dual-ended readout block (DERB) detector combines the Anger logic block detector with dual-ended readout to increase RE while measuring DOI. To investigate the trade-off between RE and spatial resolution, scalable DERB detectors with varying RE and light guide thickness were modelled with Monte- Carlo. Simulation showed RE can increase up to six-fold compared to the dual-ended readout design without significantly degrading spatial resolution. Experimental characterization of a RE = 9 : 8 DERB detector was found to achieve super-resolution <0.5 mm for resolving crystal indices, DOI resolution of ~5 mm FWHM, and mean energy resolution of 20% without recovering photons lost to neighbouring detector modules. The model was validated by agreement of simulation results adjusted for detector quantum efficiency with experimental results.