[18F]Fluorodeoxyglucose ([18F]FDG) Positron Emission Tomography/Computed Tomography (PET/CT) imaging is a powerful tool in the diagnosis of cancer and subsequent treatment planning. New state-of-the-art PET/CT scanners have the capacity to generate images of superb quality. The new scanners feature detectors with increased sensitivity and a new generation of reconstruction algorithms that produce higher quality images than the scanners they are replacing. In addition to the scanner, the scan duration, amount of administered [18F]FDG activity, and the anatomy of the patients themselves are also determining factors of image quality. There is evidence suggesting that [18F]FDG PET image quality is significantly reduced for larger patients, jeopardizing lesion detection. Two possible solutions to this problem are to (i) increase injected activity or (ii) increase scan duration. Increasing scan duration is preferable but not always possible in a busy clinic. Increasing injected activity is necessary but a proper scaling regimen with patient size must be determined in order to achieve consistent image quality. The aim of the work presented in this thesis was to achieve higher quantification accuracy and consistent image quality for all patients scanned with [18F]FDG PET. Because quantitative PET/CT images require corrections for image degrading effects, for which attenuation correction is the main contributor and is performed based on CT images, the first step in this project was to develop software tools to automate the analysis of phantom images for CT quality assurance. The next step was to optimize the reconstruction parameters for whole body [18F]FDG PET based on a phantom experiment. Finally, a retrospective study was conducted using patient [18F]FDG PET images to characterize the relationship between patient anatomical characteristics and image quality. This work concludes by suggesting optimized reconstruction parameters and a scaling regimen for injected [18F]FDG activity. With the implementation of these recommendations it can be possible to obtain images with increased quantitative value while delivering less dose to patients and maintaining a uniform level of image quality between different patients. / Graduate

Right ventricular function (RV) is one of the most important predictors of prognosis in many cardiovascular disease states. Despite the significance of RV function to survival, there are no therapies that directly nor selectively improve RV function. As well, the basis for RV failure is poorly understood. This is particularly relevant for patients with pulmonary arterial hypertension (PAH), where RV failure in the setting of pressure overload is the leading cause of death. PAH will be introduced in the 2nd chapter of this thesis by comparing and refining contemporary mortality risk assessment strategies. I will then explore 1) RV neurohormonal function and, 2) RV energetics, two molecular pathways thought to be involved in the pathogenesis and progression of maladaptive RV failure. I employed small animal molecular imaging using positron emission tomography (PET) to non-invasively investigate these pathways. The PET imaging techniques employed in this thesis have the unique potential for translation to human studies, to further explore disease mechanisms.

A lo largo de la última década hemos asistido a un desarrollo sin precedentes de las tecnologías de la salud. Los avances en la informatización, la creación de redes, las técnicas de imagen, la robótica, las micro/nano tecnologías, y la genómica, han contribuido a aumentar significativamente la cantidad y diversidad de información al alcance del personal clínico para el diagnóstico, pronóstico, tratamiento y seguimiento de los pacientes. Este aumento en la cantidad y diversidad de datos clínicos requiere del continuo desarrollo de técnicas y metodologías capaces de integrar estos datos, procesarlos, y dar soporte en su interpretación de una forma robusta y eficiente. En este contexto, esta Tesis se focaliza en el análisis y procesado de señales biomédicas y su uso en problemas de clasificación automática. Es decir, se focaliza en: el diseño e integración de algoritmos para el procesado automático de señales biomédicas, el desarrollo de nuevos métodos de extracción de características para señales, la evaluación de compatibilidad entre señales biomédicas, y el diseño de modelos de clasificación para problemas clínicos específicos. En la mayoría de casos contenidos en esta Tesis, estos problemas se sitúan en el ámbito de los sistemas de apoyo a la decisión clínica, es decir, de sistemas computacionales que proporcionan conocimiento experto para la decisión en el diagnóstico, pronóstico y tratamiento de los pacientes. Una de las principales contribuciones de esta tesis consiste en la evaluación de la compatibilidad entre espectros de resonancia magnética (ERM) obtenidos mediante dos tecnologías de escáneres de resonancia magnética coexistentes en la actualidad (escáneres de 1.5T y de 3T). Esta compatibilidad se evalúa en el contexto de clasificación automática de tumores cerebrales. Los resultados obtenidos en este trabajo sugieren que los clasificadores existentes basados en datos de ERM de 1.5T pueden ser aplicables a casos obtenidos con la nueva tecnolog / Fuster García, E. (2012). Biomedical signal analysis in automatic classification problems [Tesis doctoral]. Editorial Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/17176 / Palancia

Proton therapy, a relatively young modality in radiation therapy, has proven useful in cases where a sharp dose gradient or low secondary irradiation is required. In Canada proton therapy it was performed at the TRIUMF Proton Therapy Facility in the treatment of large or difficultly positioned ocular melanomas. This rare primary malignant cancer of the eye has a poor prognosis if untreated. Patient vision sparing is critical for quality of life and is strongly affected by the accuracy of the chosen treatment. Reduction in irradiation of critical structures is a proven strength of proton therapy due to the high dose-gradient and finite range in tissue. But, with the advantage of steep dose gradients, comes the requirement of precision target positioning and planning. Monte Carlo particle transport software is a valuable tool for understanding treat- ment doses in cases where measurement is time consuming or difficult. Accurate simulation of primary proton dose to water aids in the evaluation of beam charac- teristics and allows for study into improving dose application for patient treatment. In this work, a full Monte Carlo model of the TRIUMF proton therapy facility was developed. Measurements were taken in water to validate simulated results within 2% over the treatment depth for a wide range of beam modulations. The second advantage of proton therapy lies in its reduced dose bath to healthy tissue. This is especially important in pediatric cases where extraneous dose comes with a high risk of secondary carcinogenesis. Whereas multi-angle photon treatments necessarily irradiate large volumes of healthy tissue to produce a flat target dose, proton treatments may irradiate a target with a single beam. With this advantage comes a trade-off - protons produce a large number of neutrons as they are prepared for patient treatment. These neutrons are the largest contributor to secondary dose in proton therapy and must be well modeled and shielded to ensure patient safety. The second part of this work involves the measurement of secondary neutron doses in the TRIUMF treatment room. Measurements were validated within 20% of simulated values with uncertainties dominated by calibration of the detector. Neutron doses to an anatomic human model showed that calibrated secondary doses were in line with similar treatment facilities reporting globally. Simulations indicated that the source of neutrons was primarily in the unshieldable region of the beamline opening. Thus the total treatment time was the determining factor in secondary dose to the patient. With primary proton dose well modeled, it became possible to study the pre- cision of treatment and possible avenues for improvement. The beam modulation wheels and optimization scheme was developed in the late 90‘s when computational and manufacturing technologies were less developed. Updated optimization methods indicated that moving to a smooth scheme of energy modulation, as opposed to a stepped modulation wheel, could improve distal dose sharpness. This was contrary to the long-held belief that there was an optimal number of steps for modulation. The third portion of this work explored the use of 3D printers to enable the fabri- cation of smoothly transitioning modulator wheels. Materials and printer methods were studied, indicating a strong candidate in the PolyJet TM method for beam mod- ulation. Both stepped and newly-optimized smooth modulator wheels were printed and validated. Total turnaround time for modulator production was under 24 hours - proving the feasibility of patient-specific beam modulation. The last portion of this work explored the use of positron emitting isotopes for dose validation. Protons traversing tissue or plastic generate β + emitting isotopes via nuclear interactions. The resulting back-to-back annihilation photons can be re- constructed into the isotope distribution produced by the beam. This can potentially provide information about beam position in the target and hence position of a phan- tom or patient. An anatomic 3D printed eye phantom was designed and irradiated to test the feasibility of this method. While a strong isotope signal was reconstructed, the test did not yield a viable technique due to the low resolution of the phantom scan. The phantom position was poorly reconstructed using the transmission scan. Despite this, it could be possible to improve this method by using other methods for phantom position registration. / Graduate

Sarcoidosis is a multisystem granulomatous disease commonly involving the lungs and mediastinal lymph nodes with the exact etiology being unclear. The simultaneous presence of malignant disease such as breast cancer and sarcoidosis has been reported. Sarcoidosis preceding a diagnosis of malignancy and that occurring years after treatment of malignant disease has been noted in the past. The presence of sarcoidosis in the setting of malignant disease carries a high risk of misdiagnosis. In this article, we report the case of a 45-year-old female with stage IA invasive ductal carcinoma of left breast that was in remission for 2 years; however, radiological imaging including magnetic resonance imaging of thoracic spine and positron emission tomography–computed tomography scanning were highly suspicious for malignant disease metastasis versus lymphoma with the widespread lymphadenopathy. Multiple tissue biopsies with histopathological evaluation allowed us to definitively exclude malignant disease metastasis and to correctly diagnose her atypical presentation of sarcoidosis.

京都大学 / 0048 / 新制・課程博士 / 博士(薬学) / 甲第18550号 / 薬博第812号 / 新制||薬||238(附属図書館) / 31450 / 京都大学大学院薬学研究科医療薬科学専攻 / (主査)教授 佐治 英郎, 教授 橋田 充, 教授 髙倉 喜信 / 学位規則第4条第1項該当 / Doctor of Pharmaceutical Sciences / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(社会健康医学) / 甲第18907号 / 社医博第65号 / 新制||社医||9(附属図書館) / 31858 / 京都大学大学院医学研究科社会健康医学系専攻 / (主査)教授 川上 浩司, 教授 富樫 かおり, 教授 武藤 学 / 学位規則第4条第1項該当 / Doctor of Public Health / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20610号 / 医博第4259号 / 新制||医||1023(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 村井 俊哉, 教授 渡邉 大, 教授 林 康紀 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第20612号 / 医博第4261号 / 新制||医||1023(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 増永 慎一郎, 教授 溝脇 尚志, 教授 小泉 昭夫 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Single-photon avalanche diodes’ (SPADs) capabilities of detecting even a single photon with excellent timing resolution and compatibility with strong magnetic fields make them the most promising sensor for positron emission tomography imaging systems. With the advancements of silicon fabrication techniques, SPADs designed in standard planar complementary metal-oxide-semiconductor (CMOS) processes show competitive performance and a lower manufacturing cost. Additionally, CMOS SPADs have the potential for monolithic integration with other CMOS signal conditioning and processing circuits to achieve simple, low-cost, and high-performance imaging solutions. This work targets the design and optimization of SPAD sensors to improve their performance using low-cost standard CMOS technologies. Firstly, a detailed review on the SPADs in recent literature is presented. Then, the random telegraph signal (RTS) noise is investigated based on n+/p-well SPADs fabricated in a standard 130 nm CMOS process. Through the measurements and analysis, the RTS noise of a SPAD is found to correlate with its dark count rate and afterpulsing. Next, we design n+/p-well SPADs with field poly gates to improve the noise performance. Furthermore, a SPAD pixel, consisting of a p+/n-well SPAD and a compact and high-speed active quench and reset circuit is designed and fabricated in a standard TSMC 65 nm CMOS process. The post-layout simulations show that this pixel achieves a short 0.1 ns quenching time and a 3.35 ns minimum dead time. The measurement results show that the SPAD pixel has a dark count rate of 21 kHz, a peak photon detection probability of 23.8% at a 420 nm wavelength and a timing jitter of 139 ps using a 405 nm pulsed laser when the excess voltage is set to 0.5 V. Due to the short quenching time, almost no afterpulsing is observed even at a low operating temperature of -35 °C. Finally, a new differential quench and reset (QR) circuit consisting of two QR circuits on both the cathode and anode to quench and reset the SPAD through both terminals is proposed to reduce the reset time, to increase the count rate, to reduce the afterpulsing and to reject the common-mode noise. / Thesis / Doctor of Philosophy (PhD) / Positron emission tomography (PET) imaging is a powerful tool for diagnosis and assessment of cancers and tumors in the clinical field. Due to their capabilities of detecting even a single photon, excellent timing resolution, and their compatibility with magnetic fields to build PET/MRI (magnetic resonance imaging) multimodal imaging systems; single-photon avalanche diodes (SPADs) become the most promising sensor technology for PET imaging applications. SPADs fabricated in standard complementary metal-oxide-semiconductor (CMOS) technologies allow for a lower manufacturing cost and present the potential to integrate with other CMOS circuits to form a complete imaging system. In this thesis, random telegraph signal noise in SPADs is investigated first. Then, the poly gate is used in the design of an n+/p-well SPAD to improve the noise performance. In addition, a compact and high-speed SPAD pixel is designed and fabricated using an advanced standard CMOS process. Thanks to the fast quench and reset circuit, the SPAD pixel achieves a very short quenching time and a high-count rate. Finally, a differential quench and reset (QR) circuit consisting of two QR circuits on both the cathode and anode to quench and reset the SPAD through both terminals is proposed and studied.