Over the past decade, radiopharmaceutical therapies have demonstrated considerable potential in cancer treatment. Notably, the success of the NETTER-1 and VISION clinical trials led to FDA approval of Lu-177, a beta-emitting isotope, for treating neuroendocrine tumors in 2018 and prostate cancer in 2022. Coinciding with these advancements, there has been growing interest in exploring treatment outcomes using alternative isotopes like the alpha-emitter Ac-225, which may offer enhanced therapeutic benefits. Many therapeutic isotopes also emit photons that, while not directly contributing to therapy, can be detected using SPECT imaging. This enables concurrent delivery and evaluation of patient absorbed dose: a practice that is well-established in the field of external beam radiotherapy. Although current radiopharmaceutical treatment protocols use a standard "one-size-fits-all" approach whereby all patients receive the same injected activity, it is conjectured that image-based dosimetry can be used to tailor dosimetry on an individual basis and consequently improve treatment outcome. One of the major challenges of dosimetry is minimizing and accounting for the presence of bias and uncertainty in acquired SPECT images.
This thesis contains a collection of studies aimed at improving SPECT image quality and interpretability via improvements and modifications to existing image reconstruction protocols. Chapter 2 of the work describes the development of the open-source medical imaging software PyTomography, which enabled the subsequent innovations of this work. Chapter 3 derives a collimator detector response model for SPECT reconstruction of high energy photons, such as those emitted by the daughters of Ac-225. Chapter 4 outlines a modification to existing reconstruction algorithms to permit uncertainty estimation in medical images and subsequently in image-based dosimetry. Chapter 5 explores the optimal image acquisition and reconstruction parameters for Ac-225 imaging, and Chapter 6 explores Monte Carlo based reconstruction techniques to further improve image quality.
Add to Calendar
2025-05-27T08:00:002025-05-27T10:00:00Advancing Quantitative Dosimetry SPECT with Open-Source Image Reconstruction, Uncertainty Estimation, and Image Generation OptimizationEvent Information:
Abstract:
Over the past decade, radiopharmaceutical therapies have demonstrated considerable potential in cancer treatment. Notably, the success of the NETTER-1 and VISION clinical trials led to FDA approval of Lu-177, a beta-emitting isotope, for treating neuroendocrine tumors in 2018 and prostate cancer in 2022. Coinciding with these advancements, there has been growing interest in exploring treatment outcomes using alternative isotopes like the alpha-emitter Ac-225, which may offer enhanced therapeutic benefits. Many therapeutic isotopes also emit photons that, while not directly contributing to therapy, can be detected using SPECT imaging. This enables concurrent delivery and evaluation of patient absorbed dose: a practice that is well-established in the field of external beam radiotherapy. Although current radiopharmaceutical treatment protocols use a standard "one-size-fits-all" approach whereby all patients receive the same injected activity, it is conjectured that image-based dosimetry can be used to tailor dosimetry on an individual basis and consequently improve treatment outcome. One of the major challenges of dosimetry is minimizing and accounting for the presence of bias and uncertainty in acquired SPECT images.
This thesis contains a collection of studies aimed at improving SPECT image quality and interpretability via improvements and modifications to existing image reconstruction protocols. Chapter 2 of the work describes the development of the open-source medical imaging software PyTomography, which enabled the subsequent innovations of this work. Chapter 3 derives a collimator detector response model for SPECT reconstruction of high energy photons, such as those emitted by the daughters of Ac-225. Chapter 4 outlines a modification to existing reconstruction algorithms to permit uncertainty estimation in medical images and subsequently in image-based dosimetry. Chapter 5 explores the optimal image acquisition and reconstruction parameters for Ac-225 imaging, and Chapter 6 explores Monte Carlo based reconstruction techniques to further improve image quality. Event Location:
BC Cancer Research Agency (675 W 10th Ave, Vancouver, BC V5Z 0B4), Boardroom first floor