Implementation of new algorithms for an accurate gamma-ray impact determination in scintillation monolithic blocks for pet applications

  1. FREIRE LÓPEZ-FANDO, MARTA
Dirigida por:
  1. Andrea González Montoro Director/a
  2. Antonio Javier González Martínez Director/a

Universidad de defensa: Universitat Politècnica de València

Fecha de defensa: 14 de julio de 2023

Tribunal:
  1. Peter Bruyndonckx Presidente/a
  2. Joseba Alonso Otamendi Secretario/a
  3. Marcin Balcerzyk Vocal

Tipo: Tesis

Resumen

Positron Emission Tomography (PET) is a powerful imaging technique that provides quantitative measurements of biological and physiological processes occurring within the body at the molecular level by using specific radiopharmaceuticals. PET imaging returns functional information that allows for early diagnosis and personalized therapy treatment follow up. It has applications in several research and clinical areas, such as oncology, neurology or cardiology, among others. Efforts to improve PET systems performance are focused on increasing their sensitivity and image quality, allowing for more accurate clinical assessments. In PET imaging, a radiotracer labeled with a positron-emitting radionuclide is injected to the patient and consequently, distributed throughout the body. During the radiotracer decay, the isotope emits a positron that annihilates with an electron of the surrounding tissues, generating two 511 keV gamma-rays emitted at approximately 180º. The PET technique is based therefore on the simultaneous detection of these two gamma-rays, called annihilation photons, by usually employing a ring of detectors around the patient. Improving the design and performance of these detectors, increases the diagnostic capabilities of PET imaging. To boost PET performance, it has been suggested to use detectors based on monolithic crystals designs, due to their advantages compared to pixelated detectors. However, their implementation in commercial scanners requires overcoming some challenges mostly related to photon impact positioning methods and calibration procedures to provide the impact coordinates and time of arrival of the annihilation photons. This PhD thesis focuses on the development and experimental validation of methodologies for an accurate determination of this information in monolithic detectors, emphasizing in their practical application to full PET systems. During this thesis, the main principles of monolithic-based PET detectors have been studied to understand their behavior and limitations. Typical monolithic detector configurations based on continuous scintillation blocks coupled to flat SiPM arrays have been first considered; additionally, other novel approaches have been also validated. Two main methodologies for 3D photon interaction positioning, one based on analytical methods and another based on Deep Learning algorithms, have been developed to increase the overall detector performance. The proposed methods have been validated at the detector level but also in different PET scanners developed by our group. The present thesis is based on a compendium of the most relevant papers published in peer-reviewed journals by the PhD candidate and is organized as follows. Chapter I presents an introduction to the thesis work, composed by three sections: Medical Imaging, principles of Positron Emission Tomography and, Position estimation and calibration in monolithic-based detectors. Chapter II contains the specific objectives of this thesis and the main contributions of the candidate to the field. This chapter also includes some recent methodologies and results that have not yet been published. Chapter III collects an author copy of the four published articles selected for the compendium, in which the candidate is the first author [1]-[4]. In Chapter IV the main results and conclusion achieved during the thesis are discussed. Finally, Chapter V presents the discussion of this thesis, summarizing the main contributions and highlighting the scientific achievements.