Development and assessment of estimate methods for internal dosimetry using pet/ct
- Autores: Natàlia López Vilanova
- Directores de la Tesis: M. Amor Duch Guillen (dir. tes.), S. Bullich (codir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2019
- Idioma: español
- Tribunal Calificador de la Tesis: Mercè Ginjaume Egido (presid.), Albert Cot Sanz (secret.), Antonio Luis Gutiérrez Cardo (voc.)
- Programa de doctorado: Programa de Doctorado en Ingeniería Nuclear y de las Radiaciones Ionizantes por la Universidad Politécnica de Catalunya
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
The aim of this thesis was to assess and develop internal dose calculations methods in diagnostic and therapeutic nuclear medicine procedures to patients undergone PET/CT explorations. Towards this objective, the accuracy and precision of different classical methods commonly used to estimate internal dosimetry were investigated. Biodistribution studies were used in order to compare these methods. The main study aspects included region-of-interest (ROI) delineation methods, reconstruction algorithms, scatter correction and radiopharmaceutical's biokinetic. Optimization of internal dosimetry in this thesis was completed with the development of a Monte Carlo (MC) technique for estimating the patient-specific PET/CT dosimetry.
The development of a mathematical model using MC techniques allowed us to have a gold standard to which compare classical techniques and study the aspects discussed previously. It was observed that effective dose (ED) estimations were sensitive to whichever delineation ROI method was applied. Furthermore, it was perceived that the biokinetics of the radioligand also influences in the ED estimation. On the other hand, similar quantitative accuracy was found regarding image reconstruction (FBP and OSEM) and scatter correction methods studied (FSC and SSC). Analysis of the impact of inter- and intra-operator variability in dose estimations revealed higher reproducibility in 3D methods in comparison with 2D planar method. The last one, showed the highest interoperator variability, which implies an overestimation of the ED.
In this dissertation, specific routines were developed to be applied with the MC code PENELOPE/penEasy to perform individualized internal dosimetry estimations. Voxel-level absorbed dose maps which include self- and cross-irradiation doses were generated from the morphological and functional patient images. Further parameters such as cumulative organ dose, maximum and minimum voxel organ values, volume of the organ and dose-volume histograms of interest were reported. The model implemented was applied to a theoretical study using simulated PET images of a voxelized Zubal phantom. The results were benchmarked with the ones obtained using the OLINDA/EXM software. The comparison was in good agreement for those organs were both phantoms considered (Zubal and the reference one in OLINDA/EXM) were close.
Undoubtedly, the implementation of a patient-specific internal dosimetry method not only leads to an improvement in diagnostic examinations where the risk could be quantified, but also NM therapy could become more effective in terms that patients receiving an optimal care.
