Loading...

Using Compartment Modeling for the Non-Invasive Diagnosis of Malignant Tumors, as Well as Determining the Active Volume and Total Glycolysis of Malignant Tumors in PET/CT Imaging with FDG Radiation

Reshtebar Sofla, Niloufar | 2024

0 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 57728 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Hosseini, Abolfazl; Sheikhzadeh, Peyman
  7. Abstract:
  8. In static PET/CT imaging, the Standardized Uptake Value (SUV) parameter is used to evaluate diseases. The use of this parameter provides valuable information to physicians in many cases. However, this parameter has limitations and is not suitable for certain clinical applications, such as assessing treatment response and differentiating between benign and malignant tumors. In contrast, utilizing dynamic imaging systems can yield significant information through the extraction of kinetic features and quantitative analysis based on temporal data. However, the use of dynamic PET imaging in a clinical setting can be challenging due to factors such as prolonged scan times (> 60 minutes), complexity of data acquisition protocols, and measurement of the input function. Therefore, investigating shortened data acquisition protocols is of great importance. In this thesis, we first examine the estimation of kinetic parameters and the production of parametric Ki images using a shortened data acquisition protocol (three routine static scans) employing the SA algorithm. For this purpose, six heterogeneous tumors with varying uptake levels (in lung and liver tissues) are simulated using the XCAT phantom. Ultimately, it was concluded that the shortened protocol using three static scans at optimized times of 20 minutes, 60 minutes, and 90 minutes post-radioisotope injection can accurately estimate kinetic parameters in PET imaging. Next, the feasibility of producing parametric images using a dual time-point (DTP) scan comprising two routine static scans was investigated, and the two-point data acquisition protocol was optimized. For this purpose, six heterogeneous tumors (in lung and liver tissues) were simulated using the XCAT phantom. It was ultimately demonstrated that utilizing a DTP data acquisition protocol with two routine static scans at times of 50-53 minutes and 90-93 minutes post-radioisotope injection can serve as a good substitute for complete dynamic imaging in modeling patlak. Subsequently, the tissue features (radiomics) in the parametric images obtained from DTP imaging were examined using three different segmentation methods. For this purpose, thirty-nine heterogeneous tumors were simulated in lung tissue using the XCAT phantom. It was observed that a dual time-point data acquisition protocol at 60 minutes and 90 minutes post-radioisotope injection could serve as a suitable alternative for patlak parametric images and act as a complement to SUV images
  9. Keywords:
  10. Multi-Segment Modeling ; Four-Dimensional Extended Cardiac-Torso (XCAT)Phantom ; Simulated Annealing Method ; Compton-Positron Emission Tomography (PET)Imaging ; Patlak Modeling ; Dynamic Imaging

 Digital Object List

 Bookmark

No TOC