Computational patient models for simulation of dynamic gamma-camera imaging : Application to renography and peptide receptor radionuclide therapy
Sammanfattning: Many nuclear medicine investigations rely on gamma-camera imaging to study and quantify the distribution of radiopharmaceuticals or radionuclides in the patient as a function of time. This is typically used for diagnostic studies of physiological functions or for calculation of absorbed doses following radionuclide therapy. In this work, computational patient models (phantoms) have been developed and used for evaluation of quantitative methods and techniques relying on dynamic gamma-camera imaging.Papers I and II concern 99mTc-MAG3 dynamic renography, a well-established diagnostic modality for evaluation of renal function. In paper I, a patient model featuring the pharmacokinetics of 99mTc-MAG3 was presented. The developed framework readily allows modelling of various cases of clinical interest in a systematic manner. Dynamic image acquisition was simulated using the Monte Carlo method, and the resulting image data were encapsulated in the DICOM format to allow processing with software used in clinical practice. In paper II, this data were used to investigate the accuracy and inter-departmental variability in dynamic renography analysis, with participation from 21 nuclear medicine departments in Sweden. We found that the variability in estimates of renal TAC parameters is low and acceptable when renal function is normal, but considerably high when renal function is impaired. The accuracy of relative uptake measurements was negatively affected by the lack of attenuation correction for quantitation. Papers III-IV concern image-based, patient-specific dosimetry in peptide receptor radionuclide therapy (PRRT) with 177Lu-DOTATATE. Paper III describes the development of computational patient models for research on image-based dosimetry, based on the same approach as used in paper I. A preliminary evaluation of a realistic dosimetry protocol, based on a single SPECT and four planar scans, was performed and it was shown that absorbed doses to organs and tumours were accurate within ±25 %. In paper IV, the patient models were used in a thorough analysis of uncertainty in renal dosimetry based entirely on SPECT/CT, and a total uncertainty of approximately 6 % (1 standard deviation) was estimated in the absorbed dose to the kidneys. In paper V, the dosimetric impact of the long-lived meta-stable isomer 177mLu was studied. Furthermore, it was investigated if current dosimetry protocols, relying on measurements limited to the first week after treatment, are sufficient to predict the long-term activity retention. The results showed a negligible contribution from 177mLu to the whole-body absorbed dose, and that measurements performed more than one week after treatment are warranted for tumour and whole-body dosimetry. In conclusion, this thesis provides a contribution to the knowledge of measurement accuracy and uncertainty in dynamic renography and 177Lu PRRT.
KLICKA HÄR FÖR ATT SE AVHANDLINGEN I FULLTEXT. (PDF-format)