Challenges and opportunities in quantitative brain PET imaging

One of the main advantages of Positron Emission Tomography (PET) imaging is the potential to obtain quantitative information about physiological processes. Unfortunately, the mathematical models that allow for quantitation generally require long image acquisition protocols and potentially invasive procedures, which poses challenges in terms of patient comfort and healthcare resource utilization. As a consequence, the quantitative character of PET does not find space in the clinical practice, and many preclinical studies (small animal imaging) also do not explore quantification methods to the fullest. In this context, this thesis explored a number of opportunities to 1) bridge the gap between quantitative potential and the necessity of simple methods for clinical practice and to 2) optimize quantitative methodologies for the preclinical setting.
First, two novel simplified methods were developed and validated. The methods applied mathematical approximations to established models and were successful in significantly reducing PET scan time without compromising quantitative accuracy. As a consequence, full quantitative information might be obtained from simple, short, patient-friendly scans. Next, the thesis focused on optimizing quantitative methodology for PET imaging of the rat brain. There, a well-established radiotracer in human studies was assessed for its specific use in the preclinical setting, and optimal quantitative methods were determined. The findings of this work showed it is not possible to directly translate methodology from human to rat studies, and highlighted the importance of model validation, also in the case of radiotracers which have been established in a different species.