Innovative Kinetic Modelling Strategies for Radiometabolite Analysis using Total-Body PET and Simultaneous PET-MR Imaging

经典的PET定量方法使用动力学建模需要侵入性血液采样，如前所述。放射性代谢物实验需要6-15只动物的血液和组织样本，并且每开发一种新的示踪剂都需要重复。最近已经开发出一种方法，可以从PET扫描中提取放射性示踪剂血液曲线，而不需要血液采样[112]，但是没有方法可以直接从PET数据中提取有关放射性代谢物的信息。本博士旨在开发一种非侵入性放射性代谢物分析的新方法，以允许直接从PET图像和时间活性曲线中精确量化PET数据。本博士将利用最近开发的技术，如全身pet和PET-MR来研究新技术。首先，我将利用现有的[18F]SynVesT-1（参见第7.1.1章）大脑数据，对不同的输入函数推导方法和模型进行全面的评估和比较。我将进行金标准分析，并承担产生图像衍生全血曲线的方法。然后，我将比较这些不同输入函数推导方法的动力学建模结果，并随后将其与仅使用全血曲线来评估代谢物校正中的偏差进行比较。这个比较将对一系列的方法进行，这将使我了解目前的方法和它们的缺陷。根据目标1，我将探索全身PET的潜力，以提供产生准确输入函数所需的放射性代谢物校正。从全身PET扫描中收集信息将有可能建立器官和代谢物之间的关系。这一目标的理想结果将是了解全身PET提供的信息，这些信息可以防止在PET研究期间需要动脉血液采样。在围绕PET数据的调查之后，我将调查MR技术，目的是提供关于体内循环的放射性代谢物数量的直接或间接信息。MR也允许使用光谱学鉴定注射后产生的特定放射性代谢物。这项研究可以提供感兴趣的参数，允许从PET-MR扫描直接建立放射性代谢物。利用之前关于当前输入函数推导方法的所有信息，研究了全身PET关系和潜在的MRI方法，建立了一个可以直接从图像中推导出自由放射性代谢物校正输入函数的模型。该图像是否只需要PET或PET- mri将取决于先前目标的成功。然后需要对该方法进行验证，并在另一种具有不同放射性代谢物特征的示踪剂上进行测试- [18F]LW223。

The classical method of PET quantification using kinetic modelling requires invasive bloodsampling as previously described. The radiometabolite experiments require blood and tis-sue samples from 6-15 animals and need repeating for each new tracer developed. A recentmethod has been developed to extract the radiotracer blood curve from PET scans whichdoes not require blood sampling [112], however there have been no methods developed toextract information about the radiometabolites directly from PET data. This PhD aims todevelop a new method for non-invasive radiometabolite analysis to allow for the accuratequantification of PET data directly from the PET image and time activity curves. ThisPhD will take advantage of more recently developed technology such as total body PETand PET-MR to investigate novel techniques. Firstly, I will undertake an thorough assessment and comparison of different input function derivation methods and models in place using existing [18F]SynVesT-1 (see Chapter 7.1.1) data for the brain. I will carry out gold-standard analysis as well as undertaking methods to produce image-derived whole blood curves. I will then compare the kinetic modelling outcomes from these different input function deriving methodologies and subsequently compare these to using whole blood curves only to assess the bias in the metabolite correction. This comparison will be carried out for a range of methods and will enable me to understand current methods and their pitfalls. Following aim 1, I will explore the potential for total body PET to provide the radiometabolite correction needed to produce an accurate input function. Gathering information from the total body PET scans will potentially allow for relationships between organs and metabolites to be established. The ideal outcome of this aim would be to understand information that total-body PET provides which can prevent the need for arterial blood sampling during PET studies. Following the investigation revolving around PET only data, I shall be investigating MR techniques with the aim of providing either direct or indirect information about the quantity of radiometabolites circulating the body. MR may also allow for the specific radiometabolites produced post injection to be identified using spectroscopy. This investigation could provide parameters of interest which allow for the radiometabolites to be established directly from a PET-MR scan. Using all the information gathered from the previous aims on current input function derivation methods, total body PET relations and potential MRI methods investigated to develop a model which can derive a free radiometabolite corrected input function directly from an image. Whether that image will be required to be PET only or PET-MRI will depend on the success on prior aims. This method will then need to validated and shall be tested on another tracer which has a different radiometabolite profile - [18F ]LW223.