Reduced acquisition time PET pharmacokinetic modelling using simultaneous ASL–MRI: proof of concept

Abstract

Pharmacokinetic modelling on dynamic positron emission tomography (PET) data is a quantitative technique. However, the long acquisition time is prohibitive for routine clinical use. Instead, the semi-quantitative standardised uptake value ratio (SUVR) from a shorter static acquisition is used, despite its sensitivity to blood flow confounding longitudinal analysis. A method has been proposed to reduce the dynamic acquisition time for quantification by incorporating cerebral blood flow (CBF) information from arterial spin labelling (ASL) magnetic resonance imaging (MRI) into the pharmacokinetic modelling. In this work, we optimise and validate this framework for a study of ageing and preclinical Alzheimer's disease. This methodology adapts the simplified reference tissue model (SRTM) for a reduced acquisition time (RT-SRTM) and is applied to [F]-florbetapir PET data for amyloid-β quantification. Evaluation shows that the optimised RT-SRTM can achieve amyloid burden estimation from a 30-min PET/MR acquisition which is comparable with the gold standard SRTM applied to 60 min of PET data. Conversely, SUVR showed a significantly higher error and bias, and a statistically significant correlation with tracer delivery due to the influence of blood flow. The optimised RT-SRTM produced amyloid burden estimates which were uncorrelated with tracer delivery indicating its suitability for longitudinal studies.

Comparison of simultaneous arterial spin labeling MRI and 15O-H2O PET measurements of regional cerebral blood flow in rest and altered perfusion states

“…This strategy would retain an accurate measure of the cortical binding potential free of the influence of blood flow. 49 However, our finding of a significant systematic reduction of both cerebellar rCBF and relative reactivity in ASL (Table 3 and Figure 4(b)) discourages its use as a reference region and for normalizing ASL MRI rCBF measurements in different perfusion states. 50 Even though 15 O-H 2 O PET and ASL employ the same tracer and thereby are very similar with regard to tracer distribution in tissue and diffusion through the vessel wall (Crone-Renkin effect) 51 and the spatial resolutions were matched, there are still subtle differences between the two techniques which potentially can explain differences when comparing results from the two methods.…”

“…This strategy would retain an accurate measure of the cortical binding potential free of the influence of blood flow. 49 However, our finding of a significant systematic reduction of both cerebellar rCBF and relative reactivity in ASL ( Table 3 and Figure 4(b) ) discourages its use as a reference region and for normalizing ASL MRI rCBF measurements in different perfusion states. 50 …”

A Survey on Explainable Artificial Intelligence (XAI): Toward Medical XAI

“…GDM parameters are said to be interpretable, since they are linear combinations of the clinical variables. DL has been used for PET pharmacokinetic (PK) modeling to quantify tracer target density [132]. CNN has helped PK modeling as a part of a sequence of processes to reduce PET acquisition time, and the output is interpreted with respect to the golden standard PK model, which is the linearized version of simplified reference tissue model (SRTM).…”

“…Giving an interpretation within limited constraints may place undue emphasis on the constraint itself. Other works that use predefined models might suffer similar problems [100], [106], [132].…”

Magnetic Resonance-Based Attenuation Correction and Scatter Correction in Neurological Positron Emission Tomography/Magnetic Resonance Imaging—Current Status With Emerging Applications

“…The need to measure IDIF could also be circumvented entirely by the measurement of global CBF by phase-contrast MRI [194]. Another approach is to incorporate information from arterial spin labeling (ASL) into PET pharmacokinetic modeling [195]. Thus, methods using complementary information from MR with PET for defining IDIF might be beneficial for kinetic modeling in PET, as MRdriven or MR-assisted approaches might be less dependent on accuracy of data corrections.…”