Statistical modeling strategies for estimation of the arterial input function in dynamic positron emission tomography data analysis

Kinetic modelling of dynamic PET data requires knowledge of tracer concentration in blood plasma, described by the arterial input function (AIF). Arterial blood sampling is the gold standard for AIF measurement, but is invasive and labour intensive. A number of methods have been proposed to accurately estimate the AIF directly from blood sampling and/or imaging data. In this work, we review some of the main methodologies for AIF estimation, and present two alternatives that aim at addressing some of their major limitations. We developed a tracer travel rate projection model as an early AIF modelling attempt based on tracer travel history in the circulatory system. A penalty was considered for model parameter estimation and we used arterially sampled data for evaluation. To improve on this first model, we developed a population-based projection model that exploits historical data to estimate individual AIFs. We represent the history of a tracer atom at a sampling site by its travel time, modeled as a sum of the time for the atom to initially progress from the injection site to the right ventricle of the heart, and the time it spends in circulation before being sampled. The former is modeled as a realization from a gamma distribution, whose parameters are common to all subjects in the population, and estimated from a collection of arterial sampling data for the given tracer. The latter is represented by a subject-specific linear mixture of these population pro- files. This approach can be seen as a projection of individual AIF characteristics onto a basis of population profile components. It also incorporates knowledge of injection duration into the model fit, allowing for varying injection protocols. Analyses of arterial sampling data from 18F-FDG, 15O-H2O and 18F-FLT clinical studies show that the proposed model can outperform reference techniques. The statistically significant gain offered by using population data to train the basis components, as opposed to fitting these from the single individual sampling data, is measured on the FDG cohort. Kinetic analyses demonstrate the reliability and potential benefit of this approach in estimating physiological parameters. These results are further supported by numerical simulations that demonstrate convergence of the proposed technique with decreasing noise levels, and stable levels of performance under varying training population sizes and noise levels.

Grantor dc:publisher

University College Cork

Year dc:date.issued

2023

Author dc:creator

• Xiu, Zhaoyan

Advisors dc:contributor.advisor

• Huang, Jian
• Wolsztynski, Eric

• Arterial input function
• Positron emission tomography
• Population based projection model
• Travel rate projection model
• Dynamic PET
• AIF model
• Medical imaging
• AIF

Statement dc:rights

• © 2023, Zhaoyan Xiu.

Licence dc:rights.uri

• https://creativecommons.org/licenses/by-nc-nd/4.0/

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/10468/14535

OAI identifier oai:identifier

oai:cora.ucc.ie:10468/14535