Anatomy-constrained modelling of image-derived input functions in dynamic PET using multi-organ segmentation

TL;DR

This work addresses the accuracy of kinetic analysis in dynamic FDG PET by moving beyond a single aortic input function. It introduces an organ-aware framework that leverages CT-based multi-organ segmentation to extract four IDIFs (aorta, portal vein, pulmonary artery, ureters) and integrates them into a multi-input two-compartment model with partial blood volume correction. Across 9 patients, the approach yields substantial MSE improvements for the liver and meaningful gains for the lungs, highlighting the potential to integrate tracer kinetic modelling more fully into clinical practice. Limitations include surrogate ureter segmentation and the small sample size, suggesting directions for direct ureter segmentation and validation in larger cohorts.

Abstract

Accurate kinetic analysis of [$^{18}$F]FDG distribution in dynamic positron emission tomography (PET) requires anatomically constrained modelling of image-derived input functions (IDIFs). Traditionally, IDIFs are obtained from the aorta, neglecting anatomical variations and complex vascular contributions. This study proposes a multi-organ segmentation-based approach that integrates IDIFs from the aorta, portal vein, pulmonary artery, and ureters. Using high-resolution CT segmentations of the liver, lungs, kidneys, and bladder, we incorporate organ-specific blood supply sources to improve kinetic modelling. Our method was evaluated on dynamic [$^{18}$F]FDG PET data from nine patients, resulting in a mean squared error (MSE) reduction of $13.39\%$ for the liver and $10.42\%$ for the lungs. These initial results highlight the potential of multiple IDIFs in improving anatomical modelling and fully leveraging dynamic PET imaging. This approach could facilitate the integration of tracer kinetic modelling into clinical routine.

Figures (7)

• Figure 1: MSE values and relative error for liver and lung for all 9 patients.
• Figure 2: Predicted vs. measured TAC for liver for an exemplary patient.
• Figure 3: Predicted vs. measured TAC for lungs for an exemplary patient.
• Figure 4: Predicted vs. measured TAC for kidneys for an exemplary patient.
• Figure 5: Coronal view of calculated parametric images only using the aorta (single input) (first row, first four columns) vs. images using multiple IDIFs (multiple input) (second row, first four columns), along with calculated $\alpha$, $\beta$, $\gamma$, and $\delta$ values (third row, first four columns), the corresponding PET and CT slice (last column, top and bottom), and an MIP of PET (last column, center) as reference.