Enhancing clinical utility of respiratory-gated PET/CT using patient respiratory trace classification

Abstract

This work characterizes associations between respiratory trace parameters and improvements in PET image quantification due to gating so as to guide the selection of patients for appropriate motion compensation. Twenty-six lung and liver cancer patients underwent PET/CT exams with recorded abdominal displacement using the Varian RPM system. Static and respiratory-gated [ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">18</sup> F]FDG PET images for each patient were generated from the same list-mode acquisition in 2D or 3D mode on a GE Discovery STE PET/CT scanner, attenuation corrected with helical or phase-averaged cine CT, and reconstructed with OSEM. Patients were grouped by trace type using heuristic classification, by lesion location on diagnostic CT, or by lesion translational freedom on diagnostic CT. FDG PET lesion avidity was quantified with the maximum standardized uptake value (SUV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> ). Relative differences (% ASUV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> ) between static and gated PET images were calculated across patient groups and compared using an independent 2-sample t-test. No statistically significant differences in % ASUV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> were observed between patients of different trace type (Type 1 vs. 2-3) or lesion translational freedom (detached vs. attached). Lower lung and liver lesions had significantly higher % ASUV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> than upper lung lesions (14% vs. 3%, p <; 0.0001). Linear associations were strongest between % ASUV <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">max</sub> and motion reduction & data fraction (0.42 ± 0.10, r <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.59, p <; 0.003) metrics. Patients could be classified according to lesion location between those with significant changes in PET parameters due to respiratory gating and those with negligible changes. Further refinement in patient selection guidelines may enhance respiratory-gated PET/CT utility in therapy target definition and response assessment.