Self-Supervised Learning for Physiologically-Based Pharmacokinetic Modeling in Dynamic PET
This work addresses the need for faster and spatially-aware diagnostic modeling in clinical PET imaging, though it is incremental as it builds on existing UNet and self-supervised techniques.
The paper tackles the slow speed and spatial ignorance of conventional fitting for voxel-wise pharmacokinetic modeling in dynamic PET by training a spatio-temporal UNet with a self-supervised loss, achieving quantitatively comparable results at organ-level while generating pixel-wise parametric images consistent with physiology.
Dynamic positron emission tomography imaging (dPET) provides temporally resolved images of a tracer enabling a quantitative measure of physiological processes. Voxel-wise physiologically-based pharmacokinetic (PBPK) modeling of the time activity curves (TAC) can provide relevant diagnostic information for clinical workflow. Conventional fitting strategies for TACs are slow and ignore the spatial relation between neighboring voxels. We train a spatio-temporal UNet to estimate the kinetic parameters given TAC from F-18-fluorodeoxyglucose (FDG) dPET. This work introduces a self-supervised loss formulation to enforce the similarity between the measured TAC and those generated with the learned kinetic parameters. Our method provides quantitatively comparable results at organ-level to the significantly slower conventional approaches, while generating pixel-wise parametric images which are consistent with expected physiology. To the best of our knowledge, this is the first self-supervised network that allows voxel-wise computation of kinetic parameters consistent with a non-linear kinetic model. The code will become publicly available upon acceptance.