Low activity [¹¹C]raclopride kinetic modeling in the mouse brain using the spatiotemporal kernel method

Area of Science:

• Nuclear Medicine
• Neuroscience
• Medical Imaging

Background:

• Positron Emission Tomography (PET) studies in mice often require low tracer activity to prevent receptor saturation, leading to image noise.
• Accurate quantification of tracer kinetics and binding potential is crucial for understanding brain function and disease in mouse models.

Purpose of the Study:

• To evaluate and compare various dynamic PET reconstruction methods for quantifying binding potential (BP_ND) and R1 in mouse brain studies with reduced injected activity.
• To identify reconstruction strategies that maintain accuracy and reduce noise under low-count conditions.

Main Methods:

• Investigated independent frame reconstruction (IFR), post-reconstruction HYPR denoising, direct reconstruction with simplified reference tissue model (DIR-SRTM), spatial kernel reconstruction (KERS), and spatiotemporal kernel reconstruction (KERST).
• Evaluated HYPR denoising applied to kernel methods (KERS-HYPR, KERST-HYPR).
• Utilized *in vivo* mouse data, simulating reduced injected activity levels (2.07 to 0.260 MBq) and analyzing regional and voxel-wise quantification.

Main Results:

• All methods showed high regional correlation (r > 0.94) between reduced and full counts, with KERS-HYPR and KERST-HYPR achieving the highest (r > 0.96).
• KERST and KERST-HYPR demonstrated superior performance, maintaining Bland-Altman bias and SD below 5% for noise levels up to a 16-fold reduction.
• Spatiotemporal kernel methods provided accurate binding potential estimates even at injected activities as low as 0.26-0.518 MBq.

Conclusions:

• Spatiotemporal kernel reconstruction (KERST), particularly with HYPR denoising (KERST-HYPR), is highly effective for accurate PET quantification in mouse studies with very low injected tracer activities.
• These advanced reconstruction techniques are essential for preserving the integrity of tracer kinetic principles in preclinical neuroimaging research.