Answers are grounded in this paper and cite their sources.
To achieve high resolution and quantifiable cardiac Positron Emission Tomography (PET) images, the significant challenge of independent and highly variable respiratory and cardiac motion must be overcome. These motions often exceed the thickness of the left ventricular (LV) wall, blurring images and hindering the accurate diagnosis and quantification of cardiac diseases. While Magnetic Resonance Imaging (MRI) currently excels at resolving transmural differences in the LV wall, a capability not yet fully achieved by PET, advancements in PET resolution are now approaching 3 mm. This suggests that transmural quantification is attainable if motion is effectively mitigated. Hybrid PET/MRI systems offer a synergistic solution by providing simultaneous, co-registered data, allowing MRI’s superior anatomical resolution to guide PET reconstruction. This integration allows for motion correction and enhanced spatial resolution without the severe signal to noise ratio reduction associated with traditional gating methods. Strategies for defeating motion include using MRI derived motion vectors from Cine MRI and applying them to PET data, which significantly improves image quality. Furthermore, the integration of anatomical priors from MRI into PET reconstruction offers further resolution enhancement, particularly for the thin heart walls. The transformative potential of machine learning can further accelerate these corrections, paving the way for advanced applications like multi-tracer injection studies. Overcoming cardiac motion in PET/MRI promises to profoundly impact the diagnosis and management of heart disease by providing the quantifiable, high resolution cardiac PET images necessary for preclinical research and precise clinical theranostic assessment.
No takes yet. Share an insight, caveat, or question.
Lim et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: