Quantitation in positron emission computed tomography: 5. Physical--anatomical effects

Area of Science:

• Medical Imaging
• Neuroscience
• Radiochemistry

Background:

• Accurate quantification in Positron Emission Tomography (PET) is crucial for clinical and research applications.
• Partial volume effects (PVE) are known to introduce errors in PET quantitation, particularly in neuroimaging.
• The interplay between anatomical structure characteristics and imaging resolution requires detailed investigation.

Purpose of the Study:

• To investigate the impact of neuroanatomical structure size, shape, and position on quantitation accuracy in PET.
• To evaluate the influence of spatial tomographic resolution relative to structure volume on quantitative measurements.
• To understand how partial volume effects influence the detection of regional cerebral asymmetries.

Main Methods:

• Utilized computer simulations to analyze partial volume effects in the plane of section.
• Examined the relationship between voxel size and neuroanatomical structure volume.
• Simulated regional cerebral asymmetries to assess the impact of PVE on their measurement.

Main Results:

• Recovery coefficients decreased to ≤42% when voxel size exceeded structure volume.
• Largest errors occurred in small, thin, irregularly shaped structures with differing adjacent pixel values.
• Partial volume effects significantly enhanced or obliterated regional cerebral asymmetries based on structure size and shape.
• Large, circular, and separated asymmetries were more detectable than small, thin, adjacent ones.

Conclusions:

• Neuroanatomical structure characteristics (size, shape) are critical determinants of PET quantitation accuracy, alongside spatial resolution.
• Partial volume effects can substantially distort or obscure true regional differences, impacting the interpretation of brain asymmetries.
• Recommendations for error reduction and correction factors are essential for reliable PET quantitation in neuroimaging.