Regional compensation for statistical maximum likelihood reconstruction error of PET image pixels

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Biophysics

Background:

• Positron emission tomography (PET) imaging is crucial for assessing regional tracer concentration.
• Tissue heterogeneity, reflecting physiological variations, is of increasing interest in PET studies.
• Reconstructed PET images suffer from significant noise and model errors, obscuring true physiological variations.

Purpose of the Study:

• To develop a novel procedure for quantifying reconstruction errors in PET imaging.
• To differentiate true tissue heterogeneity from noise-induced variations in reconstructed PET data.
• To enhance the analysis of physiological variations in PET studies.

Main Methods:

• Generating and reconstructing noise realizations of virtual sinograms statistically similar to measured PET data.
• Quantifying error variation in regional reconstructed PET values.
• Utilizing physical phantom data for validation.

Main Results:

• The proposed procedure successfully quantifies error variation in regional PET values.
• The method allows for the characterization of true tissue heterogeneity despite reconstruction errors.
• Validation with phantom data demonstrates the feasibility of distinguishing error from heterogeneity.

Conclusions:

• The developed procedure effectively quantifies reconstruction errors in PET.
• This method enables the reliable assessment of tissue heterogeneity in PET imaging.
• Improved quantification of heterogeneity can lead to more accurate physiological assessments in PET.