FDG-PET parametric imaging by total variation minimization.

Area of Science:

• Neuroimaging
• Medical Physics
• Radiochemistry

Background:

• Positron emission tomography (PET) is crucial for assessing cerebral metabolic rate for glucose (CMRGlc).
• Traditional methods for parametric imaging of CMRGlc using [(18)F]-fluoro deoxyglucose (FDG) suffer from low signal-to-noise ratios at the voxel level, limiting image quality.
• This noise significantly hinders accurate regional analysis and interpretation of brain glucose metabolism.

Purpose of the Study:

• To develop an advanced method for parametric imaging of CMRGlc using FDG-PET.
• To address the limitations of low signal-to-noise ratios in traditional voxel-wise imaging.
• To improve the spatial homogeneity within brain regions and the distinction between different brain regions in the resulting images.

Main Methods:

• The proposed method minimizes the total variation of tracer uptake rates.
• It simultaneously ensures a good fit to traditional Patlak kinetic modeling equations.
• Brain phantom simulations were employed to validate the method's performance.

Main Results:

• The proposed method demonstrated significant improvements in image quality compared to standard Patlak imaging.
• Images generated by the new technique showed enhanced spatial homogeneity within brain regions.
• The distinction between different brain regions was notably improved.
• Performance was superior to Patlak images subjected to post-filtering with Gaussian or median filters.

Conclusions:

• Minimizing total variation while fitting Patlak equations offers a robust approach for parametric CMRGlc imaging.
• This technique effectively overcomes the signal-to-noise limitations of traditional methods.
• The proposed method yields higher quality and more interpretable FDG-PET images for brain glucose metabolism studies.