Quantitative [ 18 F]-fluorodeoxyglucose PET analysis with a Gaussian normative template

Area of Science:

• Neurology
• Nuclear Medicine
• Medical Imaging

Background:

• Epileptogenic focus localization in drug-resistant epilepsy is challenging.
• Standardized quantitative analysis of [18F]-fluorodeoxyglucose (FDG) Positron Emission Tomography (PET) is needed.
• Current methods lack objective criteria for focus identification.

Purpose of the Study:

• To develop a standardized quantitative [18F]-FDG PET normative template.
• To evaluate its clinical utility for objective epileptogenic focus localization in drug-resistant epilepsy.

Main Methods:

• Harmonized [18F]-FDG PET data from 89 participants (healthy subjects and epilepsy patients).
• Developed and validated four candidate models (Gaussian, Gamma, Smoothed Gaussian, Smoothed Gamma).
• Selected the optimal template based on normality validation and assessed clinical utility using z-score threshold.

Main Results:

• The Smoothed Gaussian model was established as the final template with the highest voxel pass rate.
• Validation in independent healthy subjects showed no false positives, confirming high specificity.
• The template accurately identified metabolic abnormalities in all 24 temporal lobe epilepsy patients.

Conclusions:

• The developed Smoothed Gaussian-derived template enables objective and reproducible [18F]-FDG PET analysis.
• It significantly improves focus localization accuracy in drug-resistant epilepsy.
• Provides a reliable quantitative tool for guiding epilepsy surgery.