Optimizing statistical parametric mapping analysis of 18F-FDG PET in children.

Area of Science:

• Neuroimaging
• Pediatric Neurology
• Medical Image Analysis

Background:

• Statistical Parametric Mapping (SPM) is an objective tool for analyzing 18F-fluoro-2-deoxy-d-glucose-positron-emission tomography (FDG-PET) images.
• Ethical constraints prevent obtaining control data from healthy children, limiting SPM application in pediatric populations.
• Using adult controls for pediatric FDG-PET analysis presents significant limitations.

Purpose of the Study:

• To develop and validate a pseudo-control group of children for FDG-PET studies in pediatric epilepsy.
• To optimize SPM analysis for pediatric FDG-PET by establishing an age-matched control dataset.

Main Methods:

• Analyzed FDG-PET images from 47 children with refractory focal epilepsy using visual and SPM analysis.
• Compared SPM performance using two control groups: healthy young adults and a pediatric pseudo-control group.
• The pediatric pseudo-control group comprised patients with epilepsy and normal-appearing PET scans (visual and SPM).

Main Results:

• Visual analysis detected hypometabolic areas in 87% of pediatric epilepsy cases.
• The pediatric pseudo-control group significantly improved SPM specificity (97% vs. 89%) and positive predictive value (86% vs. 65%).
• This approach reduced hypometabolic cortical artifacts by 41%, particularly in younger patients.

Conclusions:

• An age-matched pseudo-control group optimizes SPM analysis of FDG-PET in pediatric epilepsy.
• This methodology offers a viable solution for addressing control group limitations in pediatric neuroimaging.
• The approach may be applicable to other pediatric brain pathologies requiring FDG-PET analysis.