Abstract
18F-FDG PET is an important tool for the presurgical assessment of children with drug-resistant epilepsy. Standard assessment is performed visually and is often subjective and highly user-dependent. Voxelwise statistics can be used to remove user-dependent biases by automatically identifying areas of significant hypo- or hypermetabolism associated with the epileptogenic area. In the clinical setting, this analysis is performed using commercially available software. These software packages suffer from two main limitations when applied to pediatric PET data: pediatric scans are spatially normalized to an adult standard template, and statistical comparisons use an adult control dataset. The aim of this work was to provide a reliable observer-independent pipeline for the analysis of pediatric 18F-FDG PET scans, as part of presurgical planning in epilepsy. Methods: A pseudocontrol dataset (19 subjects 6–9 y old, and 93 subjects 10–20 y old) was used to create two age-specific 18F-FDG PET pediatric templates in standard pediatric space. The 18F-FDG PET scans of 46 epilepsy patients (16 patients 6–9 y old, and 30 patients 10–17 y old) were retrospectively collated and analyzed using voxelwise statistics. This procedure was implemented with the standard pipeline available in the commercial software Scenium and an in-house Statistical Parametric Mapping, version 8 (SPM8), pipeline (including age-specific pediatric templates and reference database). A κ-test was used to assess the level of agreement between the findings of voxelwise analyses and the clinical diagnosis of each patient. The SPM8 pipeline was further validated using postsurgical seizure-free patients. Results: Improved agreement with the clinical diagnosis was reported using SPM8, in terms of focus localization, especially for the younger patient group: κ = 0.489 for Scenium versus 0.826 for SPM. The proposed pipeline also showed a sensitivity of about 70% in both age ranges for the localization of hypometabolic areas on pediatric 18F-FDG PET scans in postsurgical seizure-free patients. Conclusion: We showed that by creating age-specific templates and using pediatric control databases, our pipeline provides an accurate and sensitive semiquantitative method for assessing the 18F-FDG PET scans of patients under 18 y old.
Highlights
Multimodal imaging in patients with drug-resistant epilepsy is used to guide presurgical assessment in potential candidates for surgery [1]
We investigated whether more accurate and reliable results can be achieved using age-specific templates and databases of 18F-FDG PET images, and we introduced a processing pipeline to improve the accuracy with which the epileptogenic zone is localized on pediatric 18F-FDG PET scans
The adopted methodology included the voxelwise analysis of a pediatric patient dataset using a commercial software and an ad-hoc pipeline based on the Statistical Parametric Mapping, version 8 (SPM8, http://www.fil.ion.ucl.ac.uk/spm/)
Summary
Multimodal imaging in patients with drug-resistant epilepsy is used to guide presurgical assessment in potential candidates for surgery [1]. More sophisticated voxelwise statistical methods can be used to remove user-dependent biases, thereby improving reproducibility and preventing disagreement arising from visual assessment of PET scans [2] These methods spatially normalize the patient brain scan to a common standard template, where it is compared with a normative dataset. Our aim was to create 18F-FDG PET templates representing two age ranges (6–9 y and 10–20 y) and use age-appropriate control datasets for voxelwise analysis This pipeline was quantitatively compared against the commercial software Scenium, currently adopted in clinical settings, and was further validated in postsurgical seizure-free patients, whose clinical followup diagnosis was used as the gold standard
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
