Structural and Quantitative MRI Signal Features in Pediatric Focal Cortical Dysplasia: Clinical Relevance.
Authors
Affiliations (1)
Affiliations (1)
- Department of Medical Imaging, Second Faculty of Medicine, Charles University and Motol University Hospital, Prague, Czech Republic; Department of Pathophysiology, Second Faculty of Medicine, Charles University, Prague, Czech Republic. [email protected] or [email protected].
Abstract
The aim of this study was to quantitatively characterize MRI signal-intensity (SI) abnormalities in focal cortical dysplasia (FCD) and to compare expert visual lesion delineation with objective SI metrics across conventional structural MRI sequences and quantitative T1 mapping (qT1). Thirteen pediatric patients with histologically confirmed FCD type I or II underwent presurgical MRI including 3D T1-weighted, T2-weighted, FLAIR, and qT1 sequences. Two expert-defined lesion masks were generated: a conservative feature-focused mask (L1) and a broader contextual mask reflecting overall radiological judgement (L2). Signal-intensity values were extracted from lesions, perilesional zones, and contralateral homologous regions, and spatial correspondence with the resection cavity was evaluated using the Dice score. L2 delineations were substantially larger and showed markedly higher overlap with resected tissue than L1, suggesting that expert contextual assessment captures clinically relevant but quantitatively subtle abnormalities. In contrast, L1 showed stronger and more consistent SI abnormalities, particularly on T2-weighted and FLAIR images, indicating that expert visual judgement and quantitative SI metrics do not fully converge. Perilesional tissue closely resembled healthy brain, with only minimal SI gradients beyond the lesion border. Quantitative T1 mapping provided limited additional value over conventional T1-weighted imaging for the detection of white-matter abnormalities. These findings indicate that expert assessment and quantitative MRI capture complementary but non-identical aspects of FCD pathology, and suggest that future automated detection models should combine quantitative features with multiple layers of radiological expertise. Key words Focal cortical dysplasia " Signal intensity " Structural MRI " Pediatric " Quantitative T1.