Utility of MR fingerprinting in differentiating epileptogenic from non-epileptogenic cortical malformations

This study investigates the potential of magnetic resonance fingerprinting (MRF) as a non-invasive method to differentiate epileptogenic from non-epileptogenic cortical malformations.

Sixty-nine subjects were included: four patients with complex cortical malformations who underwent detailed pre-surgical assessments including Stereo-EEG (SEEG) and/or subsequent surgery, 17 with histopathologically confirmed FCD II, and 48 healthy controls (HC). All subjects underwent a whole-brain 3 T MRF acquisition, the reconstruction of which generated T1 and T2 relaxometry maps. A 3D ROI was created for each lesion, and ROI-based z-score normalization using HC data was performed to minimize bias from lesion location. Within-patient comparisons were performed in Cases 1-3 (each with multiple lesions); across-patient comparisons were performed in Cases 1, 2, and 4 (with radiologically diagnosed FCD II), using a reference library of histopathologically confirmed, epileptogenic FCD II lesions. The most effective MRF measures for distinguishing epileptogenic from non-epileptogenic cortical malformations were identified based on their concordance with SEEG findings, surgical location and seizure outcomes.

In all four cases, epileptogenic malformations showed significantly increased gray-matter (GM) T1 values compared to non-epileptogenic ones, both within and across patients. In Case 4, an MRI-negative epileptogenic region exhibited elevated GM and white-matter (WM) T1 and T2 values than normal cortex, which was undetectable on conventional MRI, highlighting the sensitivity of these measures.

Quantitative MRF metrics, especially GM T1, hold promise as a non-invasive tool to to differentiate epileptogenic from non-epileptogenic cortical malformations. Our approach could aid SEEG implantation and surgical planning when complex/multiple cortical malformations are present.