Clinical and high-field MRI in MRI negative epilepsy

The assessment of MRI as negative in epilepsy patients depends on the MRI protocol used, the magnetic field strength of the MRI, the expertise of the examiner, the background information on the presumed location of the epileptogenic zone and if necessary, the automated image analysis methods available at the center. Indeed, in the majority of surgically treated MRI negative patients an epileptogenic lesion can be demonstrated histopathologically, which then usually corresponds to subtle focal cortical dysplasia (FCD), hippocampus sclerosis (HS) or gliosis; however, preoperative visualization of the lesion is crucial for adequate resection of the epileptogenic zone and thus for a favorable postoperative outcome. This is why optimized MRI methods for epilepsy detection should be exploited in patients with focal epilepsy refractory to medication. These include the HARNESS MRI protocol with volumetric T1 and FLAIR sequences as well as a high-resolution T2 sequence at 3 T, which already enables a significantly higher lesion detection visually than with conventional sequences. The human eye should be supported by automated postprocessing methods, which have now achieved a high sensitivity in FCD and HS detection through continuous technical developments and the application of machine learning, but which require interpretation by experienced users due to the limited specificity. In the medium term, the superiority of high-field MRI with 7 T and more in selected patients, which has already been described in the experimental environment, will also bring further advantages in preoperative visualization in the clinical context.