A method for automated volumetry to assess disease progression in Rasmussen Encephalitis

Introduction: Rasmussen Encephalitis (RE) is a chronic inflammatory disease of the brain that affects and subsequently destroys one hemisphere and is causing intractable seizures and progressive neurologic deficits. Manual planimetric volumetry based on single slices of MRI-studies has been performed before, resulting in a number of indices to assess disease progression of which the hemispheric ratio proved to be the most valuable. We have applied a novel automated volumetric approach to assess disease progression in Rasmussen Encephalitis.

Methods: Structural MRI was performed on a 3T scanner (Siemens Magnetom Trio) with an 8-channel head coil. 3D-T1-weighted images were obtained using an MP-RAGE sequence with 160 slices (resolution 1.0mm³). FSL tools were used for brain-extraction and segmentation of the whole-brain. The 3D-dataset was then transformed to the MNI-space with an in-house script using Flirt and Fnirt. The left and right hemispheres were isolated using predefined masks and thereby excluding cerebellum, brainstem and the opposite hemisphere. The hemispheres are then back-transformed using the calculated inverse transformation matrices. Whole brain and single hemisphere volumes as well as the HR were calculated using FSLmath and FSLstats (HR=volume of the affected hemisphere divided by the volume of the unaffected hemisphere). Until now 10 patients were evaluated using the approach.

Results and Conclusion: The HR has proven to be a valuable parameter to assess disease progression in RE patients. As manual volumetry is based only on single slices and is time consuming, automated approaches are needed. We could show that automated serial MRI volumetry is feasible in both adult and pediatric RE patients. HR is especially valuable in children because it shows disease progression even in the growing brain.

Fig.1: Serial MRI-based volumetry of a patient with the adult form of RE (born 1955) between 2007 and 2010

Fig.2: Serial MRI-based volumetry of a patient with RE (born 1994) between 2007 and 2010