Multimodal noninvasive evaluation in MRI-negative operculoinsular epilepsy

Restricted access


Presurgical evaluation of patients with operculoinsular epilepsy and negative MRI presents major challenges. Here the authors examined the yield of noninvasive modalities such as voxel-based morphometric MRI postprocessing, FDG-PET, subtraction ictal SPECT coregistered to MRI (SISCOM), and magnetoencephalography (MEG) in a cohort of patients with operculoinsular epilepsy and negative MRI.


Twenty-two MRI-negative patients were included who had focal ictal onset from the operculoinsular cortex on intracranial EEG, and underwent focal resection limited to the operculoinsular cortex. MRI postprocessing was applied to presurgical T1-weighted volumetric MRI using a morphometric analysis program (MAP). Individual and combined localization yields of MAP, FDG-PET, MEG, and SISCOM were compared with the ictal onset location on intracranial EEG. Seizure outcomes were reported at 1 year and 2 years (when available) using the Engel classification.


Ten patients (45.5%, 10/22) had operculoinsular abnormalities on MAP; 5 (23.8%, 5/21) had operculoinsular hypometabolism on FDG-PET; 4 (26.7%, 4/15) had operculoinsular hyperperfusion on SISCOM; and 6 (30.0%, 6/20) had an MEG cluster (3 tight, 3 loose) within the operculoinsular cortex. The highest yield of a 2-test combination was 59.1%, seen with MAP and SISCOM, followed by 54.5% with MAP and FDG-PET, and also 54.5% with MAP and MEG. The highest yield of a 3-test combination was 68.2%, seen with MAP, MEG, and SISCOM. The yield of the 4-test combination remained at 68.2%. When all other tests were negative or nonlocalizing, unique information was provided by MAP in 5, MEG in 1, SISCOM in 2, and FDG-PET in none of the patients. One-year follow-up was available in all patients, and showed 11 Engel class IA, 4 class IB, 4 class II, and 3 class III/IV. Two-year follow-up was available in 19 patients, and showed 9 class IA, 3 class IB, 1 class ID, 3 class II, and 3 class III/IV.


This study highlights the individual and combined values of multiple noninvasive modalities for the evaluation of nonlesional operculoinsular epilepsy. The 3-test combination of MAP, MEG, and SISCOM represented structural, interictal, and ictal localization information, and constituted the highest yield. MAP showed the highest yield of unique information when other tests were negative or nonlocalizing.

ABBREVIATIONS CCF = Cleveland Clinic Foundation; EZ = epileptogenic zone; FCD = focal cortical dysplasia; ICEEG = intracranial EEG; ILAE = International League Against Epilepsy; MAP = morphometric analysis program; MEG = magnetoencephalography; MPRAGE = magnetization-prepared rapid acquisition gradient echo; PMC = patient management conference; SAHZU = Second Affiliated Hospital of Zhejiang University; SEEG = stereo-EEG; SISCOM = subtraction ictal SPECT coregistered to MRI.

Article Information

Correspondence Zhong Irene Wang: Cleveland Clinic, Epilepsy Center, Cleveland, OH.

INCLUDE WHEN CITING Published online April 12, 2019; DOI: 10.3171/2018.12.JNS182746.

Disclosures Dr. Gonzalez-Martinez received an educational grant from Zimmer Biomet.

© AANS, except where prohibited by US copyright law.



  • View in gallery

    Schematic diagram showing the location distribution of the subgroups of patients included in our study. Red circles represent insular epilepsy; yellow circles represent operculoinsular epilepsy; green circles represent opercular epilepsy (hollow circles represent lesions located in the lateral aspect of the insula). Locations of the circles were approximate from ICEEG ictal onset. Figure is available in color online only.

  • View in gallery

    Examples of 4 MRI-negative patients with operculoinsular epilepsy (P5 = frontal opercular; P6 = anterior frontoparietal opercular; P8 = posterior frontoparietal opercular; P12 = frontal opercular and anterior insular) with MAP-positive findings shown by gray matter–white matter junction map, and one example of an MAP-negative patient (P17 = anterior insular). First column, presurgical T1-weighted and FLAIR images; second column, coregistered MAP junction file; third column, postsurgical MRI indicating resection of operculoinsular region. The red cross shows the location of subtle abnormalities. Fourth column, MRI overlaid with ICEEG. Red electrode contacts indicate ictal onset overlapping with MAP-positive findings in the operculoinsular region. Fifth column contains illustrations of FDG-PET, MEG, or SISCOM when concordant with ICEEG. P5, P6, P8, P12, P17 = case numbers. Figure is available in color online only.

  • View in gallery

    Localization yields of MAP, FDG-PET, SISCOM, and MEG, individually and combined (+).

  • View in gallery

    Distribution of patients who had localizing studies with MAP, MEG, and/or FDG-PET, illustrating their complementary values (MAP: solid ellipse; MEG: dotted ellipse; FDG-PET: dashed ellipse; SISCOM: dash-dotted ellipse). The numbers within the ellipsoids indicate the total localizing studies for each test. The number within the overlapping part of the ellipsoids indicates the number of patients with 2, 3, or 4 corresponding tests that had localizing findings. Take MAP for example: within the episode there are 10 localizing studies in total, of which 5 occurred when all other tests were not localized to the operculoinsular region (thus for these 5 studies, MAP had unique information). Similarly, unique information was seen on MEG in 1, on SISCOM in 2, and on FDG-PET in none.





All Time Past Year Past 30 Days
Abstract Views 466 466 87
Full Text Views 52 52 19
PDF Downloads 32 32 7
EPUB Downloads 0 0 0


Google Scholar