The Relationship between Magnetic Source Imaging (MSI) of Epileptic Spikes and the Findings of the Magnetic Resonance Imaging, Positron Emission Tomography (PET), and Single Photon Emission CT (SPECT) in Epileptic Patients

Abstract

Purpose: Several noniiivasive imaging methods can detect epileptic ioci. We compared tnagnetic source imaging (MSI) of epilcptic spikes with the results of other imaging mcthoda, including positron emission tomography (PET), single photon emission CT (SPECT), and MRI, in order to determine characteristics of each. Methods: We studied 9 patients, age ranging from 1–23 years. Etiologies included acute encephalopathy, tubcrous sclerosis, pcrisylvian syndrome, and cerebral infarct in I patient each, unknown in the rest. Eight exhibited bca/ization‐related epilepsy (LRE) and the remaining iut syndrome (LGS). Informed cnsent was obtained from the patients or their parents. Thc magnctocncephalograins (MEG) were measured in 21 magnetically shielded room using a layingtype whole‐cortex MEG with 160‐channel first‐order gradiometers. We estimated the equivalent current dipoles (ECD) of spikes in the simultaneously recorded EEG. MRI scanning was performed at 5 m m thickness in 3 directions including the markers for superimposition with MEG. IXF‐FDG PET was performed with a Headtome IV in 4 paticnts and 123Tc ECD SPECT with a Toshiba GCA9300 in 8 patients. We visually evaluated these images. Results: We could estimate the ECD in 8 patients as having I or multiple locations. All 4 patients examined by PET cxhibitcd 1 or more hypometabolic arcas on PET. SPECT studies revealed hypoperfusion areas in 3 patients m d hyperperfusion areas in 2 patients. One id the patients with hypcrpcrfusion was on ictal study. Structural ahnormalities were presented on MRI in 3 patients. Casc I had advcrsivc seizures with turning to the lcft and exhibited a broad right frontotemporal hypometabolic area on PET. Her ECDs wcrc cstimatcd to originate from the right prcmotor area, in agreement with a hypoperfusion spot on ictal SPECT. Case 2 had epilepsy as a residuum of past cnccphalopathy and had complcx partial seizures (CPS) with rapid generalim tion. The foci detected on MSI, PET, and SPECT did not agree. Casc3 had CPS and anterotemporal hypomctabolism on PET. ECDs were focused in the lateral mterotemporal area. Case 4 had tuberous xlc‐ rosis and a past history of status cpilcpticus. MRI revealed cortical tubers and white matter lesions. Cortical tubers exhibited hypoperlu‐ sion on SPECT. However, ECDs werc cstimatcd to originate from the right occipital hypoperfusion area and not thc cortical tubcrs. C was suspected to have idiopathic LRE and exhibited no focus on PET and SPECT. ECDs were estimated to originate from the left frontal cortex. Case 6 had pcrisylvian syndrome with LGS. Bilaterally abnormal cortex exhibited hyperperfusion on SPECT, and ECDs wcrc citimated to originate lrom this cortex. Casc 7 had CPS and left temporal sporadic spikes on previous EEGs. The findings of SPECT and PET studies were normal, and ECDs could not be estimated because there was no spike in the MEG study. Case 8 had right leg pain and frontal slow wave bursts on the EEG. PET rcvcaled right frontal and left parietal hypometabolism, and ECDs were estimated to originate from the left parietal area. Case 9 had an old inraretion in the lcft occipital lobe and CPS. SPECT revealed perfusion defect in the infarction and right frontal hypoperfusion. ECDs were estimated to originatc from the margin of the infrarction. Conclusions: Interictal MEG studics can dctcct irritable zones of epileptic foci, hut their findings sometimes do not accurately reveal the epileptogcnic zone. However the majority of our cstimatcd ECDs agreed with the ictal symptomatology and the foci detected by PET and SPECT. The estimated area of ECDs was usually smaller than hypometabolic areas on PET and hypoperfusion area on SPECT. Multimodal interictal studies with noninvasive technologics will enablc more enable detection of epileptic foci.