W. Van Paesschen

Canonical Correlation Analysis Applied to Remove Muscle Artifacts From the Electroencephalogram

The electroencephalogram (EEG) is often contaminated by muscle artifacts. In this paper, a new method for muscle artifact removal in EEG is presented, based on canonical correlation analysis (CCA) as a blind source separation (BSS) technique. This method is demonstrated on a synthetic data set. The method outperformed a low-pass filter with different cutoff frequencies and an independent component analysis (ICA)-based technique for muscle artifact removal. In addition, the method is applied on a real ictal EEG recording contaminated with muscle artifacts. The proposed method removed successfully the muscle artifact without altering the recorded underlying ictal activity

Quantitative hippocampal MRI and intractable temporal lobe epilepsy

Objectives: To evaluate and compare T2 relaxometry and volumetrics of hippocampus in the presurgical evaluation of patients with intractable temporal lobe epilepsy (TLE), and to correlate these quantitative MRI measures with the pathology of the resected hippocampus. Patients: Forty patients with intractable TLE who underwent presurgical evaluation and subsequent temporal lobe surgery. Main outcome measures: Hippocampal T2 (HCT2), volumes of hippocampi and hippocampal volume ratio (HCVR) (volume of hippocampus with higher HCT2 divided by volume of hippocampus with lower HCT2), and qualitative pathology. Results: Thirty-two patients had hippocampal sclerosis, three patients had end-folium sclerosis, one patient had amygdala sclerosis, and four patients had a foreign tissue lesion in the temporal lobe. HCT (2) ratio (R/L) correlated inversely with the ratio of hippocampal volumes (R/L) (r equals minus 0.91; p less than 0.0001). A high T2 signal in an atrophic hippocampus was characteristic of hippocampal sclerosis. All patients with hippocampal sclerosis had an HCVR below control values, and only one of these had an HCT2 in the normal range. HCVR produced one false-positive result. The patients with end-folium sclerosis had normal HCT2 and HCVR. The patient with amygdala sclerosis had a normal hippocampus on qualitative and quantitative assessment. Of the four patients with a lesion, one had a mildly increased HCT2 and one had mild volume asymmetry. Hippocampal volume asymmetry could be reliably detected on visual inspection of the MRI with an HCVR of 0.85 or less, and an increase of HCT2 with a T2 of 115 msec or higher. Conclusion: Quantitative MRI combining HCT2 and HCVR is a reliable method for diagnosing hippocampal sclerosis noninvasively. End-folium sclerosis and amygdala sclerosis should be considered in patients with intractable TLE and negative findings on MRI studies, including quantitative measures of the hippocampus. NEUROLOGY 1995;45: 2233-2240

The Relationship Between Quantitative MRI and Neuropsychological Functioning in Temporal Lobe Epilepsy

Summary: Purpose: Quantitative MRI techniques provide an unparalleled opportunity to examine in vivo the relationship between the extent and laterality of hippocampal pathology and associated neuropsychological deficits. The purpose of this study was to examine the nature of the relationship between quantitative measures of hippocampal pathology and neuropsychological measures, using a multivariate approach.

Proton magnetic resonance spectroscopy in MRI-negative temporal lobe epilepsy

Objective: To determine the 1H MRS findings in patients with intractable temporal lobe epilepsy (TLE) who had no detectable abnormality on either qualitative or quantitative MRI. Background: Previous work has indicated that ∼20% of patients with TLE remain MRI negative after extensive qualitative and quantitative investigation. Single-voxel 1H MRS provides a means of identifying potentially diffuse disease. Methods: Seven patients with intractable TLE, normal routine MRI, normal hippocampal volumes, and normal hippocampal and amygdala quantitative T2 values underwent single-voxel 1H MRS of the medial temporal lobes. The results are compared with those from 13 control subjects and 15 patients with evidence of hippocampal sclerosis (HS). Results: The ratio N-acetylaspartate/(cholinecreatine + phosphocreatine) was abnormally low in five of the seven MRI-negative patients. In two of these patients, the ratios were low bilaterally. The observed MRS ratios in the MRI-negative group with abnormal EEG were similar to those from temporal lobes ipsilateral to HS, suggesting the presence of widespread or diffuse disease of a similar degree in both groups. Conclusions: These results demonstrate that MRS can provide evidence of temporal lobe abnormalities in TLE patients who show no abnormality on extensive MRI investigation.

Correlations of interictal FDG-PET metabolism and ictal SPECT perfusion changes in human temporal lobe epilepsy with hippocampal sclerosis

BACKGROUND The pathophysiological role of the extensive interictal cerebral hypometabolism in complex partial seizures (CPS) in refractory mesial temporal lobe epilepsy with hippocampal sclerosis (mTLE-HS) is poorly understood. Our aim was to study ictal-interictal SPECT perfusion versus interictal fluorodeoxyglucose (FDG)-PET metabolic patterns. METHODS Eleven adults with refractory unilateral mTLE-HS, who were rendered seizure free after epilepsy surgery, were included. All had an interictal FDG-PET and an interictal and ictal perfusion SPECT scan. FDG-PET data were reconstructed using an anatomy-based reconstruction algorithm, which corrected for partial volume effects, and analyzed semi-quantitatively after normalization to white matter activity. Using Statistical Parametric Mapping (SPM), we compared interictal metabolism of the patient group with a control group. We correlated metabolic with ictal perfusion changes in the patient group. RESULTS Global cerebral grey matter glucose metabolism in patients was decreased 10-25% compared with control subjects. Interictal PET hypometabolism and ictal SPECT hypoperfusion were maximal in the ipsilateral frontal lobe. Ictal frontal lobe hypoperfusion was associated with crossed cerebellar diaschisis. The ipsilateral temporal lobe showed maximal ictal hyperperfusion and interictal hypometabolism, which was relatively mild compared with the degree of hypometabolism affecting the frontal lobes. CONCLUSION Interictal hypometabolism in mTLE-HS was greatest in the ipsilateral frontal lobe and represented a seizure-related dynamic process in view of further ictal decreases. Crossed cerebellar diaschisis suggested that there is a strong ipsilateral frontal lobe inhibition during CPS. We speculate that surround inhibition in the frontal lobe is a dynamic defense mechanism against seizure propagation, and may be responsible for functional deficits observed in mTLE.

The amygdala and intractable temporal lobe epilepsy A quantitative magnetic resonance imaging study

Objective: To establish a quantitative MRI technique using T2 relaxation time mapping to study systematically the amygdala in patients with intractable temporal lobe epilepsy (TLE). Background: Identification of a focal abnormality on MRI in patients with intractable TLE is important, because outcome from surgery depends largely on the removal of the underlying pathology. Hippocampal sclerosis (HS) is the most common cause of intractable TLE, but epileptogenic lesions can be confined to the amygdala. Methods: Twenty control subjects and 82 patients with intractable TLE were studied. Patients who had foreign tissue lesions visible on routine MRI were excluded. All subjects had a hippocampal T2 map and volumetry and an amygdala T2 (AT2) map. Results: Forty-four of the 82 patients (54%) had an abnormal AT2, which was bilateral in 18. Forty-four patients (54%) had unilateral HS on MRI, 25 (57%) of whom had an abnormal AT2. Seven patients (8%) had bilateral HS, four of whom had an abnormal AT2. Thirty-one patients (38%) had normal quantitative hippocampal measures, 15 of whom had an abnormal AT2, which was bilateral in seven. Fluid attenuated inversion recovery (FLAIR) imaging, where appropriate, confirmed that the increased AT2 signal was due to parenchymal changes. Neuropathologic correlates of an increased AT2 included microdysgenesis in one and gliosis in three patients. Patients with an isolated AT2 abnormality were significantly older at the onset of habitual epilepsy and rarely had a history of febrile convulsions, in comparison with patients who had HS. An isolated AT2 abnormality correlated well with interictal EEG findings. Conclusions: The combination of AT2 mapping and FLAIR is a sensitive method to detect lesions that are not seen on routine MRI in the amygdalae of patients with intractable TLE. Further correlational studies will be required to define the role of this technique in the presurgical evaluation of patients with intractable TLE. NEUROLOGY 1996;47: 1021-1031

Anatomical-based FDG-PET reconstruction for the detection of hypo-metabolic regions in epilepsy

Positron emission tomography (PET) of the cerebral glucose metabolism has shown to be useful in the presurgical evaluation of patients with epilepsy. Between seizures, PET images using fluorodeoxyglucose (FDG) show a decreased glucose metabolism in areas of the gray matter (GM) tissue that are associated with the epileptogenic region. However, detection of subtle hypo-metabolic regions is limited by noise in the projection data and the relatively small thickness of the GM tissue compared to the spatial resolution of the PET system. Therefore, we present an iterative maximum-a-posteriori based reconstruction algorithm, dedicated to the detection of hypo-metabolic regions in FDG-PET images of the brain of epilepsy patients. Anatomical information, derived from magnetic resonance imaging data, and pathophysiological knowledge was included in the reconstruction algorithm. Two Monte Carlo based brain software phantom experiments were used to examine the performance of the algorithm. In the first experiment, we used perfect, and in the second, imperfect anatomical knowledge during the reconstruction process. In both experiments, we measured signal-to-noise ratio (SNR), root mean squared (rms) bias and rms standard deviation. For both experiments, bias was reduced at matched noise levels, when compared to post-smoothed maximum-likelihood expectation-maximization (ML-EM) and maximum a posteriori reconstruction without anatomical priors. The SNR was similar to that of ML-EM with optimal post-smoothing, although the parameters of the prior distributions were not optimized. We can conclude that the use of anatomical information combined with prior information about the underlying pathology is very promising for the detection of subtle hypo-metabolic regions in the brain of patients with epilepsy.

11C-flumazenil PET, volumetric MRI, and quantitative pathology in mesial temporal lobe epilepsy

Background: Using statistical parametric mapping and 11C-flumazenil (FMZ) PET we have previously shown reduction of central benzodiazepine receptor (cBZR) binding restricted to the hippocampus in mesial temporal lobe epilepsy (mTLE) due to hippocampal sclerosis (HS). The limited spatial resolution of PET, however, results in partial-volume averaging that affects quantitative analysis of cBZR density. Method: We determined hippocampal volume loss and reduction in cBZR binding using an MRI-based method for partial-volume effect correction of 11C-FMZ volume of distribution (FMZ-Vd) in 17 patients with refractory mTLE and an MRI diagnosis of HS that was subsequently histologically verified in all cases. Quantitative neuropathology was performed with assessment of neuron density in 14 of the 17 patients. Absolute FMZ-Vd and asymmetry indices(FMZ-AI) were compared before and after partial-volume effect correction with MRI-determined hippocampal volumes (HCV), hippocampal T2 measurements, and, if available, neuronal cell densities. Results: Compared with 15 age-matched healthy volunteers, significant reductions of absolute hippocampal FMZ-Vd were found before correction for partial-volume effects in 11 of 17 patients (65%) and only abnormal FMZ-AI in the other six patients. After partial-volume effect correction all 17 patients (100%) showed both significant unilateral reduction of absolute FMZ-Vd and abnormal FMZ-AI. There was no correlation between corrected absolute FMZ-Vd and HCV or neuronal cell density. After correction for partial-volume effect we found a mean 38% reduction of FMZ-Vd in the sclerosed hippocampus, over and above the reduction of HCV. Conclusion: Correction for partial-volume effect allows absolute quantitation of FMZ-PET and increases its sensitivity for detecting abnormalities in TLE due to HS. The lack of correlation between cBZR binding and neuronal density implies that atrophy with neuron loss is not the sole determinant of reduced cBZR binding in patients with mTLE and hippocampal sclerosis.

Paroxysmal choreoathetosis/spasticity (DYT9) is caused by a GLUT1 defect

Objective: Mutations in SLC2A1, encoding the glucose transporter type 1 (GLUT1), cause a broad spectrum of neurologic disorders including classic GLUT1 deficiency syndrome, paroxysmal exercise-induced dyskinesia (PED, DYT18), and absence epilepsy. A large German/Dutch pedigree has formerly been described as paroxysmal choreoathetosis/spasticity (DYT9) and linked close to but not including the SLC2A1 locus on chromosome 1p. We tested whether 1) progressive spastic paraparesis, in addition to PED, as described in DYT9, and 2) autosomal dominant forms of hereditary spastic paraparesis (HSP) without PED are caused by SLC2A1 defects. Methods: The German/Dutch family and an Australian monozygotic twin pair were clinically (re-)investigated, and 139 index cases with dominant or sporadic HSP in which relevant dominant genes were partially excluded were identified from databanks. SLC2A1 was sequenced in all cases in this observational study and the functional effects of identified sequence variations were tested in glucose uptake and protein expression assays. Results: We identified causative mutations in SLC2A1 in both families, which were absent in 400 control chromosomes, cosegregated with the affection status, and decreased glucose uptake in functional assays. In the 139 index patients with HSP without paroxysmal dyskinesias, we only identified one sequence variation, which, however, neither decreased glucose uptake nor altered protein expression. Conclusions: This study shows that DYT9 and DYT18 are allelic disorders and enlarges the spectrum of GLUT1 phenotypes, now also including slowly progressive spastic paraparesis combined with PED. SLC2A1 mutations were excluded as a cause of HSP without PED in our cohort.

Automatic Removal of Ocular Artifacts in the EEG without an EOG Reference Channel

This paper presents a method for removing electroocular (EOG) artifacts in the electroencephalogram (EEG). The procedure is based on blind source separation (BSS) and, in contrast to methods already available in the literature, it is completely automated and does not require the availability of peri-ocular EOG electrodes. The proposed approach removed most EOG artifacts in 6 long-term EEG recordings containing epilectic seizures without distorting the recorded ictal activity