Steven L. Weinstein

Noninvasive assessment of language dominance in children and adolescents with functional MRI: a preliminary study.

Background Assessment of language organization is crucial in patients considered for epilepsy surgery. In children, the current techniques, intra-carotid amobarbital test (IAT) for language dominance, and cortical electrostimulation mapping (ESM), are invasive and risky. Functional magnetic resonance imaging (fMRI) is an alternative method for noninvasive functional mapping, through the detection of the hemodynamic changes associated with neuronal activation. We used fMRI to assess language dominance in children with partial epilepsy. Methods Eleven right handed children and adolescents performed a word generation task during fMRI acquisition focused on the frontal lobes. Areas where the signal time course correlated with the test paradigm (r = 0.7) were considered activated. Extent and magnitude of signal changes were used to calculate asymmetry indices. Seven patients had IAT, ESM, or surgery outcome available for comparison. Results fMRI language dominance always agreed with IAT (6 cases) and ESM (1 case), showing left dominance in six and bilateral language in one. fMRI demonstrated left dominance in three additional children, and right dominance in one with early onset of left temporal epilepsy. Four children whose initial studies were equivocal due to noncompliance or motion artifacts were restudied successfully. Conclusions fMRI can be used to assess language lateralization noninvasively in children. It has the potential to replace current functional mapping techniques in patients, and to provide important data on brain development.

Language dominance in partial epilepsy patients identified with an fMRI reading task

Background fMRI language tasks readily identify frontal language areas; temporal activation has been less consistent. No studies have compared clinical visual judgment to quantitative region of interest (ROI) analysis. Objective To identify temporal language areas in patients with partial epilepsy using a reading paradigm with clinical and ROI interpretation. Methods Thirty patients with temporal lobe epilepsy, aged 8 to 56 years, had 1.5-T fMRI. Patients silently named an object described by a sentence compared to a visual control. Data were analyzed with ROI analysis from t-maps. Regional asymmetry indices (AI) were calculated ([L−R]/[L+R]) and language dominance defined as >0.20. t-Maps were visually rated by three readers at three t thresholds. Twenty-one patients had intracarotid amobarbital test (IAT). Results The fMRI reading task provided evidence of language lateralization in 27 of 30 patients with ROI analysis. Twenty-five were left dominant, two right, one bilateral, and two were nondiagnostic; IAT and fMRI agreed in most patients, three had partial agreement, none overtly disagreed. Interrater agreement ranged between 0.77 to 0.82 (Cramer V;p < 0.0001); agreement between visual and ROI reading with IAT was 0.71 to 0.77 (Cramer V;p < 0.0001). Viewing data at lower thresholds added interpretation to 12 patients on visual analysis and 8 with ROI analysis. Conclusions An fMRI reading paradigm can identify language dominance in frontal and temporal areas. Clinical visual interpretation is comparable to quantitative ROI analysis.

fMRI language task panel improves determination of language dominance.

Background: fMRI language tasks reliably identify language areas in presurgical epilepsy patients, but activation using single paradigms may disagree with the intracarotid amobarbital test (IAT). Objective: To determine whether a panel of fMRI tasks targeting different aspects of language processing increases accuracy in determining hemisphere language dominance. Methods: Twenty-six patients age 12 to 56 years, predominantly with temporal lobe epilepsy, were studied using whole-brain 1.5 T fMRI (echo planar imaging, blood oxygenation level–dependent) with three task categories using a block design: verbal fluency, reading comprehension, and auditory comprehension. fMRI t maps were visually rated at three thresholds. All patients had assessment of language lateralization by IAT. Results: fMRI showed left dominance in 21 patients, right dominance in 2, and bilateral activation in 2; raters disagreed over a left vs right bilateral rating in 1 patient. There was full agreement between IAT and fMRI in 21 of 25 patients (IAT failed in 1). In three instances of partial disparity with IAT, the fMRI panel showed consistent findings across raters. Agreement between raters was excellent (partial disagreement in only one patient); the panel of tasks was superior to any single task for interrater agreement (Cramer V 0.93 [range 0.91 to 1.0] vs 0.72 [range 0.60 to 0.86]). Conclusions: A panel of fMRI language paradigms may be more accurate for evaluating partial epilepsy patients than a single task. A panel of tasks reduces the likelihood of nondiagnostic findings, improves interrater reliability, and helps confirm language laterality.

Early seizure detection.

Summary: For patients with medically intractable epilepsy, there have been few effective alternatives to resective surgery, a destructive, irreversible treatment. A strategy receiving increased attention is using interictal spike patterns and continuous EEG measurements from epileptic patients to predict and ultimately control seizure activity via chemical or electrical control systems. This work compares results of seven linear and nonlinear methods (analysis of power spectra, cross‐correlation, principal components, phase, wavelets, correlation integral, and mutual prediction) in detecting the earliest dynamical changes preceding 12 intracranially‐recorded seizures from 4 patients. A method of counting standard deviations was used to compare across methods, and the earliest departures from thresholds determined from non‐seizure EEG were compared to a neurologists judgement. For these data, the nonlinear methods offered no predictive advantage over the linear methods. All the methods described here were successful in detecting changes leading to a seizure between one and two minutes before the first changes noted by the neurologist, although analysis of phase correlation proved the most robust. The success of phase analysis may be due in part to its complete insensitivity to amplitude, which may provide a significant source of error.

Atypical language in lesional and nonlesional complex partial epilepsy

Objective: We investigated the relationship between partial epilepsy, MRI findings, and atypical language representation. Methods: A total of 102 patients (4 to 55 years) with left hemisphere epileptogenic zones were evaluated using three fMRI language tasks obtained at 1.5 or 3T with EPI BOLD techniques: verbal fluency, reading comprehension, and auditory comprehension. fMRI maps were visually interpreted at a standard threshold and rated as left or atypical language. Results: Atypical language dominance occurred in 30 patients (29%) and varied with MRI type (p < 0.01). Atypical language representation occurred in 36% (13/36) with normal MRI, 21% (6/29) with mesial temporal sclerosis, 14% (4/28) with focal cortical lesions (dysplasia, tumor, vascular malformation), and all (6/6) with a history of stroke. Multivariate logistic regression analysis found handedness, seizure onset, and MRI type accounted for much of the variance in language activation patterns (χ2 = 24.09, p < 0.01). Atypical language was more prevalent in patients with early seizure onset (43.2%, p < 0.05) and atypical handedness (60%, p < 0.01). None of the three clinical factors were correlated with each other (p > 0.40). Patients with atypical language had lower verbal abilities (F = 6.96, p = 0.01) and a trend toward lower nonverbal abilities (F = 3.58, p = 0.06). There were no differences in rates of atypical language across time, age groups, or MRI scanner. Conclusion: Early seizure onset and atypical handedness, as well as the location and nature of pathologic substrate, are important factors in language reorganization. GLOSSARY: FOV = field of view; MTS = mesial temporal sclerosis; RRN = read response naming; TE = echo time; TR = repetition time; WAIS = Wechsler Adult Intelligence Scale; WISC = Wechsler Intelligence Scale for Children.

Neuronal spatiotemporal pattern discrimination: The dynamical evolution of seizures

We developed a modern numerical approach to the multivariate linear discrimination of Fisher from 1936 based upon singular value decomposition that is sufficiently stable to permit widespread application to spatiotemporal neuronal patterns. We demonstrate this approach on an old problem in neuroscience--whether seizures have distinct dynamical states as they evolve with time. A practical result was the first demonstration that human seizures have distinct initiation and termination dynamics, an important characterization as we seek to better understand how seizures start and stop. Our approach is broadly applicable to a wide variety of neuronal data, from multichannel EEG or MEG, to sequentially acquired optical imaging data or fMRI.

Seizure focus affects regional language networks assessed by fMRI.

Objective: To investigate the degree of language dominance in patients with left and right hemisphere seizure foci compared to normal volunteers using a fMRI reading comprehension task. Methods: Fifty patients with complex partial epilepsy, aged 8 to 56 years and 33 normal volunteers, aged 7 to 34 had fMRI (1.5 T) and neuropsychological testing. Participants silently named an object described by a sentence compared to a visual control. Data were analyzed with region of interest (ROI) analysis based on t maps for inferior frontal gyrus (IFG), midfrontal gyrus (MFG), and Wernicke area (WA). Regional asymmetry indices (AIs) were calculated [(L − R)/(L + R)]; AI >0.20 was deemed left dominant and AI <0.20 as atypical language. Results: Left hemisphere focus patients had a higher likelihood of atypical language than right hemisphere focus patients (21% vs 0%, χ2 < 0.002). Left hemisphere focus patients, excluding those with atypical language, had lower regional AI in IFG, MFG, and WA than controls. Right hemisphere focus patients were all left language dominant and had a lower AI than controls in WA and MFG, but not for IFG. AI in MFG and WA were similar between left hemisphere focus/left language patients and right hemisphere focus patients. Patients activated more voxels than healthy volunteers. Lower AIs were attributable to greater activation in right homologous regions. Less activation in the right-side WA correlated with better verbal memory performance in right focus/left hemisphere-dominant patients, whereas less strongly lateralized activation in IFG correlated better with Verbal IQ in left focus/left hemisphere-dominant patients. Conclusions: Patients had lower asymmetry indices than healthy controls, reflecting increased recruitment of homologous right hemisphere areas for language processing. Greater right hemisphere activation may reflect greater cognitive effort in patient populations, the effect of epilepsy, or its treatment. Regional activation patterns reflect adaptive efforts at recruiting more widespread language processing networks that are differentially affected based on hemisphere of seizure focus.

Parenting stress and childhood epilepsy: the impact of depression, learning, and seizure-related factors.

OBJECTIVE The purpose of this study was to evaluate stress in parents of children with epilepsy relative to the impact of childhood depression, learning disorders, and seizure-related risk factors. METHODS Sixty-five parents and their children completed the Parenting Stress Index, Child Depression Inventory, and behavior and demographic forms. Kruska-Wallis ANOVAs and Spearmans rank correlations were used in a cross-sectional study design. RESULTS High levels of stress were found among the parents (45%). Overall, child depression (23%) was found to significantly increase the distress parents experienced in their role (P<0.05). Another risk factor found to impact parenting stress was learning disabilities (P<0.01). The seizure-related factors of polytherapy, duration, and age at onset were correlated with depression (P<0.05). CONCLUSION To effectively manage children with epilepsy, assessments of depression and learning must be considered because of their potential impact on parenting stress and the childs overall quality of life.

Multivariate linear discrimination of seizures

OBJECTIVE To discriminate seizures from interictal dynamics based on multivariate synchrony measures, and to identify dynamics of a pre-seizure state. METHODS A linear discriminator was constructed from two different measures of synchronization: cross-correlation and phase synchronization. We applied this discriminator to a sequence of seizures recorded from the intracranial EEG of a patient monitored over 6 days. RESULTS Surprisingly, we found that this bivariate measure of synchronization was not a reliable seizure discriminator for 7 of 9 seizures. Furthermore, the method did not appear to reliably detect a pre-seizure state. An association between anti-convulsant dosage, frequency of clinical seizures, and discriminator performance was noted. CONCLUSIONS Using a bivariate measure of synchronization failed to reliably differentiate seizures from non-seizure periods in these data, nor did such methods show reliable detection of a synchronous pre-seizure state. The non-stationary variables of decreasing antiepileptic medication (without available serum concentration measurements), and concomitant increasing seizure frequency contributed to the difficulties in validating a seizure prediction tool on such data. SIGNIFICANCE The finding that these seizures were not a simple reflection of increasing synchronization in the EEG has important implications. The non-stationary characteristics of human post-implantation intracranial EEG is an inherent limitation of pre-resection data sets.

Low incidence of abnormal 18FDG-PET in children with new-onset partial epilepsy A prospective study

ObjectivePatients with refractory partial epilepsy often exhibit regional hypometabolism. It is unknown whether the metabolic abnormalities are present at seizure onset or develop over time. MethodsThe authors studied 40 children within 1 year of their third unprovoked partial seizure with EEG, MRI, and [18F]-fluorodeoxyglucose (18FDG)-PET (mean age at seizure onset = 5.8 years, range 0.9 to 11.9 years; mean epilepsy duration = 1.1 years, range 0.3 to 2.3 years; mean number of seizures = 30, range 3 to 200). The authors excluded children with abnormal structural MRI, except four with mesial temporal sclerosis and two with subtle hippocampal dysgenesis. 18FDG-PET was analyzed with a region of interest template. An absolute asymmetry index, |AI|, greater than 0.15 was considered abnormal. ResultsThirty-three children had a presumptive temporal lobe focus, five frontotemporal, and two frontal. Mean AI for all regions was not different from 10 normal young adults, even when children less likely to have a temporal focus were excluded. Eight of 40 children (20%) had focal hypometabolism, all restricted to the temporal lobe, especially inferior mesial and inferior lateral regions. Abnormalities were ipsilateral to the presumed temporal lobe ictal focus. ConclusionsAbnormalities of glucose utilization may be less common and profound in children with new-onset partial seizures than in adults with chronic partial epilepsy. Although these patients’ prognosis is uncertain, resolution of epilepsy after three documented seizures is uncommon. If the subjects develop a higher incidence of hypometabolism in the future with planned follow-up studies, metabolic dysfunction may be related to persistent epilepsy rather than present at seizure onset.