Patients with uncontrolled epilepsy may have impairment of language and memory associated with their seizures. In addition, it is very important to be able to map these functions when surgery is being considered as a possible treatment. In this study, we are using imaging methods to try to replace more invasive approaches to detecting the effects of epilepsy on language and memory, and for preoperative mapping. Methods: We use positron emission tomography (PET) and functional magnetic resonance imaging (fMRI)to map language and memory neuroimaging in patients with temporal lobe epilepsy to perform non-invasive evaluation of functional cortex, and study the effect of epilepsy on cognitive anatomy in children and adults with seizures. Our studies in both normal volunteers and patients with uncontrolled seizures have shown that imaging evaluation compares well with more invasive procdures such as electrostimulation mapping and the intracarotid sodium amytal test. We also evaluate the effect of seizures on the development of functional cognitive anatomy. Recent findings: We used fMRI and PET to study language subsystems in adults and children. we used O15 PET to identify cortical regions crucial for the computation of speech codes (i.e., the phonological system). Brain lesions may selectively impair the ability to read pseudowords (i.e., pronounceable non-words), leaving the ability to read familiar words largely intact. We found strong left lateralized epicenters of neural activity in rhyming both for words and pseudowords, suggesting conjoint neural networks for phonological processing. However, pseudoword rhyming recruited significantly more cortical tissue in the left posterior prefrontal cortex and the left inferior occipital-temporal junction, and activated the left supramarginal gyrus, which was not apparent in word rhyming. Brain activation shows relative regional variation with engagement of different language subsystems. Our results suggest that regions beyond traditional speech areas may be crucial for the computation of speech codes, reading, and comprehension in general. Although semantic categorization did not require computation of speech codes, it automatically activated the phonological system. The phonological system appears to be crucial for word reading, as well as text reading. Effects of multiple tasks in functional imaging have not been examined. Such cross-task interference would render functional mapping difficult to interpret. We found that including multiple cognitive tasks that engage in complex cognitive processes (e.g., semantics vs. phonology) in the same experimental session may induce brain activation uncharacteristic for cognitive processes under investigation. In order to activate temporal cortex, we used a semantic recall task based on our prior PET studies, adapted for reading, as well as a new fable reading task. Both paradigms proved robust activators of temporal cortex, as well as anterior language regions (possibly related to working memory or semantic recall), though reading was more strongly lateralized. It was particularly important that the tasks reliably activated left temporal language regions within as well as across subjects. Interestingly, none of our studies with either fMRI or PET, in contrast to some earlier reports, have shown significant gender effects on lateralization of activation for verbal fluency, auditory responsive naming, reading or rhyming tasks. Current studies: We are using several new tasks to investigate functional reorganization in patients, particularly children with epileptic foci in language processing regions.
Showing the most recent 10 out of 19 publications
