The goals of this project are: (i) to generate an empirical model to predict the long-term language outcome following epilepsy surgery and (ii) to better understand 'how' the language system works during speech. About 1% of the general population has epilepsy, while one-fifth of epilepsy is medically intractable. Subsets of patients with intractable focal epilepsy benefit from surgical resection of the seizure focus with functionally-important areas preserved. Yet, in reality, accurate identification of language areas is difficult, especially in children, since electrical stimulation mapping lacks sufficient sensitiity, often takes hours to complete, and has a risk of stimulation-induced seizures. In the first funding period, we demonstrated that naming-related augmentation of gamma activity (50-120 Hz) on electrocorticography (ECoG) recording can delineate the language circuitry, and that surgical damage of sites showing such gamma-augmentation predicted the acute postoperative language outcome better than electrical stimulation mapping. An important next step is to determine how well the long-term language outcome can be predicted, since some but not all children recover language function well after the resection of language networks. To maximize the predictive performance, we will determine the language cortex and subcortical pathway, while combining ECoG gamma mapping with diffusion weighted imaging (DWI) fiber tractography. Furthermore, the prediction model will take into account the chronic effect of functional recovery in addition to the acute effect of damaged language networks on neuropsychological outcome measures. This project is significant since the results will be directly translatable into patient management, and our innovative multimodality technique has the potential to become a mainstream method to localize functionally-important brain structures. We will also determine the anatomical and functional connectivity within the language networks, using ECoG gamma mapping, DWI tractography and cortico-cortical evoked potentials (CCEPs). Theoretical models of human speech propose that phonologic and semantic information is transferred, via the arcuate fasciculus, between the temporal and frontal language areas. Yet, the exact location of each arcuate pathway for phonologic and semantic information has not been elucidated. Furthermore, directional efficiency of signal transferring in each pathway has not been clarified, although a modern theoretical model indicates the presence of bi-directional signal transferring between the frontal and temporal lobes. In this project, we will specifically determine if these sites involved in phonological and semantic functions are differentially connected via distinct arcuate fasciculus fibers. We will also determine 'in which direction' electrical signals propagate more efficiently within and across the two lobes involved in language. The model refined or revised by this empirical study will help in prediction of specific language symptoms following focal brain damage, and ultimately may better localize the therapeutic targets for improving different types of language impairments in neurological diseases.
The goals of this project are: (i) to validate a new language mapping method to predict the long-term language outcome following epilepsy surgery and (ii) to better understand how the language system works during speech. Augmentation of intracranial EEG activity at 50-120 Hz during speech will localize cortical sites involved in language. Diffusion weighted imaging fiber tractography (a quantitative MRI technique) will localize subcortical pathways connecting such language cortical areas. Propagation of electrical signals safely evoked by tiny electrical pulses will determine how cortical signals are transferred from a language cortex to other areas. A two-year follow-up of neuropsychological function will determine what diagnostic techniques and what patient profile are useful to minimize the risk of long-term language deficits following brain surgery. Combination of our language mapping techniques will also determine how cortical areas processing 'meaning' and 'sounds' of words are connected within the temporal and frontal lobes. The results of this study will help in prediction of specific language symptoms following brain damage, and ultimately may identify the therapeutic targets for improving different types of language impairments in neurological diseases.
Showing the most recent 10 out of 65 publications
