Project |01 Predicting seizure severity and duration in patients with temporal lobe epilepsy using EEG activity tracker data
In this project we build predictive models on data from sensors on patients with severe epilepsy. These patients were in the hospital and were monitored continously for as much as a week. In reviewing data from autism studies, we have found that seizures were sometimes preceded by huge spikes in skin conductance. It seemed that their sensors might actually be able to predict the onset of seizures. Here we use a logistic regression combined with k-means classification on activity tracker data to predict the onset and duration of siezures in TLE patients. We use a novel wrist activity tracker sensor combined with a cranial EEG sensor for the same.
Manuscript in publication. Patent pending.
Project |02 Logistic models for the optogenetical silencing of Dentate Granule Cells
The development of temporal lobe epilepsy, the most common form of Epilepsy, is associated with altered proliferation and impaired integration of adult-generated Dentate Granule Cells (DGCs). These DGCs normally integrate into the granule cell layer of the dentate gyrus and are important for limiting the flow of excitation within the hippocampus.
Our lab and others have previously shown that hippocampal DGCs born shortly before or after an epileptogenic brain insult, termed Status Epilepticus (SE), display multiple morphological/physiological abnormalities and integrate aberrantly into the dentate circuitry.
These abnormal cells mediate the formation of anomalous recurrent excitatory circuits within the dentate, increasing hippocampal excitability and potentially contributing to the occurrence of spontaneous recurrent seizures (the defining feature of epilepsy). Previous studies using cell ablation strategies have produced some encouraging results implicating adult born DGCs in epileptogenesis, however, potential non-specific effects of cell loss have lead many investigators to view these results as inconclusive. We hypothesize that adult born DGCs are required for the progression of epileptiform activity and accordingly propose that silencing adult born DGCs post SE would lead to a blockage of this epileptogenic activity in-vitro. The current study is designed to investigate whether DGCs born before or after SE play a role in seizure progression, and whether silencing hilar ectopic DGCs, born after SE, can alone sustain a seizure. The ability to transiently and reversibly silence these cells provides a much more powerful test of their role over previous approaches. We use optogenetics data to build logistic models which can predict the silencing/no silencing of these DGC in-vivo. The results of this study will define the role of adult born DGCs in seizure progression and will serve as a stepping stone to further designing a DGC silencing based therapeutic approach for post-stroke seizures and epilepsy.
Project |03 The Kindling Study
Temporal lobe epilepsy is associated with changes in the morphology of hippocampal dentate granule cells, including formation of hilar basal dendrites. These changes are evident in numerous models that are associated with substantial neuron loss and spontaneous recurrent seizures. By contrast, previous studies have shown that in the kindling model, it is possible to administer a limited number of stimulations sufficient to produce a lifelong enhanced sensitivity to stimulus evoked seizures without associated spontaneous seizures or overt neuronal loss. Here we asked whether stimulations sufficient to induce enhanced sensitivity to evoked seizures, but not frank epilepsy will induce the formation of basal dendrites and other morphological changes similar to those observed in models of epilepsy associated with substantial cell loss. The morphology of GFP-expressing granule cells from Thy-1 GFP mice was examined either one day or one month after the last of five amygdala kindling-evoked seizures. Interestingly, significant reductions in dendritic spine density were evident one day after the last seizure, the magnitude of which had diminished by one month. Further, the there was an increase in the thickness of the granule cell layer, a day after kindling, which was absent a month later. No differences in the number of basal dendrites were detected at either time point. The time course and direction of spine density changes support the idea that this plasticity represents a homeostatic response within some limbic circuits aimed at reducing excitatory synaptic function. These findings demonstrate that the early stages of kindling epileptogenesis does not produce the striking rearrangements of granule cell structure seen in other models of epilepsy.
Just a sample of my work. To see more or discuss possible work >>