Epileptic encephalopathies (EE) are severe forms of childhood epilepsies. Recent studies have shown that some cases of pediatric EE are caused by copy number variants (CNV), which are chromosomal deletions, duplications, or insertions of DNA. CNVs may contain genes that are important for neural health, causing disease when there is a copy number change. I am investigating rare CNVs in EE patients where the genetic cause is not known. I will screen EE patients for CNVs using a technique called array comparative genomic hybridization (aCGH) to identify rare CNVs that may be disease-causing. If a CNV is found, I will conduct aCGH on parents to determine if the CNV is de novo (arising for the first time in that individual and not inherited from either parent). We also expect that in cases with multiple affected family members, patients may inherit rare mutations from an affected parent. Preliminary results show that out of 69 patients screened thus far, 15 rare CNVs that range in size and number of genes affected have been identified in 13 patients. Of these, 3 de novo CNVs were identified that include a single exon deletion of the gene SLC2A1, a partial deletion of the gene NRXN1, and a 2.4 mb deletion in chromosomal region 7q11.23, affecting 40 genes. In addition, 2 patients each have a 1.4-1.5 mb deletion encompassing the region 16p13.11, which has been reported to be associated with epilepsy and developmental delay. Parental and segregation testing is ongoing for many patients. Rare CNVs can highlight genes implicated for proper neural function and development. This will help advance our understanding of the pathology and genetic basis of epilepsy. Furthermore, identification of particular pathogenic genes will be useful in the clinical setting for genetic counseling of affected patients and families.

Cognitive impairment in the form of dementia is a hallmark of Alzheimer’s disease (AD), an aging-related neurodegenerative disorder that is characterized by increasing deficits in memory and cognition. To allow for targeted, early intervention, it is necessary to develop reliable prognostic tools that can identify individuals at risk of developing AD. Studies suggest that several neuroimaging markers may be predictive of cognitive decline; these include atrophy of medial temporal lobe structures, reduced cerebral blood flow to the hippocampus, reduced functional connectivity between memory-related structures, and lower density of cholinergic cells within the nucleus basalis. In this study, we modeled the relationship between episodic memory test scores from Wechsler’s Adult Intelligence Scale using both established and novel neuroimaging markers while controlling for age, sex and education. Standard neuropsychological battery test scores and structural, diffusion and resting-state functional MRI data were collected from 54 participants (n = 29 women) aged 65 and above (mean = 78.6 years) recruited from Group Health Research Institute’s Adult Changes in Thought study. We found left hippocampal and left entorhinal cortex (LEC) volumes to be significantly smaller in cognitively impaired subjects as compared to healthy controls (p <0.05), and episodic memory scores were most positively correlated with LEC volumes (r = 0.45, p < 0.05) as compared to other MRI measures. Stepwise regression analysis also showed that LEC volume explained a significant portion of variance in memory scores (R² = 0.31, F = 6.88, p <0.05). In comparison to structural MRI markers, functional connectivity measures did not add predictive value, suggesting that physiological measures play a more mediating role in the relationship between cortical structure and performance. Overall, volumetric MRI measures served as better predictors of cognitive performance than other examined MRI measures, and the inclusion of these measures in future AD risk assessments may help to predict cognitive decline with greater sensitivity.

Although efficacious treatment options are available for children who stutter, currently existing treatments for adults who stutter are generally ineffective and time-consuming with relapse rates as high as 50-70%. Thus, there is an urgent need for innovative stuttering treatment approaches that are both effective and efficient. Here, we explore the use of neurofeedback training as a novel tool in reducing stuttering frequency. Neurofeedback is a type of biofeedback that enables individuals to self-regulate particular aspects of brain activity (e.g., slow cortical potentials or activity in specific frequency bands) through real-time visual feedback. Neurofeedback protocols targeting various activation patterns have already been tested extensively for the clinical management of disorders such as epilepsy, ADHD, and Tourette’s syndrome. In the present preliminary study, we examined whether or not the direct regulation of a brain wave called the mu rhythm (8-13 Hz) shows promise for reducing the primary symptoms of stuttering. Three adults who stutter completed 12 sessions of neurofeedback training over a 6-week period. Brain activity was recorded with an electroencephalography (EEG) system and, after on-line filtering, these data were used to provide visual feedback about activity in the mu frequency band. Participants played a video game in which the goal was to change the color of a balloon by increasing or decreasing their mu rhythm activity. Reading and conversational speech samples were audio- and video-recorded before and after each training session for the offline calculation of stuttering frequency. Data collection has been completed and analysis of each participant’s stuttering frequency within and across sessions is ongoing. Results will provide initial data regarding the potential use of EEG-based neurofeedback as a new clinical tool in the treatment of stuttering.

Auditory-motor adaptation is a form of sensorimotor learning in which human subjects gradually (i.e., over several trials) adjust the movements of their speech articulators when experiencing experimentally manipulated auditory feedback. For example, it is well documented that subjects lower their formant frequencies (i.e., acoustic output resonant frequencies that are determined by the articulators’ positions) in response to a real-time upward shift of these formants in the perceived auditory feedback. Our long-term goal of translating such auditory-motor learning protocols into clinical applications for individuals with speech disorders requires an understanding of the central nervous system’s tolerance for feedback delays. Indeed, one potential problem for efficient sensorimotor learning arises when movement-related feedback is delayed: perceiving the consequences of one’s own actions with a delay results in this sensory information being processed similar to externally-generated, rather than self-generated, sensory input. In the present study, we follow up on recent work in which we found that auditory-motor learning in speech is completely abolished with feedback delays of 100 ms or more. Here, we tested whether shorter delays also have a negative impact on speech auditory-motor learning. Twelve adult subjects read out loud monosyllabic words while a digital vocal processor applied a 2.5 semitones upward shift to all formants in the auditory feedback signal that was presented through insert earphones. In separate conditions, this feedback was delayed by 0, 33, 66, or 100 ms. Measurements of the subjects’ extent of auditory-motor learning across all conditions will be presented and interpreted in the context of current models of sensorimotor learning.
 

Previous work in my lab has shown that extracranial applications of high intensity focal ultrasound can affect the activity of specific regions in the brain, even going as far as producing specific movements in the extremities of rodents if applied to the motor cortex. With the recent discovery that activation using light stimulation of neurons whose conductive coating (myelin) has been damaged results in higher-than-normal remyelination processes, we proposed that the activation of similarly damaged neurons using high frequency focused ultrasound will also result in increased remyelination and restoration of the targeted neurons. Such a treatment could be used for various neurodegenerative diseases, including multiple sclerosis. Since multiple sclerosis is a disease which damages the myelin coating of neurons and impairs their function, this treatment could revolutionazie multiple sclerosis therapy. Earlier this year, I began work on developing such a treatment using a mouse model which induces brain lesions similar to those found in human multiple sclerosis patients. These experiments include therapies using ultrasound transducers at three different center frequencies, and effectiveness of the treatment is assessed by EEG recordings of the mouse brains during application of ultrasound, as well as MRI scans and stained brain slices of treated and control mice. This work will run through May, and by then I will be able to determine whether high frequency focused ultrasound therapy can be effective in either slowing down demyelination or speeding up remyelination in a well-known animal model of multiple sclerosis. So far we have collected about half of our mouse data, and the post-therapy MRI scans look promising but brain slice staining is currently underway, which will provide much more concrete data on the extent of remyelination.