Research lab of Dr. Beth A. Lopour
University of California, Irvine, Biomedical Engineering
We are generally interested in the study of the brain as a complex dynamical system, including signal processing of neural data and simulation via computational models. We are especially interested in projects with clinical applications that have the potential to improve diagnosis or treatment of a neurological condition.
The current focus of the lab is on projects related to epilepsy, due to unique circumstances that allow us to record data directly from living human brains. When medication is inadequate for the control of seizures, patients will sometimes opt to have surgery to remove the portion of the brain that is generating the seizure activity. In order to identify a target for the surgery, physicians use various types of data, including electrophyisiological measurements. These electrode measurements may come from the surface of the skull (EEG), the surface of the brain (ECoG), or they may be implanted directly into various brain structures (LFP, single unit). In addition to guiding clinical decisions, these measurements provide a unique and exciting opportunity for research:
High frequency oscillations as a biomarker of epileptic tissue
High frequency oscillations (HFOs) were first identified as a marker of epileptic tissue in rodent models of epilepsy, and they are now being extensively studied in humans. These events have a peak frequency > 80 Hz and are characterized as four or more oscillations that stand out from the background activity. It has been shown that removal of the brain regions with high frequency oscillations correlates to a seizure-free outcome after epilepsy surgery; however, the identification of these “ripples” of activity is not a trivial task. They can be seen both inside and outside of the seizure focus, and they can be physiological as well as pathological. We are characterizing these high frequency oscillations using human electrocorticogram (ECoG) data and studying their relationship to the network of brain regions that is generating the seizures.
Specific areas of interest:
- Novel algorithms for automated detection of HFOs
- Techniques for optimization of HFO measurement and detection
- Identification of new computational high frequency biomarkers of epileptic tissue that are more accurate and easier to implement in a clinical setting
This project is being done in collaboration with Dr. Jack Lin at the UC Irvine Medical Center.
EEG markers of infantile spasms
Infantile spasms are a rare form of epilepsy that occurs in children under 2 years of age. Because this is a such a critical period of development for the infant’s brain, early diagnosis and treatment is crucial. In diagnosing infantile spasms, physicians look for a pattern in the EEG called hypsarrhythmia; while there is a classic presentation of this rhythm, many variations are possible, which makes diagnosis difficult in some cases. We are developing computational markers of EEG to (1) Help us learn about the underlying neuronal networks that contribute to this disease, and (2) Be an unbiased tool for the diagnosis and treatment of infantile spasms, as a supplement to standard visual EEG analysis.
Specific areas of interest:
- Computational markers of EEG for the diagnosis of infantile spasms, assessment of treatment response, and prediction of treatment response and relapse
- Functional connectivity networks in healthy infants and those with infantile spasms
- Temporal dynamics of computational EEG markers during the sleep/wake cycle, during the course of treatment, and during normal development.
This project is being done in collaboration with Dr. Shaun Hussain at UCLA and Dr. Daniel Shrey at CHOC.