Tracking Epilepsy

 Sridevi Sarma, an assistant professor in the Department of Biomedical Engineering, was educated as an electrical engineer, and now she’s using that training to help doctors understand how the brain’s electrical activity changes during an epileptic seizure.

Working with William S. Anderson and Nathan Crone, of the Johns Hopkins Department of Neurology, Sarma is refining a software product, called EZTrack, that analyzes a brain’s electrical impulses to pinpoint the location of the epileptogenic zone (EZ)—the region with abnormal electrical signals that causes seizures. This is important because a small percentage of people with epilepsy have such severe seizures that surgery is required to remove portions of the brain.

Today, surgeons rely on teams of experts who pore over a patient’s EEG data (which often takes weeks to collect) to determine the location in errant signals. An expert in mathematical models and algorithms, Sarma is using graph theory to discover how the brain’s electrical activity changes during a seizure. This lets her show doctors a map of a brain’s electrical misfiring in batch mode. She does this by analyzing data transmitted by electrodes placed directly on the brain. “This approach to automated localization is a paradigm shift,” says Gregory Kent Bergey, co-director of the Johns Hopkins Epilepsy Research Laboratory.

Sarma was awarded a $100,000 Coulter Foundation grant to build a prototype of the software product. To date, she has tested it in 19 clinical patients, with 100 percent accuracy in predicting surgical outcome. She plans to expand the clinical testing.

Modeling a Hurricane’s Wrath

As a hurricane spins toward the coast, electric power utility companies and major retailers scramble to prepare for its effects. How many crew members and what equipment should be on the ground, ready to help restore power? What will be the demand for bottled water, generators, and batteries?

Using publicly available data, Geography and Environmental Engineering’s Seth Guikema, together with collaborator Steven Quiring of Texas A&M University, has developed a new model to predict power outages four to six days in advance of a hurricane’s making landfall, by combining census data with factors such as three-second wind gust speeds. A paper on this research was just accepted for publication in IEEE Access, and Guikema now is seeking to commercialize the technology.

“Our results can impact response time and efficiency and provide cost savings,” Guikema says. “There’s interest in our work from Gulf Coast and East Coast power companies, big retailers, and companies that provide analytical services to power electric power utilities.”

Self-Assembling Tumor Fighters

Safely and effectively delivering anti-cancer drugs directly to tumor sites remains one of the greatest challenges in battling that disease. Currently, many researchers are focusing on encapsulating chemotherapeutic agents within nanosized carriers—made of synthetic materials—that help transport them to the tumor. But those small vehicles carry very small drug portions, and also present potential toxicity and side effects.

Honggang Cui, an assistant professor of chemical and biomolecular engineering and affiliated with the Institute for NanoBioTechnology, has an alternative approach: anti-cancer drugs that assemble themselves into nanostructures that may be more effective against tumors. “These structures have a high drug-loading capacity and they minimize potential toxicity,” Cui says. “Furthermore, we have shown that we can tailor the molecular design according to how much of the drug is needed.”

Cui and his team are now in the process of testing their product for commercialization.