Data and visualization help to personalize epilepsy treatment

A doctor at Boston Children’s Hospital checks on a young epilepsy patient during treatment. Photo credit: Katherine C. Cohen, Boston Children’s Hospital

A doctor at Boston Children’s Hospital checks on a young epilepsy patient during
treatment. Photo credit: Katherine C. Cohen, Boston Children’s Hospital

CHAPEL HILL, NC – For patients with epilepsy and their doctors, determining the best treatment plan often involves playing “medical detective.”

Non-routine visits to the doctor often take place after the patient has endured a seizure, and patient and doctor must piece together what happened just before the seizure, its length and severity, and possible triggers in an effort to determine whether treatments or medications need to be changed. If a clinician wants to compare a patient’s latest seizure to his or her medical history or to historic data in Electronic Medical Records (EMRs), they often must wade through reams of paper to see relationships and find correlations that could lead to better treatments.

In the age of mobile apps for everything, their world has yet to see an app that integrates EMRs and individual medical histories, but a project led by RENCI’s Ketan Mane, PhD, aims to change that.

The project unites RENCI’s expertise in data infrastructure and information visualization with epilepsy experts at the UNC Chapel Hill School of Medicine, Duke University School of Medicine and Boston Children’s Hospital at Harvard University.  The goal is to put patient medical data and related information at the clinician’s fingertips in visual formats that make it easy to spot trends and find correlations between variables, such as seizures and medications, or medications and side effects.

“Clinicians usually learn about a new epileptic seizure afterwards, when the patient comes in for an appointment,” said Mane, a senior research scientist and medical informatics specialist. “We want the clinician to be able to get this information soon after the event and before the office visit. We want the clinician to be able to view the data outside the confines of a silo by merging it with the patient’s medical history to build overlays and make sense of what is going in a way that is easy to absorb and includes context.”

To achieve their goals, the research team will explore using both mobile and Web technologies to allow patients or their family members to record details about a seizure immediately after it happens. The infrastructure will include a mobile app for documenting seizure data on a smart phone, a streamlined system for collecting that data and delivering it to the patient’s caregiver, and web and visualization technologies that integrate patient data with EMR data, and then display that data on a visual dashboard that enables clinicians to view a patient’s profile, overlay it with other data and spot trends.  RENCI plans to build a prototype system using a sample of data from collaborators at Boston Children’s Hospital, UNC Hospitals, and Duke Medical Center.

When fully functional, the project could illustrate how digital medical records and personal medical data can help clinicians make treatment decisions based on records of what has worked well in similar cases and the histories of each patient. It could also takes some of the time-consuming detective work out of medical care by making it quick and easy to see relationships between symptoms, side effects, medications, activities and seizures. As a result, clinicians will be able to adjust treatment plans based on a patient’s overall medical history and relevant medical evidence.

“EMRs mean there is a wealth of medical data available that can be used to help us improve diagnoses and develop treatment plans that are targeted to individuals and their medical histories,” said Tobias Loddenkemper, MD, a co-PI on the project, neurologist at Boston Children’s Hospital and associate professor of neurology at Harvard Medical School. “The problem is that data alone can’t help a busy clinician, who struggles just to keep up with the daily patient load. Our research aims to put this data into the hands of clinicians in ways that will be useful and will have positive impacts on patients.”

In addition to Loddenkemper, partners in the project are Mohammed Mikati, MD, chief of pediatric neurology at the Duke University School of Medicine; Michael Tennison, MD, professor of child neurology in the UNC School of Medicine, and Alan Leviton, MD, Boston Children’s Hospital.

For more information, see