Team 7
Real-Time EEG Analysis Software Add-on for Monitoring of Status Epilepticus
Team Members:
Real-Time EEG Analysis Software Add-on for Monitoring of Status Epilepticus
Team Members:
Aubrey Berger
Lauren Everett
Nyah Kshatriya
Margarito Hernandez Fuentes
Team Mentors:
Dr. Leon Iasemidis, PhD - Barrow Neurological Institute
Dr. Noah Hutson, PhD - Barrow Neurological Institute
YouTube Link:
View the video link below before joining the zoom meeting
Zoom Link:
https://asu.zoom.us/j/83380473869
Abstract
Status Epilepticus (SE) is a form of seizure that prolongs longer than 5 minutes or consists of concurrent seizures within a 5 minute period. In the United States, there are approximately 150,000 cases of SE annually, with an incidence rate of 6.2 in 100,000 people. Patients with SE face a grim mortality rate of 22%. Over-medicating SE patients can lead to adverse effects, but delays in treatment can lead to neuronal damage. Physicians require a better approach to efficiently detect and evaluate the effectiveness of medication on SE. Our device aims to provide valuable information to physicians to better inform medication responses.
Our device consists of four components: (1) real-time architecture, (2) a cleaning algorithm, (3) a feature classification algorithm, and (4) a user interface. The device imports patient EEG data in real-time and uses preprocessing techniques to remove artifacts. Multiple features are extracted, such as theta band amplitude and connectivity of various parts of the brain. These features are classified via a support vector machine. The classification is displayed to physicians on the user interface. Physicians can add annotations and view the raw EEG data. To optimize our product, we performed multiple designs of experiments and iterated accordingly. Ultimately, we fulfilled a majority of our proposed target specifications, including an accuracy of 97.6%, real-time processing of 1.25 seconds, and a file size of approximately 2.6MB. We aim to market our device as a Class II medical device as it informs medication decisions. We will submit a 510(k) premarket notification and pursue classification as Software as a Medical Device (SaMD) through the FDA. We will market our device to hospitals for integration with current EEG systems and license the software to EEG manufacturing companies as an add-on to their products.
