Detection of epileptogenic brains with non-linear analysis of electromagnetic signals

Background

Epilepsy, one of the most prevalent chronic neurological diseases, affects approximately 0.6-1% of the general population and represents an economic burden of .1-1% of the total budget of health care systems in some parts of the world, with a significant proportion of these costs arising from the way epilepsy is diagnosed. Presently, an epilepsy diagnosis relies on a patient's clinical history and is based on identifying surrogate markers for unprovoked seizures (epileptogenicity) such as the presence of repeated seizures. However, seizures can occur in any brain, including in apparently normal brains, if the appropriate stimulus is applied. Because of their unpredictable nature, seizures are often difficult to observe while a patient is in the clinic, and this inability to observe seizures promptly often leads to a delay in the diagnosis. Clinical events that are suggestive of seizures are confirmed with a concomitant electroencephalogram (EEG) recording that demonstrates their epileptic nature. However, as epilepsy is the predisposition to spontaneously develop seizures, this predisposition cannot currently be diagnosed by direct methods.

Technology Overview

Boston Children’s Hospital researchers recognized that conventional methods might be improved by identifying biomarkers for epilepsy by analyzing electric or electromagnetic signals from patients prone to seizures without the need to identify the occurrence of seizures. This approach involves:

• Receiving EEG data recorded from a patient;
• Applying a multiscale algorithm to the received EEG data to produce a scaled EEG;
• Determining nonlinear feature values, such as entropy and recurrence rates, from this data;
• Diagnosing the patient as having epilepsy, based, at least in part, on the nonlinear feature values.

It includes methods, computer-readable media, and computer systems for the analysis of EEG data for the detection of reliable biomarkers for epilepsy. By identifying specific nonlinear biomarkers, it provides an accurate diagnosis of epilepsy as well as other cognitive disorders with similar EEG patterns.

Applications

• Prompt diagnosis of epilepsy without the need to wait for seizures;
• Guidance of tailored treatment based on specific nonlinear features;
• Identification of other cognitive disorders with similar EEG patterns.

Advantages

• This invention can increase diagnostic accuracy through the identification of specific nonlinear biomarkers;
• It would allow remote diagnosis and monitoring of disease—EEG data may be sent directly from a wireless EEG headset to a smartphone or cell phone and relayed directly to a remote computing device;
• It can provide additional data to advance the understanding and treatment of epilepsy.

Publications

• Sathyanarayana, A., El Atrache, R., Jackson, M., Alter, A. S., Mandl, K. D., Loddenkemper, T., & Bosl, W. J. (2020). Nonlinear Analysis of Visually Normal EEGs to Differentiate Benign Childhood Epilepsy with Centrotemporal Spikes (BECTS). Scientific reports, 10(1), 8419. https://doi.org/10.1038/s41598-020-65112-y