Japanese researchers are hoping to predict epileptic seizures by incorporating new heart rate variability (HRV) methods to the development of a wearable, daily use device.
Testing of the heart rate method on patients with drug-resistant epilepsy has so far revealed that an algorithm created by the scientists has a high sensitivity for predicting incoming seizures; and a low rate of false alarms.
The research paper,“Epileptic Seizure Prediction Based on Multivariate Statistical Process Control of Heart Rate Variability Features,” was published in IEEE Transactions on Biomedical Engineering.
Previous research had proposed that heart rate variability (HRV) monitoring, the RR interval (RRI) fluctuation in an electrocardiogram (ECG), can predict an incoming epileptic seizure, as the excessive neuronal activity that occurs before a seizure also affects the autonomic nervous system and, in turn, autonomic nervous function affects HRV.
Prevention and warning of uncontrolled seizures can significantly improve the quality of life of epileptic patients, especially the approximately 30% that have not seizure control even with appropriate medication.
In order to surpass the current constraints in the development of an RRI measurement device in daily life, the researcher proposed a new HRV-based epileptic seizure prediction method. It consists of two parts: HRV feature extraction from RRI data of epileptic patients, and epileptic seizure prediction by using an anomaly monitoring technique.
The scientists applied the proposed new method to clinical data collected from 14 epileptic patients. The data indicated that eight patients had 11 awakening preictal (immediately before seizure) episodes. The length of interictal (between seizure) episodes was about 57 hours..
The new method was able to predict 10 of the 11 preictal episodes, which translates into a sensitivity of 91%. False positives, which are predicted seizures that do not actually happen, only occurred once per hour on average.
researchers concluded: “This study proposed a new HRV-based epileptic seizure prediction method, and the possibility of realizing an HRV-based epileptic seizure prediction system was shown”, and that the method could be used in daily life because heart rate can be measured easily with a wearable sensor.
The scientists are currently developing a mobile seizure prediction system via a smartphone and a wearable RRI sensor. The wearable RRI sensor will measure RRI data and send the information immediatly to the smartphone in order to warn the patients of potential “interictal” and “preictal” occurances in real-time.
