Research from Japan shows that heart rate variability (HRV – the variation in the time interval between heartbeats) could potentially be used to predict epileptic seizures.Heart rate and rhythm can be easily measured using a wearable sensor, and evidence shows that they are affected by seizure activity. If the ‘predictive value’ of this is confirmed, we will be a step closer to the development of a device that can alert people when a seizure is about to take place. This would have clear life-changing, and even life-saving, benefits for people with recurrent epileptic seizures, and it is something that researchers having long strived for.During the current study, the team used eight features of HRV, combined with a recognised technique called ‘anomaly monitoring’, to develop an algorithm for predicting seizures. They then tested the method on simultaneous 24-72-hour video-EEG (looking for physical and electrical signs of seizures) and ECG (examining heart activity) recordings from 14 people with drug-resistant epilepsy. Having the two sets of data allowed the researchers to check how effective the algorithm was at identifying when a seizure was oncoming (at the time of assessment) based on the person’s ECG recordings (and HRV characteristics) in the lead-up to it.The data showed that eight of the 14 subjects had a total of 11 ‘preictal’ episodes (brain activity seen shortly before a seizure) when they were waking from sleep. From the ECG (and hence the HRV) data obtained at these times, the algorithm successfully predicted 10 of the 11 seizures that followed. Also important is the frequency with which the model predicted seizures when they were not about to happen (false positives), as this can be extremely disruptive. It was found to be less than once per hour on average.These findings are very encouraging and the authors are now developing a mobile seizure prediction system based on a smartphone and wearable heart beat interval sensor. In this system, the sensor will measure the person’s heart beat intervals and ‘send them’ to the smartphone wirelessly. The smartphone app will analyse the data and assess, in real time, whether the person is between seizures (interictal) or about to have a seizure (preictal). If the latter is true, an alert will be raised.The authors write, “The developing system has the potential for improving QoL (quality of life) of epileptic patients through real-time seizure warning.”This study is published in the scientific journal IEEE Transactions on Biomedical Engineering.Click here to read more stories about living with epilepsy.