Seizures Can Also be Predicted a Few Days in Advance Like a Weather Forecast!

Epilepsy is a chronic disease with repeated episodes of convulsions. The short electrical activity of the brain can cause convulsions, hallucinations or loss of consciousness. Seizures are divided into partial/focal seizures, generalized seizures, and unclassified seizures. If the seizure is not treated in time, it may cause great harm to the patient.
Recently, a new analysis of clinically approved brain implant data by neuroscientists from the University of California, San Francisco, the University of Berne, and the University of Geneva shows that patterns of brain activity can be used to predict the risk of seizures in patients with epilepsy several days in advance. The research was published recently in the journal The Lancet.

For more than four decades, research on predicting seizures has focused on developing early warning systems, which at best can only warn patients a few seconds or minutes before the seizure.
This research result is the first time that epileptic seizures can be reliably predicted several days in advance, which can greatly facilitate patients to avoid the possible risks of seizures. Epilepsy can cause convulsions, hallucinations, or loss of consciousness. For decades, researchers around the world have been trying to identify patterns of electrical activity in the brain that indicate impending onset, but the results have been limited.
The researchers used an implantable brain stimulation device that can quickly stop seizures by precisely stimulating the patient’s brain when the patient has the first signs of an imminent seizure. This equipment, called the Neurospace RNS System, also allows the research team to study the epilepsy-related brain activity recorded by patients during months or even years of normal life.
By analyzing these data, it is found that seizures are not as random as they seem. In the latest study, researchers began to test whether these laws can be used to establish a clinically reliable seizure risk prediction mechanism.
The researchers built a statistical model to match the recorded brain activity patterns with the subsequent seizures of 18 epileptic patients who had implanted neurospace devices. Subsequently, the team tested these prediction algorithms using data from 157 participants who participated in the multi-center long-term treatment trial of the RNS system between 2004 and 2018.
Reviewing the trial data, the researchers were able to determine a period of time when the patient’s likelihood of seizures was nearly 10 times higher than baseline, and in some patients, signs of these high-risk periods could be detected several days in advance. Of course, an increased risk of seizures does not necessarily mean seizures will occur.

Although many people report factors that promote seizures, such as stress, alcohol, wrong dosage of drugs, or lack of sleep, neurologists still cannot be completely sure what caused the epilepsy to occur at a specific time. Like the weather forecast, this prediction cannot be accurate at a few minutes and a few minutes. However, based on brain activity, patients can know the chance of seizures. If the risk is greater, then they should avoid activities that may trigger seizures.
In addition, accurate prediction of seizure risk also allows neurologists to adjust the patient’s drug dose accordingly. Most of the time, keep the dose low to minimize side effects, and only increase the dose when the risk of seizures is high. The researchers found that there are significant differences in the degree to which the brain activity of study participants can predict future seizure risk.
The risk of 40% of RNS system trial participants can be predicted several days in advance, the brain data of other participants can only predict the risk of the next day, and some participants have not shown a reliable activity cycle at all. In response to this difference, the researchers said that more work is needed.