A new Approach to Rapid Diagnosis of Epilepsy Found at IKBFU

Epilepsy is a dangerous neurological disease that, according to WHO, affects up to 50 million people worldwide. Its main symptoms are so-called seizures, in which there is a pathological synchronization of neurons of the brain. As explained by experts, seizures often occur without outward signs, so many people may not suspect that they are sick.

To diagnose and treat this disease in clinical practice, electroencephalography (EEG) is used to record the electrical activity of the brain. Existing methods involve "manual" processing of EEG results by a doctor, which can take up to six hours for one patient.

Scientists at Immanuel Kant Baltic Federal University joined efforts with specialists of the Pirogov National Medical and Surgical Center, have developed a program capable of independently searching the EEG data array for signs of abnormal electrical activity characteristic of a seizure state

The new program, as the scientists explained, is based on machine learning without a «teacher» —in contrast to neural network systems, it independently identifies patterns and hidden relationships in unlabeled data without reliance on predetermined templates. This approach to diagnosing epilepsy, according to the authors, is the first in the world.

In the future, the research team intends to adapt the proposed program for combined use with the method of photostimulation of the patient's brain, which allows to significantly reduce the time required for a full examination using EEG.