IISc Researchers Develop Algorithm To Identify Occurrence And Type Of Epilepsy By Decoding Brain Scans

Based on the electrical patterns in different analyses, the algorithm was able to develop whether a human subject could have epilepsy or not, with a high degree of accuracy. 

Researchers at the Indian Institute of Science (IISc) have developed an algorithm that can interpret electroencephalograms (EEGs), a recording of brain activity, and detect seizures. The algorithm can identify the occurrence and type of epilepsy by decoding brain scans. ( Image Source :Getty )

Researchers at the Indian Institute of Science (IISc), in collaboration with AIIMS (All India Institute of Medical Sciences) Rishikesh, have developed an algorithm that can interpret electroencephalograms (EEGs), a recording of brain activity, and detect seizures. The algorithm can identify the occurrence and type of epilepsy by decoding brain scans. The study describing the findings was recently published in the journal Biomedical Signal Processing and Control.

What is epilepsy?

Epilepsy is a central nervous system disorder in which the brain activity becomes abnormal, resulting in seizures or periods of unusual behaviour, sensations and sometimes loss of awareness, and maybe the result of a genetic disorder or an acquired brain injury. It is a neurological disease in which the brain emits sudden bursts of electrical signals in a short amount of time. 

Types of seizures

Based on the region of the brain in which an abnormal electrical activity begins, seizures can be classified as focal seizures and generalised seizures. Absent seizures are a type of generalised seizures. 

When the abnormal electrical activity causing a seizure begins in both hemispheres of the brain at the same time, a generalised seizure occurs. Meanwhile, focal seizures begin in one area of the brain, but become generalised and spread to other areas, according to Johns Hopkins Medicine. 

An absence seizure causes one to blank out and stare into space for a few seconds. 

Neurologists need to manually inspect EEGs in order to identify whether a patient is epileptic. EEGs can capture erratic signals in the brain. However, visual inspection of EEG can become tiring after prolonged periods. This may occasionally lead to errors, Hardik J Pandya, corresponding author on the new paper, said in an IISc statement. 

Pandya explained that the research aims to differentiate EEGs of normal subjects from epileptic EEGs, and the developed algorithm aims to identify the types of seizures. He added that the work will help neurologists make an efficient and quick automated screening and diagnosis. 

What does the algorithm do?

The researchers developed a novel algorithm that can examine EEG data and identify signatures of epilepsy from the electrical signal patterns. Based on the electrical patterns in different analyses, the algorithm was able to develop whether a human subject could have epilepsy or not, with a high degree of accuracy. 

How were tests conducted on human subjects?

As part of the study, the researchers examined EEG data from 88 human subjects acquired at AIIMS Rishikesh. Each subject was made to undergo a 45-minute EEG test. The test was divided into two parts. 

Initially, a 10-minute test was conducted wherein the subject was awake. It included photic stimulation and hyperventilation. Photic stimulation is a technique used to investigate abnormal brain activity triggered by visual stimuli such as flashing lights or patterns. Hyperventilation is an activation technique used during EEG recording to elicit abnormalities that may imply epilepsy.

The 10-minute test was followed by a 35-minute sleep period. After this, the team analysed the data and classified different wave patterns into sharp signals, spikes, and slow waves. 

Different types of wave patterns in EEGs

The patterns in which a signal rises and falls within a very short duration of time (about 70 milliseconds) are called spikes. The patterns with rises and falls spread over a slightly longer duration (about 250 milliseconds) are called sharps. Meanwhile, slow waves spread over about 400 milliseconds, a much longer duration. 

What do different EEG patterns indicate?

According to the study, a normal EEG pattern indicates a normal subject, partial discharges or wave patterns indicate focal seizures, widespread discharges indicate generalised seizures, and spike wave discharges indicate absence seizures. 

How does the algorithm help determine the occurrence and type of epilepsy?

In comparison to a healthy individual, an epileptic subject shows a different set of patterns. The total number of sharp waves seen in the EEG is called the Cumulative Sharp Count. The researchers developed an algorithm to calculate the Cumulative Sharp Count, and use it as a parameter to detect if the subject is epileptic or not.

According to the study, a higher value of Cumulative Sharp Count indicates a greater chance that the subject is epileptic. The algorithm also calculates the areas under the spikes and sharp curves, enabling one to distinguish between focal and generalised epilepsy. A greater value of the sum indicates generalised epilepsy, while a lower value indicates focal epilepsy.

According to the researchers, the study shows a way to identify absence seizures, that are characterised by a brief, sudden lapse of consciousness, using a Cumulative Spike-Wave Count. In some cases, absence seizures can be critical and fatal. 

The researchers checked the accuracy of the algorithm by running it on a new set of EEG data from subjects for whom the type of epilepsy was already known to the doctors. The blind validation study successfully classified the subjects accurately in nearly 91 per cent of the cases, the statement said.

Physicians at AIIMS Rishikesh are testing the reliability of the algorithm.