EPINOV: A Computer Model of Epileptic Brain Undergoes Clinical Trial for Transforming Surgery

Clinical trials have a familiar pattern. Groups of volunteers are administered a new drug or vaccine to test their efficacy in preventing a disease or killing a germ. Along with testing how harmful or safe they are for the body. And with the search for effective vaccines, the pandemic brought clinical trials closer to the public discourse. However, clinical trials are not restricted to drugs or vaccines alone; now, computer models of diseased parts of the body are being used for clinical trials for surgeries.

Imagine a situation where scientists prepare a computer model using high-level computing techniques representing a diseased condition of a part of the body. This model will tell about the details of the disease and with some computer modulations doctors can decide where exactly a surgery should be done. Scientists have developed such a model of epileptic brains and their model is undergoing clinical trials. In the trials, it is being evaluated how far the model correctly depicts the physiological conditions and how accurately it can predict the doctors to carry on a surgery.

In the clinical trial named EPINOV, scientists of the Institute De Neuroscience Des Systems (INS) in France are looking at how far a computer model of the epileptic brain can improve the rate of epilepsy surgeries. They have reported on the approach and the clinical trial in their research paper published in the journal Science.

The models are created using the Virtual Epileptic Patient (VEP), which is a computational system developed by scientists. This project is a part of the Human Brain Project (HBP), a 10-year research project started in 2013 with an aim of allowing researchers across Europe to advance their understanding of neuroscience and brain diseases.

In an article by Miryam Naddaf published at Nature Briefing, UCL (University College London) neurosurgeon Aswin Chari, commenting on VEP was quoted as saying, “It’s an example of personalized medicine. VEP uses “the patient’s own brain scans [and] the patient’s own brainwave-recording data to build a model and improve our understanding of where their seizures are coming from.”

Nearly 50 million people that suffer from epilepsy across the globe, have abnormal brain activities and one-third of them do not respond to drugs applied for stopping epileptic seizures. According to Chari, surgeries are huge reliefs for these patients. In such surgeries, a portion of the brain known as the epileptogenic zone, which initiates the seizures, are removed. It is still a challenge for neurosurgeons to find out the exact portion of the epileptogenic zone. Remember, brain surgeries always come with challenges, especially in finding the exact location a surgeon will make the incision. A cut in an unwanted portion can lead to severe conditions like paralysis.

In epileptic surgeries, clinicians take advantage of several techniques, especially to find the epileptogenic zone, such as MRI (Magnetic Resonance Imaging), EEG (Electroencephalogram) etc. Surgeons also use the SEEG (Stereoelectroencephalography) technique. Here, seven centimetres-long 16 electrodes are placed over the skull and the electrodes record the brain waves (depicting brain activities) of specific areas for a week or two.

It is worth mentioning here that the hallmark of brain functioning is weak electrical pulses or signals. Neurons in the brain produces those signals and transmit amongst one another. These electrical pulses have particular patterns and in diseased conditions, these changes. The techniques such as EEG or SEEG etc. detect the brain pulses and find how the patterns are changed and then analyse from which portion the abnormal signals are coming and decide about surgery. This is in a nutshell, the real problem, especially analysing the signals detected by the techniques is more complicated.

But the techniques have some limitations. For example, the SEEG can detect only high frequency electric signals and the lower frequencies go undetected, that account for 20% of the seizure cases, as written by Miryam Naddaf in Nature Briefing. These limitations make surgeries challenging, especially finding the epileptogenic zones and restrict the current success rate of surgery to only 50%. “The failure of the surgery is often attributed to a misidentification of the epileptogenic zone,” said Viktor Jirsa, the corresponding author of the research paper published in Science and also the director of the Institute De Neuroscience Des Systems.

The EPINOV trial began in June 2019 and across 11 French hospitals, the trial has so far enrolled 356 people. They evaluated the predictive power of the computer model VEP and observed participants for a year after they underwent surgery with the help of the model. Enrolment for EPINOV is going to be over by this year and data analysis will start from late of 2024, as the one year observation finishes by then.

The researchers however evaluated the VEP model with 53 patients having drug-resistant epilepsy. The authors in their research paper in Science said that VEP showed a 60% precision in identifying the epileptogenic zones. The VEP predicted zones showed very small distance from the clinically defined epileptogenic zones.

No clinical trial is a short-duration exercise, and here also we will have to wait for some time to have the final results. However, if the trial results turn out to be promising then this computer model-based surgeries will pave a new path in the realm of personalised medicine, where individuals are treated precisely according to their physiological specificities.