AI algorithm can detect brain defects and help treat epilepsy disorders

Scientists have now developed a new AI algorithm that can detect early symptoms in patients with epilepsy, including even those that experienced senior physicians may miss.

The AI algorithm was able to detect patients with a rare disease with an accuracy of over 60%, compared to MRI which did not detect any hidden symptoms. The rare condition is epilepsy, which is said to occur in 1 per cent of the UK and US population, and causes uncontrolled bursts of brain currents in the brain, resulting in spasms and convulsions.

Anyone can have convulsive spasms, but that doesn't mean they all have epilepsy, and often people with epilepsy may have more than one convulsive seizure before being diagnosed. Epilepsy occurs when there is a sudden burst of brain current, causing disruption to the functioning of the brain. In some patients, people remain alert and are able to perceive their surroundings when they have a seizure, while in others they lose consciousness completely and are in an extremely dangerous situation when they have a seizure.

Sometimes people with epilepsy experience unusual sensations, perceptions and activities, or they may fall stiffly to the floor and keep convulsing. If medication does not work, the patient may have to opt for brain surgery.

One of the causes of epilepsy is drug-resistant focal cortical dysplasia (FCD), a subtle abnormality of the brain that causes a failure of signal transmission in the brain. Epilepsy can be treated surgically, but the changes the disease produces in the brain are so subtle that even experienced radiologists performing MRI scans may miss the signs of the disease.

But an AI algorithm based on a team of researchers at University College London can detect 63% of the symptoms of FCD, something that has previously been difficult for healthcare professionals to detect and is one of the key triggers of seizures.

The researchers claim that their AI model could provide the best chance of a cure for more epilepsy patients undergoing brain surgery. It is reported that in the UK, there are around 600,000 people with epilepsy, but only 20-30% of them do not respond to medication.

Brain cells, or neuronal cells, usually form organised layers of cells that form the cerebral cortex. In people with FCD, the brain cells are disorganised and disordered, leading to a higher risk of spastic convulsions. FCD symptoms are the most common cause of epilepsy in children undergoing surgery to control the condition, and the third most common cause in the adult population requiring surgical treatment. In the latest study, researchers collected more than 1,000 MRI scans from 22 global studies of epilepsy, and a team of radiologists flagged the scans as healthy or FCD, and then ran an AI algorithm to detect the abnormalities.