Scientists use AI to spot 'invisible' brain abnormalities

Researchers in London have developed an Artificial Intelligence tool that detects hidden brain abnormalities. It was trained on hundreds of patient scans and it can now spot two-thirds of these tiny and invisible irregularities.

Scientists in London have developed an Artificial Intelligence tool called a MELD Graph to detect tiny brain abnormalities that can cause epilepsy—often too difficult for standard MRI scans to identify.

The team has spent 10 years developing this process. The device was trained on MRI data from more than 700 people with focal cortical dysplasia, which causes epilepsy. Known as FCDs, these can be hard to spot with the human eye and half of these lesions are missed by radiologists.

As well as finding these irregularities, it explains the cause.

With this analysis, radiologists say they can quickly diagnose and be able to provide surgery that could potentially cure the seizures.

Epilepsy affects about 1 in 100 people globally, with 1 in 5 cases linked to brain structural abnormalities.

Although the tool isn't clinically available yet, the team has released the software as open-source, and is training clinicians and researchers on its use.
 

Copied