Australian researchers announced on Tuesday that an artificial intelligence tool is able to monitor accurate brain distortions that are difficult to discover in children with epilepsy, which may help patients to undergo a faster surgery that may change their lives.
According to experts, epilepsy have multiple reasons, and about 3 out of 10 cases are attributed to structural abnormalities in the brain. But MRI often neglects these deformities, especially the smallest pests that may be hidden in the depths of the fold of the brain.
A team led by a pediatric neurologist at the Royal Melbourne Hospital, Emma McDonald Lorses, worked to train an artificial intelligence tool on children’s brain images to detect pests the size of blue berries, or even what is smaller than them.
McDonald Lawers noticed – at a press conference before publishing the study in the magazine “Eiblipsea” – that “these pests are often unable, and many children are not considered in need of surgery.”
She explained that “this tool is not indispensable for radiologists or those who specialize in epilepsy. Rather, it is more like an investigator that helps to collect parts of the disassembled image more quickly, allowing the suggestion of surgery that may change the life of” the patient.
Among the patients participating in the study who suffer from peel and focal epilepsy, 80% had previously undergone a magnetic resonance examination whose results were normal.
When the researchers used the artificial intelligence tool to analyze both magnetic resonance tests and another type of medical examinations, the oxidon tomography, its success rate was 94% for one of the two test groups and 91% for the other group.
Among the 17 children in the first group, 12 surgery underwent the removal of the brain lesions, and 11 seizures, according to the McDonald Lors team at the Murdoch Institute for Children’s Research.
She pointed out that “the next step is to test this detection tool in a hospital environment that is more realistic on new patients whose injury has not been diagnosed.”
It affects epilepsy that causes repeated seizures about one in 200 children, and is effective of medicines in about a third of patients.
“This study is interesting” and its “extremely important” results.
In a similar research published in February conducted by a team of “Kings College London”, an artificial intelligence analyzes of MRI data reached 64% of the epilepsy brain lesions that were not monitored by radiologists.
Konrad Wagstille pointed out that Australian researchers used MRI and azuretone tomography, but the latter is “costly, and is not widely available such as MRI, and has a accompanying radiation dose, such as computerized tomography or X -rays.”