AI may aid early detection and treatment of epilepsy, Australian study finds

Coloured positron emission tomography (PET) sagittal scans of the brain of a 22 year old male patient with temporal lobe epilepsy, showing abnormal metabolic activity in the right temporal pole. Temporal lobe epilepsy (TLE) is the most common type of focal epilepsy and is characterised by seizures that originate in the temporal lobes. These areas are responsible for memory, emotions, and language. Temporal lobe seizures often impact these functions, and individuals may experience unusual sensations during a seizure, such as joy, deja vu, or fear. The cause of TLE is often unknown but may be linked to brain injury, brain defects, or tumors. Anti-epileptic medications are commonly used to manage seizures. The brain has been highlighted using the radioactive tracer fluorodeoxyglucose (18F-FDG).No use: BBC, Pearson, Oxford University Press, Springer Nature, Dorling Kindersley, Cambridge University Press & Assessment, New Scientist, Bonnier Publications, John Wiley & Sons, The Lancet, BMJ

Source: CFL

Australian researchers have developed an artificial intelligence tool that can detect tiny brain malformations in children with severe epilepsy, potentially fast-tracking life-changing surgery.

The detector, created at the Murdoch Children’s Research Institute and the Royal Children’s Hospital in Melbourne, identified lesions the size of a blueberry in up to 94% of cases, compared with an 80% miss rate in standard MRI scans.

In a recent study published in Epilepsia, 17 children were tested with the tool; 12 underwent surgery to remove lesions and 11 are now seizure-free. Experts say the advance could transform treatment for the roughly one in 200 children living with epilepsy.

How it works

The tool was trained on MRI and PET scans of 71 children and 23 adults with cortical dysplasia and focal epilepsy.
Cortical dysplasias develop in the womb and are a common cause of drug-resistant seizures, yet are often missed on MRIs because lesions can be as small as a blueberry and hidden deep in brain folds.
By combining MRI and PET imaging, the detector reached success rates of 94% in one test group and 91% in another.
Of the 17 children in the test group, 12 had surgery and 11 are now seizure-free.
Lead researcher Dr. Emma Macdonald-Laurs said the AI works “like a detective,” helping doctors piece together scans more quickly without replacing radiologists or epilepsy specialists.

Why it matters

Epilepsy affects about one in 200 children; around a third of cases do not respond to drugs, making surgery the best hope for a cure.
Early identification of lesions enables faster referrals for surgery, fewer seizures and improved developmental outcomes.
Frequent seizures disrupt learning, behaviour and mood, with long-term risk of intellectual disability if untreated.
Traditional MRIs miss more than half of cortical dysplasias on first scan, delaying diagnosis and treatment.
AI can help neurosurgeons plan safer operations by pinpointing abnormal tissue while avoiding healthy brain regions controlling speech, movement and thought.
Researchers plan to expand testing of the tool in real hospital settings, though PET scans remain expensive, less available than MRIs, and involve a radiation dose similar to CT scans.

This story is written and edited by the Global South World team, you can contact us here.

Logan Zapanta