Main content

AI can predict brain cancer survival rates

A neurologist from King鈥檚 College London says the tool 鈥渆mpowers clinicians and patients to consider the choices that they might have鈥.

Artificial Intelligence (AI) can predict whether adult patients with brain cancer will survive more than eight months after receiving treatment.

Glioblastoma is a fast-growing type of brain tumour that is difficult to treat with just one in four patients surviving more than one year after diagnosis.

Researchers from King's College London say they have created a deep learning model to allow them to predict outcomes more reliably for patients following treatment with radiotherapy and chemotherapy.

Dr Thomas Booth is a Neurology Consultant at King鈥檚 College. He told Newsday: 鈥淲e鈥檝e looked at tens and thousands of images (of non-cancer brains) and there are components within the images, which are teased out by the AI, which then when we add cancer images to it, the AI is picking out various鈥eatures, which can differentiate between poor survival and good survival.鈥

He said the resulting information 鈥渆mpowers clinicians and patients to consider the choices that they might have.鈥

(Picture: Shows a doctor鈥檚 hand pointing out the location of a glioblastoma brain tumour on a computer screen. Credit: Getty Images.)

Release date:

Duration:

4 minutes