AI can predict brain cancer survival rate in patients suffering from ‘most lethal form’ of the disease, scientists reveal

• Scientist trained system on a dataset of 10,000 brain cancer scans
• AI can predict if a patient will survive eight months following treatment 
• Read more: AI develops cancer treatment in just 30 days 

Artificial intelligence can predict the survival rate of patient’s suffering from the deadliest type of cancer, scientists have revealed.

Researchers at King’s College London trained a system to determine if individuals diagnosed with glioblastoma would live for at least eight months following radiotherapy, allowing doctors to explore other treatments if patients would not.

The average glioblastoma survival time is 12 to 18 months – only 25 percent of patients survive more than one year and just five  percent of patients survive more than five years.

Patients typically need regular scans before clinicians can see drug effectiveness, but the new system showed it could provide the same results by analyzing a single MRI scan.

The AI was trained with a dataset of 10,000 MRI scans of different types of cancers.

Alysha Chelliah, who was involved in the study, said: ‘We applied deep learning to predict whether glioblastoma patients will survive the first eight months after completing radiotherapy.

‘This approach is intended to improve the ability to identify patients who require early second-line treatment or clinical trial enrolment, compared to those showing initial treatment response.’

The National Foundation for Cancer Research — which funds research into cancers — says glioblastomas are the ‘most lethal form of brain cancer’.

It develops when cells supporting nerves in the brain begin to divide uncontrollably.

The fast-growing cells invade nearby brain tissue, making them hard to remove, but generally do not spread to other areas of the body.

About one in 30,000 people have the condition, estimates have suggested.

Dr Thomas Booth, who was involved in the research, said: ‘This study was motivated by a clinically-attuned and critical research question regarding aggressive brain tumors, and delivered by leveraging cutting edge artificial intelligence. 

‘Whilst less common than other cancers, the devastation is disproportionate with a two-year survival rate of 18 percent.’

The team setout to find a better method for predicting short- and long-term survival rates, which would eliminate the need for multiple scans that can delay treatments.

‘Instead of trying to grapple with interpreting each and every non-specific follow-up brain scan, we simply looked at one routine scan after radiotherapy and gave an accurate prediction using artificial intelligence to answer a simple question: which patients will not survive the next eight months,’ said Booth.

‘The AI was able to give us an immediate and accurate prediction which means clinicians can empower patients to make choices about their treatment.’