Data Science Investment Counter —Funding raised by UK data science companies in 2018.
£ 5.640 Million

How AI Could Help Find a Cure for Alzheimer’s

From Google DeepMind’s efforts to detect eye disease and kidney injury, to IBM’s and Microsoft’s moonshot pledges to end cancer, machine learning and Artificial Intelligence are increasingly being deployed to fight pernicious or outright untreatable diseases.

Now a London AI company is on a quest to defeat Alzheimer’s, a neurodegenerative illness that affects 1.3 percent of the UK population and is among the prime causes of dementia.


Avalon AI is using machine learning algorithms to detect early signs of Alzheimer’s disease in brain scans.

Avalon AI—a startup supported by Techstars and EF accelerators— is using machine learning algorithms to detect early signs of Alzheimer’s disease in brain scans.

“We’re streamlining the work of researchers and doctors,” explains Avalon AI’s cofounder Olivier van den Biggelaar.  “They send us brain scans of possible patients, we analyse them with machine learning algorithms looking at some key features associated with the condition, and we send back a report where those features are highlighted.

Avalon AI’s technology has been trained on the Alzheimer’s Disease Neuroimaging Initiative’s brain scan bank, and its unsupervised diagnostic tools assess the health of brain parts like the hippocampus, an area that atrophies in Alzheimer’s patients.

Compared to human doctors, van den Biggelaar says, Avalon AI’s algorithms can pick up telltale signs of the disease’s onset that are discernible in the scans but frequently go unnoticed by practitioners.

In this way, the technology could solve one of the issues making Alzheimer’s particularly insidious: patients often remain undiagnosed until they reach an advanced stage and their symptoms become evident.


By detecting early Alzheimer’s patients, Avalon AI will find suitable subjects for clinical trials of new therapies.

That is not only bad for patients, but also poses a challenge to researchers looking to cure Alzheimer’s for good: experimental new treatments, in fact, need to be tested on early stage patients whose brain tissue has not yet been savaged by the disease. But as things stand, pharmaceutical companies have a lot of troubles finding subjects for their trials.

Hence the second prong of Avalon AI’s business, besides the diagnostic service.

“We’ll match the patients we diagnose with clinical trials, as they’ll likely be in the stage of the disease you’re looking for if you are running a trial,” van den Biggelaar says. “ In this way, we’ll make subject recruitment much easier and much cheaper.”

Over time, as Avalon AI processes more and more scans, its technology will grow more accurate at delivering early Alzheimer’s diagnoses.

“Our deep learning predictive models will help identify patients very early , and early detection will eventually beget a treatment, “ says van den Biggelaar . “The overarching goal of our company is actually to find a cure for Alzheimer’s disease.”

Follow Avalon AI on Twitter


Co-working space and blog dedicated to all things data science.

Subscribe to our newsletter