Babylon study finds new AI with ‘imagination’ is more effective than GPs at medical diagnosis

Researchers used causal machine learning approach to increase algorithm’s accuracy.
By Tammy Lovell
06:30 am
Share

Credit: Babylon

Research by UK-based digital firm Babylon found that artificial intelligence (AI) using the principles of causal reasoning scored higher than 70% of GPs in diagnosing written test cases.

Up until now, the firm’s AI symptom checkers have provided advice using an algorithm which relies on correlations with patient symptoms and medical history.

But a peer-reviewed study, published in Nature Communications, concluded that disentangling correlation from causation made the AI more accurate.

Researchers used a new approach, known as causal machine learning, to act like an ‘imagination’ so the AI could consider what symptoms it might see if the patient had a different illness to the one it was considering.

Dr Jonathan Richens, Babylon scientist and lead author, said: “We took an AI with a powerful algorithm, and gave it the ability to imagine alternate realities and consider ‘would this symptom be present if it was a different disease’? This allows the AI to tease apart the potential causes of a patient’s illness and score more highly than over 70% of the doctors on these written test cases.”

In the study, a group of 44 GPs were tasked with diagnosing 1,671 written test cases, including typical and atypical examples of symptoms for more than 350 illnesses.

Babylon’s AI took the same tests and used both an older algorithm based on correlations and the newer, causal one.

Each GP assessed an average of 159 cases and achieved a mean score of 71%. Babylon’s older AI achieved 72%, whilst the new AI achieved 77%.

This new algorithm is not yet present in Babylon’s publicly available app and will only be released once it has met all necessary regulatory approvals.

WHY IT MATTERS

This technology could pave the way for a future partnership between clinicians and AI that would speed up doctors’ diagnosis, improve accuracy, free up time for clinicians and improve patient outcomes.

THE LARGER CONTEXT

Babylon recently announced plans to launch in the US, following a $550 million Series C round, reported to put the company at a valuation of more than $2 billion.

Another player in the symptom checker space is Polish firm Infermedica, which announced a $10.2million Series A round for its call centre triage solution earlier this month.

Meanwhile, Finnish startup, Klinik Healthcare Solutions, recently expanded its online triage service to the UK, with 60,000 patients now able to access its AI symptom checker.

ON THE RECORD

Dr Ali Parsa, CEO and founder, Babylon, said: “AI will be an important tool to help us all end the injustice in the uneven distribution of healthcare, and to make it more accessible and affordable for every person on Earth.”

Share