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

Artificial Intelligence beats doctors at predicting heart disease deaths

A new artificial intelligence (AI) developed model can predict risk of death in patients with heart disease better than models created by medical experts, according to a new study from the Francis Crick Institute.

The study was published in PLOS One, and it is the result of a collaboration between scientists at the Crick, the Farr Institute of Health Informatics Research and University College London Hospitals NHS Foundation Trust.

Coronary artery disease is the number one cause of death in the UK. The researchers behind the study developed the AI model in an attempt to create a method of preventing heart disease-related deaths which would be more efficient than medical experts models using self-taught machine learning techniques.

A prognostic model built by medical experts was compared to the AI one created by the Crick team. The first one made predictions based on 27 variables chosen by doctors, such as age, gender and chest pains. By contrast, the AI model’s algorithms trained themselves, searching for patterns among the electronic health data of over 80,000 patients and picking the most relevant variables from a set of 600.

“Doctors already use computer-based tools to work out whether a patient is at risk of heart disease, and machine-learning will allow more accurate models to be developed for a wider range of conditions.

Andrew Steele

Results showed the data-driven model outperformed the expert-designed model at predicting patient mortality, but it also identified new variables that doctors hadn’t considered.

“Along with factors like age and whether or not a patient smoked, our models pulled out a home visit from their GP as a good predictor of patient mortality,” said Crick scientist Andrew Steele, first author of the paper.

“Home visits are not something a cardiologist might say is important in the biology of heart disease, but perhaps a good indication that the patient is too unwell to make it to the doctor themselves, and a useful variable to help the model make accurate predictions.”

Although still in a proof-of-concept stage, this study is another example of how AI is slowly changing Healthcare in the UK and globally, and similar models could be soon implemented in real-world medical scenarios.

“It won’t be long before doctors are routinely using these sorts of tools in the clinic to make better diagnoses and prognoses,” Steele said.

“Doctors already use computer-based tools to work out whether a patient is at risk of heart disease, and machine-learning will allow more accurate models to be developed for a wider range of conditions.”

SHARE THIS ARTICLE:
BY SHACK15

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

Subscribe to our newsletter