A research funded by the British Heart Foundation (BHF) has found that an algorithm developed using artificial intelligence (AI) can help diagnose heart attacks in women more quickly and accurately.
The University of Edinburgh researchers have combined the data obtained from 10,038 people, of which 48% are women, who visited hospital with suspected heart attack.
This data was utilised for developing an AI-based tool, known as CoDE-ACS, to help healthcare providers diagnose heart attacks more accurately.
Later, the researchers have validated the tool on 3,035 people, of which 31% were women, outside of the UK.
Using AI, the tool combines the usually collected patient information, such as sex, age, observations, ECG findings, and medical history along with the troponin blood test results, when they arrive at hospital.
Then, it produces a score of zero to 100.
At present, the protein troponin in the blood is measured as the gold standard for diagnosing heart attack, but its levels released by the heart differ between men and women, along with age as well as other health conditions.
The researchers found that the CoDE-ACS tool has the potential to exclude a heart attack with 99.5% accuracy, which confirms that the people are safe.
It has also identified the people who need to stay at hospital for further tests, with 83.7% accuracy.
The researchers noted that the tool’s performance was consistent regardless of age, sex, and pre-existing health conditions.
BHF associate medical director James Leiper said: “This is a huge step forward which promises to ensure everyone is on a level playing field when it comes to heart attack diagnosis and treatment.
“We know that women are more likely to receive a misdiagnosis, but by harnessing the power of AI, this team could provide one solution that helps to make that an issue of the past.”
BHF Centre for Cardiovascular Science data scientist Dimitrios Doudesis said that the new algorithm is now embedded into an easy-to-use mobile app for supporting healthcare providers in guiding treatment decisions.
Doudesis added: “Whilst the troponin test takes 30 minutes to process, we take an array of other health information and add it into the app alongside the troponin measurement.
“This provides doctors with a precise and instantaneous score to confirm if they can reassure their patient that they haven’t had a heart attack and send them home, or if they require further tests.”