ARTIFICIAL INTELLIGENCE Can Diagnose Heart Attacks Faster and More Accurately Than Human Doctors, Study Finds
Research funded by the British Heart Foundation found that an algorithm excludes accurately Heart attack in 99.6 percent of suspected patients.
It will reduce the rate of misdiagnosis and get life-saving treatments to the right patients faster.
Doctors in Scotland are already testing the system in hospitals.
Professor Nicholas Mills from the University of Edinburgh said: “Early diagnosis and treatment saves lives.
“Unfortunately, Many medical conditions cause chest pain and the diagnosis is not always easy.
“This has tremendous potential to improve patient care and efficiency in our busy emergency departments.”
In the UK there are around 100,000 hospital admissions for heart attacks each year.
Human health professionals assess whether they believe someone has such a condition by testing levels of a blood protein called troponin, which spikes during a seizure.
However, results vary and a dangerous result for one patient may be safe for another.
Artificial intelligence takes into account a patient’s troponin levels, age, gender, heartbeat data and medical history – and compares all this data to a huge database of previous patients.
It then advises the doctor on the likelihood of a heart attack by comparing the patient’s results to those of previous patients.
The entire analysis can be performed much faster by a machine than by a human.
Professor Sir Nilesh Samani, Medical Director of the British Heart Foundation, said: “This has the potential to rule out or rule out a heart attack more accurately than current approaches.”
“It could be transformative and reduce the time it takes to get a diagnosis, which is much better for patients.”
The results of the experiments with the CoDE-ACS system were published in the journal Nature Medicine.