Researchers have come up with the three facts needed to speed up diagnosis and treatment of people with chest pain due to coronary artery disease.
In a paper published in JMIR Cardio, a team of researchers from Computing and Life Sciences at the OU and the Cardiovascular Research Unit at Milton Keynes University Hospital (MKUH), describe how they worked together on the dataset supplied by the hospital to decide whether a patient should be prioritised.
Professor Attila Kardos, Head of the Cardiovascular Research Unit at MKUH, supplied a unique dataset from 17 years of stress echocardiograms, a test used to detect blocked arteries in the heart. Patients are often sent for a stress echocardiogram if they go to their GP with chest pain.
This dataset allowed lead author, OU machine learning specialist Dr Mohamed Bennasar, to find that he could predict the test result with a very high accuracy knowing only three facts about the patient: prior cardiac history, gender and prescribed medication.
Professor Kardos said “This work has the potential to improve patient care by prioritising patients who have a very high probability of a positive result, which means heart disease can be detected faster and appropriate treatment such as medications or corrective surgery, i.e. stent placement or bypass surgery, can be proposed sooner.”
The paper’s second author, OU physiologist Dr Duncan Banks, noted: “This kind of multidisciplinary collaboration between computing and life sciences is just the tip of the iceberg, we expect to see many more fascinating results in the near future.”