Through the use of machine learning and AI, scientists at Google along with its subsidiary Verily, which specialises in health-tech, have devised a new way to assess the risk of heart disease in patients.
The company has created software that utilises a new algorithm to infer a person's age, blood pressure and if they are a smoker by analysing scans of the back of their eye. This data is then used to predict the likelihood of the risk of the patient suffering from a heart attack or other major cardiac event with almost the same accuracy as the current methods employed by doctors today.
Since Google's new method does not require a blood test, doctors will be able to determine an individual's cardiovascular risk faster and more efficiently. However, it still needs to be tested thoroughly before being implemented in a real world setting.
A recently published paper in the Nature journal Biomedical Engineering provided further insight into the technology behind the software and how it functions. Instead of replacing healthcare professionals, this new use of AI and machine learning can help improve the diagnostic tools used by doctors worldwide.
The scientists at Google and Verily analysed a medical dataset containing the eye scans and general medical data of almost 300,000 patients using machine learning to train the algorithm. The company's algorithm was successfully able to determine which patients had suffered from a cardiovascular event during the past five years 70 per cent of the time. This puts the algorithm just behind the current SCORE method that utilise a blood test to correctly predict a patient's cardiovascular health 72 per cent of the time.
Although Google and Verily's new algorithm is still a ways off from being used in hospitals and doctor's offices, it shows the possibility that AI and machine learning will augment current jobs as opposed to replacing them.
Image Credit: Everything Possible / Shutterstock