Algorithm learns to diagnose genetic disease

By training an algorithm using patient data, researchers at the University of Gothenburg, along with colleagues from Portugal, have now found a way of estimating the probability that a patient is suffering from a common genetic disease.

”Depending on how well you train the algorithm, the test would be more or less precise,” Saga Helgadottir explains.

Familial hypercholesterolemia (FH) is a common genetic disease affecting one in 250 people. It is a metabolic disease that causes raised levels of cholesterol. A genetic test is required for a definitive diagnosis. However, these tests are only available at a limited number of university hospitals. All other hospitals use a method called the Dutch Lipid Clinic Network Score (DLCNS), which is based on the patient’s clinical features and family history.

A group of researchers have now invented an alternative method which has been shown to produce better results than the DLCNS.
“We decided to try and use machine learning to predict whether a patient has the hereditary genetic disease or whether their symptoms have arisen as a result of their lifestyle,” says Saga Helgadottir, a doctoral student at the Department of Physics.

Link to full article