Machine learning helps predicting the risk of cardiovascular disease for diabetes patients

Machine learning helps predicting the risk of cardiovascular disease for diabetes patients

People suffering from type 2 diabetes are much more likely to develop a cardiovascular disease. This is, of course, very sad and health professionals cannot predict which patients will develop heart conditions and which ones will avoid them. Now scientists from the University of Sydney used machine learning to develop a model, which can help predict the risk of cardiovascular disease for people living with type 2 diabetes.

Type 2 diabetes significantly degrades the quality of life for millions of people. Image credit: Omstaal via Wikimedia (CC BY-SA 4.0)

Type 2 diabetes is a progressive disease – it becomes worse and worse over time. It is estimated that nearly half a billion people in the world live with type 2 diabetes. And since this condition affects pretty much the entire body, they are more likely to develop cardiovascular disease.

Prevention is the best treatment, but it is very tricky to predict which diabetes patients are facing that increased risk. Scientists now gathered data from private hospitals in Australia, paying attention to heart conditions and their relation with type 2 diabetes. Researchers found that some factors increased the risk of cardiovascular disease for people with type 2 diabetes.

Dr Shahadat Uddin, lead researcher in this study, said: “Our study found that the prevalence of renal failure, fluid and electrolyte disorders, hypertension and obesity was significantly higher in patients with both cardiovascular disease and type 2 diabetes than patients with only type 2 diabetes”.

Scientists use machine learning algorithms to analyse a huge pool of data and produce reliable results. Scientists found that their model achieves accuracy of 79-88 % – for medical predictions that’s very high. Researchers say that a universal coding practice would be useful for doctors working with type 2 diabetes patients. It would help researchers to analyse more data quicker, improving the model and having even more reliable, accurate predictions.

This new prediction model could be used for research as well as in a clinical setting. Scientists believe that it could also be useful for government and private health insurers. Probably the most important application of this comorbidity risk prediction model is going to be the development of health management programs for patients at high risk of developing multiple chronic diseases.

Type 2 diabetes is a lifestyle disease. Although there are exceptions, it develops because people are abusing unhealthy food, beverages, alcohol. Most type 2 diabetes patients are overweight. Hopefully, studies like this will help predicting which patients are more likely to develop cardiovascular conditions and need to take some prevention measures.

 

Source: University of Sydney


Source link