Scientists have created a neural network that accurately identifies heart failure

An international team of scientists has developed a new neural network that defines heart failure. She works accurately and is able to diagnose this problem by the result of an electrocardiogram.

Such indicators seem unbelievable, but the studies conducted confirm these statements - the accuracy of the neural network is 100%. To create it, scientists from the UK and Italy used previous developments in this direction, but they created their product from scratch. This allowed them to avoid the problems of previous developments and not to think about the fact that faultlessness is impossible to achieve. To do this, they used a neural network with a hierarchical structure that allows it to well recognize repeating patterns and structure existing data. For her training, scientists used the public data of electrocardiograms of different people.

A new development can determine the presence of progressive heart failure in just one heart beat. Also, the novelty has the ability to establish the degree of development of the disease. This will help doctors to make an accurate diagnosis in advance, and also to understand when a progressive pathology has gone into a more serious form. Recall, in total, 26 million people live in the world with such a disease. Timely diagnosis can save many lives.

dni24.com