Artificial Intelligence Is Now Able to Diagnose Children

Modern technologies are developing at a speed of light. The good thing is that aside from new smartphones with three cameras and transparent OLED televisions, the mankind is also working on the improvement of one of the most important spheres of human life – healthcare.

Artificial intelligence is capable of doing things that seemed a mystery just a couple of years ago. Many medical and technological experts believe that if AI is merged with medicine, it could create a revolution in the way illnesses are diagnosed and treated.

How can AI help out the littlest human beings who are seriously ill?

Just like the law is based on precedents, healthcare can be the same. As reported by Daily Mail, this concept was developed by the Chinese scientists who created a deep learning system that consists of almost one and a half billion “precedents” (medical files, test results, health records, patients’ history etc.)

Currently, the system is based only on children’s cases, since child diseases are an urgent and pressing problem in the healthcare sector.

Basically, the system works in the following way: the AI program analyzes the symptoms and then selects the suitable diagnosis out of a huge database of child illnesses. The scientists state that the system is 90% accurate.

The history of creation

The scientists, doctors and AI developers from Guangzhou, China took data on patients less than 18 from the major medical establishment in the city Guangzhou Women and Children’s Medical Centre for the time period between the January of 2016 and July of 2017.

The total amount of health records amounted to around 1.36 pieces, and after the program was fully “trained”, it contained over 100 million items of data.

Despite the skepticism of some experts who claim that no real doctor can be replaced by a machine, the program proved its efficiency and, hopefully, will save many lives in the future.

Subscribe to The Coin Shark news in Facebook: https://www.facebook.com/coinshark/