Last Updated: 22nd January, 2019
Yes, it’s true. Artificial Intelligence can be used as a death factoring calculator. Case studies have proved that AI can predict metastatic cancer and deaths caused by heart attack. Let’s deep dive how? Google is creating an Artificial Intelligence (AI) system that can be used to predict a patient’s death—Simply astonishing.
Google’s Medical Brain team is using a new type of Artificial Intelligence algorithm that can make predictions about the likelihood of death among hospital patients. The company claims that the AI can predict when a patient might die more accurately than other traditional models used by doctors.
Artificial Intelligence (AI) systems are clearly taking the world by storm. AI technology is revolutionizing the world and is prevalent in almost every facet of our lives:
- The way we purchase items.
- The way we communicate.
- The way we move across the world.
These days, AI is pushing its boundaries further and entering into the deadly territories: Death Prediction.
What is Artificial Intelligence?
Artificial Intelligence is a branch of Computer Science which aims to make machines simulate Human Intelligence Tasks/Processes. These systems usually make predictions using a technique referred to as Machine Learning.
How does Google AI make Predictions?
The system works by chewing up data about various details of a patient such as age, gender, ethnicity, previous diagnosis, present signs, and laboratory results.
What’s more, the system can also take data buried on charts and PDFs and use it for making predictions.
After testing the algorithm, Google found out that it could make death predictions with an astonishing accuracy level of 95%, which is 10% more accurate than the traditional models.
In one of the case studies, the Google AI software crunched about 176,000 data points of the records of a woman with metastatic cancer and pronounced that she had a 19.9% chance of dying in the hospital. The doctors at the hospital had given her a 9.3% chance of dying.
Amazingly, as predicted by the AI software, the woman died after two weeks—beating the doctors at their own game.
Although the Google’s AI may not be perfect at predicting deaths, it would continue to perform better as it is ‘taught’ with more data.
Therefore, if this breakthrough healthcare technology is improved, it can reduce human error in medicine and lead to more lives being saved.