A new AI can predict if you’ll die soon based on your electrocardiogram. But, scientists don’t understand what patters the algorithm is picking up.

Thanks to recent researches, we can now use artificial intelligence for various purposes – even predicting the future.

We have an AI technique that can forecast volcanic eruption, and an algorithm that can write believable fake news within seconds. There’s also an Artificial Intelligence system that can predict the future.

But, an AI model that can predict the likelihood of Dying within a year may be the most interesting of all. Yes, that exists now.

Brandon Fornwalt and his colleagues trained AI to examine ECG data and predict who has a higher risk of dying in the following year.

So, how does it work?

New AI can Predict the Likelihood of Death

An ECG records the electrical activity of the heart. So, patients with cardiac conditions such as heart attacks and atrial fibrillation show a noticeable pattern change.

Fornwalt and his colleagues trained two versions of the AI.

In the first algorithm, they used raw ECG data, which measures the voltage over time. Meanwhile, they fed the second algorithm ECG data along with the patient age and sex.

Then, they tasked the AI models with examining 1.77 million ECG results from almost 400,000 people. Both AI versions had to predict who was at a higher risk of dying within a year.

To measure the AI’s performance, the researchers used a metric know as AUC. It provides a picture of how well distinguishes between the patients who died and those that survived.

AI Can Catch What Humans Miss

Where a perfect score is 1 and 0.5 indicates no distinction between the two groups, the AI consistently scored above 0.85. Current risk scoring models used by doctors range between 0.65 and 0.8, says Fornwalt.

In other words, the AI accurately predicted the risk of death, even in people whose ECG appeared normal. What’s more, three cardiologists who separately reviewed seemingly ordinary ECG data missed the risk patterns that the AI picked up.

Fornwalt noted:

That finding suggests that the model is seeing things that humans probably can’t see, or at least that we ignore and think are normal. AI can potentially teach us things that we’ve been maybe misinterpreting for decades.”

Scientists are still not sure how the AI is picking up the data. As a result, some physicians have reservations about using the algorithm.

Read More: Researchers Discover New Fast Radio Bursts Using Artificial Intelligence

The post New AI can Predict Likelihood of a Patient Dying Within a Year appeared first on EDGY_ Labs.