Download the all-new Republic app:

Published 09:05 IST, November 13th 2019

Artificial intelligence is better than doctors at predicting abnormalities: Reports

Artificial intelligence was used to analyze 1.77 million ECG results from nearly 400,000 people & predict patients who had a higher chance to die within a year

Reported by: Digital Desk
Follow: Google News Icon
  • share
null | Image: self
Advertisement

Artificial intelligence can accurately predict how likely someone is to die within a year by looking at heart test results - even when doctors believe the person is healthy. In a mystery diagnosis, the machines are able to detect abnormalities in electrocardiogram (ECG) results and predict who is at higher risk of dying. Researchers recently performed a study in which Brandon Fornwalt and his team at Pennsylvania's Geisinger healthcare provider took historical data of electrocardiogram (ECG) results and sort them into groups based on which patients had died within a year.

The experiment

Artificial intelligence was used to analyze 1.77 million ECG results from nearly 400,000 people to develop its ability to predict patients who had a higher chance to die within the next year. An ECG is a simple test that can be used to check a patient's heart rhythm and electrical activity as a graph of voltage over time. The raw data from this was used for training the AI, along with the patient's age and sex.

Advertisement

READ | AI Tool Can Now Help Farmers Protect Their Crops From Climate Change

The researchers then used a measure called the area under the curve (AUC) as a method to compare how well the algorithm can perform its task. A score can range from 0 to 1, where 1 is where predictions are 100% correct, and 0.5 implies that it can't distinguish between the patients living more than a year and those who die within a year.

Advertisement

The AI was able to consistently predict the correct result with a score of 0.85, which is impressive on its own, however, comparing it to the success rate of trained doctors highlights the extent to this AI can prove useful in the future. The AUC scoring models currently used by doctors range between 0.65 and 0.8, says Fornwalt. Also, the Geisinger team also made another algorithm to spot certain ECG patterns that doctors use when predicting the risk of death.

READ | First 'AI' University Launched In Abu Dhabi, Could India Be Next?

Advertisement

What is an ECG?

An ECG, an electrocardiogram, traces the electrical activity of the heart and can help Doctors to detect abnormal rhythms that may occur due to heart attack or heart conditions. 

READ | Now, AI Can Predict Where Lightning Will Strike 30 Minutes In Advance

Advertisement

READ | This AI System May Help Save Children Or Pets Left Alone In Vehicles

01:40 IST, November 13th 2019