CAN AI PREDICT DEATH RISKS?
Artificial Intelligence is getting smarter and creepier day by day, and it can even predict when a person will die. Yes, a new machine-learning algorithm developed using echocardiogram (ECG) videos of the heart can accurately predict patients who will die within a year by looking at heart test results — even when they look normal to doctors. And AI algorithms are proving to be an effective solution in predicting death.
But we have no idea how it works. How it does is so is a mystery.
The algorithm — an example of what is known as machine learning, or artificial intelligence — outperformed other clinically used predictors, including pooled cohort equations and the Seattle heart failure score.
For their study, the research team used specialized computational hardware to train the machine learning model on 812,278 ECG videos collected from 34,362 Geisinger patients over the last ten years. The study compared the results of the model to the cardiologist’s predictions based on multiple surveys. A subsequent survey showed that the cardiologist’s prediction accuracy improved by 13 percent when assisted by the model. Leveraging nearly 50 million images, this study represents one of the largest medical image datasets ever published.
The AI accurately predicted the risk of death even in people deemed by cardiologists to have a normal ECG. Three cardiologists who separately reviewed normal-looking ECGs weren’t able to pick up the risk patterns that the AI detected.
The neural network model that directly analyzed the ECG signals was found to be superior for predicting one year risk of death. “This is the most important finding of this study. This could completely alter the way we interpret ECGs in the future.” Says Brandon Fornwalt, Chairman of the Department of Imaging Science & Innovation at Geisinger in Danville, Pennsylvania.
Another study by the same group of researchers found that AI-based models can analyze ECG test results and pinpoint patients at higher risk of developing a potentially dangerous irregular heartbeat (arrhythmia) and the risk of early death due to chronic diseases in a largely middle-aged population. The team used more than 2 million ECG results from more than 3 decades of archived medical records in Pennsylvania/ New Jersey’s Geisinger Health System to train deep neural networks. They found that AI can examine ECG that results to predict irregular heartbeat and the death risk.
“Incorporating these models into routine ECG analysis would be simple. However, developing appropriate care plans for patients based on computer predictions would be a bigger challenge”, said lead author Sushravya Raghunath.
Both studies are among the first to use AI to predict future events from an ECG rather than to detect current health problems.
More and more industries today are being influenced by AI and MI, and healthcare is no exception. It’s one of the industries that could benefit the most from AI, which will undoubtedly transform the future of healthcare. Thanks to smart algorithms, doctors would accurately predict patients outcomes and even the risk of premature death. Such valuable information would help healthcare professionals to better personalize treatments and improve care delivery before it’s too late.
