Artificial intelligence algorithms are everywhere in healthcare. They sift through patients’ records to predict who will develop medical conditions like heart disease or diabetes, they help doctors figure out which people in the emergency room are the sickest, and they screen medical images to find evidence of disease. But even as AI algorithms become more important to medicine, they are often invisible to people receiving care.
Artificial intelligence tools are complicated computer programs that suck in huge amounts of data, look for patterns or trajectories, and make a prediction or recommendation to help make a decision. Sometimes the way algorithms process all the information they receive is a black box – unfathomable even to the people who designed the program. But even if a program is not a black box, the math can be so complex that it’s difficult for anyone who isn’t a data scientist to understand exactly what goes on inside.
Patients don’t need to understand these algorithms at the data scientist level, but it’s still helpful for people to have a general idea of how AI-based healthcare tools work, said Suresh Balu, program director at the Duke Institute for Health Innovation. That way they can understand their limitations and ask questions about how they are used.
Some patients can get a little nervous when they hear that algorithms are being used in their care, says Mark Sendak, a data scientist at the Duke Institute for Health Innovation. “People get really scared and nervous and feel threatened by the algorithms,” he says.
Those patients may be nervous about how hospitals are using their data or concerned that their medical care is being run by computers rather than their doctors. Understanding how algorithms work — and don’t work — can help alleviate these concerns. Right now, healthcare algorithms have very limited use: A computer program doesn’t make important medical decisions, but it can help a doctor decide whether a series of medical scans should be examined more closely.
To help demystify the AI tools used in medicine today, let’s break down the components of one specific algorithm and see how it works. We chose an algorithm that marks patients in the early stages of sepsis – a life-threatening complication of infection leading to widespread inflammation throughout the body. It can be difficult for doctors to identify sepsis because the signs are subtle, especially in the beginning, so it’s a common target for artificial intelligence-based tools. This particular program also uses mathematical techniques, such as neural networks, which are typical of medical algorithms.
The algorithm we’re looking at underlies a program called Sepsis Watch, which Sendak and Balu helped develop at Duke University. Sepsis Watch went online in 2018 after about three years of development. If someone is admitted to a Duke hospital today, the program can keep an eye out.
This is what it would do.
SOURCE – www.theverge.com