Artificial intelligence algorithms are ubiquitous in healthcare. They sort through patient data to predict who will develop diseases like heart disease or diabetes, they help doctors figure out which people are the sickest in an emergency room, and they search medical images to find clues to disease. But even if AI algorithms are becoming increasingly important for medicine, they are often invisible to people receiving care.
Artificial intelligence tools are complicated computer programs that take in massive 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 take in is a black box – unfathomable even to the people who designed the program. But even if a program is not With 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’s going on inside.
Patients don’t need to understand these algorithms at a data science level, but it’s still useful to have a general idea of how AI-based health tools work, says Suresh Balu, program director at the Duke Institute for Health Innovation. This allows them to understand their limitations and ask questions about how they are used.
Some patients can get a little nervous when they hear about algorithms being used in their treatment, 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.
These patients may be nervous about how hospitals are using their data or concerned that their medical care is being guided by computers rather than their doctors. Understanding how algorithms work—and don’t work—can help allay these concerns. Right now, algorithms in healthcare have a very limited use: a computer program won’t make important medical decisions, but it could help a doctor decide whether a series of medical scans needs closer scrutiny.
To demystify the AI tools used in medicine today, we’re going to break down the components of a specific algorithm and see how it works. We selected an algorithm that flags patients in the early stages of sepsis — a life-threatening complication of infection that leads to widespread inflammation throughout the body. It can be difficult for doctors to recognize sepsis because the signs are subtle, especially in the early stages. Therefore, it is a common target for tools based on artificial intelligence. This particular program also uses mathematical techniques such as neural networks, which are typical of medical algorithms.
The algorithm we’re looking at underpins a program called Sepsis Watch, which Sendak and Balu helped develop at Duke University. Sepsis Watch went online in 2018 after around three years of development. Today, if someone is hospitalized at a Duke hospital, the program could keep an eye on them.
Here’s what it would do.
https://www.theverge.com/c/22927811/medical-algorithm-explainer-sepsis-risk-watch How an algorithm guides a medical decision