Artificial intelligence has become a revolutionary tool in many industries, but the technology is having a particularly big impact when it comes to healthcare.
For example, researchers having been using AI to combat the flu, by building improved seasonal forecasts that inform the development of influenza vaccines. With the technology, doctors can diagnose rare diseases better so that patient can get the treatment they need faster.
Then scientists started developing more ways for further application for AI. This time, is predicting death for seriously ill patients.
According to a study by Stanford, most terminally ill patients want to spend their last days at home with the people they love, rather than receiving aggressive treatments to extend just a few days of their lives. While 80 percent of Americans say they would prefer to die at home, 60 percent die in acute care hospitals, according to the study.
Stephanie Harman, an internal medicine physician and founding medical director of Palliative Care Services for Stanford Health Care, saw an opportunity to give palliative care physicians the ability to identify and proactively reach out to patients.
At first, Harman took her idea to Nigam Shah, associate professor of medicine and biomedical informatics at Stanford University, who had previously talked about possible collaborations involving AI in healthcare with Andrew Ng, a Chinese American computer scientist.
Here, the scientists have created an algorithms which can predict if a hospital inpatient will die within 3 to 12 months, with over 90 percent accuracy.
With the help of AI, specialists would review the records of those people to discuss whether they would benefit from palliative care. That in effect, will ensure that terminally ill patients will receive the most comfort as possible in their last final months, and receive the care that best reflects their preferences.
This predictive model uses deep learning. In Stanford's case, the researchers uses AI to scan medical records of 160,000 deceased patients of Stanford Health Care and Lucile Packard Children’s Hospital who had varying illnesses (ranging from cancer to organ failure to neurological issues), medical histories, and other diseases.
And to know the exact date for each patient's death, the AI searched the given records for patterns indicative of advanced illness and encroaching death, to then assign weights to the various pieces of medical information.
When tested on the records of another 40,000 patients whose deaths were withheld, the algorithm was able to correctly determine if the patient died within a 3- to 12-month from a specific date nine times out of 10.
"It's not about replacing a doctor's judgment, ever," said Kenneth Jung, a research scientist at Stanford University School of Medicine who helped develop the AI algorithm. "This really is about providing extra care."
