Artificial intelligence wants to make you happy. Streaming apps suggest music and songs based on your past choices. Advertising apps note when icy white snow is piling up outside and tempt you with warm beach vacations. Web mapping services suggest routes to avoid traffic jams. Dating apps may even help you find true love.
New tools are helping combine artificial intelligence technologies with medical expertise to help doctors make faster, more informed and humane decisions.
Clinical researchers at Stanford Medicine are now using AI technologies to help you with another kind of relationship, the one with your doctor. These software algorithms are empowering your physician to deliver better, faster and more personalized medical advice to you. Here is a short list of these projects.
The Green Button project
Physicians can fill out a short online form and hit a green button to submit medical questions to specialists experienced in mining data within Stanford’s 150 million-plus patient records for evidence-based advice.
These experts use AI algorithms to answer a physician’s question within seconds, followed by a live conversation to discuss the results. This saves time associated with manually searching medical literature for answers. Also, real-world evidence from the patient records answers clinical questions that aren’t addressed in randomized clinical trials or medical guidelines.
The service, which originated at Stanford, enables patients to receive personalized medical advice based on the experiences of people with similar health and demographic profiles.
Mining for hidden disease risk
Researchers have developed an AI program that can detect familial hypercholesterolemia, a rare genetic condition in which high levels of cholesterol in the blood result in a twentyfold increase in risk of coronary artery disease. By prescreening all patients at a clinic for this disease, high-risk individuals of this largely hidden disease can be treated early, avoiding serious complications such as heart attacks.
The researchers are exploring the feasibility of using this program for other diseases to provide cost-effective, population-level screening for conditions that might otherwise go undetected and untreated.
Updating heart attack, stroke risk data
A team of disease prevention researchers recently used AI technology to update the data traditionally used to determine a person’s risk of a heart attack or stroke. The old data included information about the health, treatments and outcomes of patients from more than 70 years ago. Researchers developed more accurate risk equations by extracting contemporary data from the records of more than 26,000 racially diverse patients, using sophisticated AI statistical methods.
“A lot has changed in terms of diets, environments and medical treatment since the 1940s,” said Sanjay Basu, MD, PhD, assistant professor of primary care outcomes research. “So, relying on our grandparents’ data to make our treatment choices is probably not the best idea.”
The research suggests that 11 million Americans may need to talk to their doctors about taking different prescriptions of aspirin, statins and blood pressure medications.