Health Life

The geographic bias in medical AI tools

Credit: Pixabay/CC0 Public Domain

Just a few decades ago, scientists didn’t think much about diversity when studying new medications. Most clinical trials enrolled mainly white men living near urban research institutes, with the assumption that any findings would apply equally to the rest of the country. Later research demonstrated that assumption to be false; examples accumulated of medications that were later determined to be less effective or caused more side effects in populations that were underrepresented in the initial study.

To address these inequities, federal requirements for participation in were broadened in the 1990s, and now attempt to enroll diverse populations from the onset of the study.

But we are now at risk of repeating these same mistakes as we develop new technologies, such as AI. Researchers from Stanford University examined clinical applications of machine learning to find that most algorithms are trained on datasets from patients in only three , and that the majority of states have no represented patients whatsoever.

“AI algorithms should mirror the community,” says Amit Kaushal, an attending physician at VA Palo Alto Hospital and Stanford adjunct professor of bioengineering. “If we’re building AI-based tools for patients across the United States, as a field, we can’t have the data to train these tools all coming from the same handful of places.”

Kaushal, along with Russ Altman, a Stanford professor of bioengineering, genetics, medicine, and , and Curt Langlotz, a professor of radiology and biomedical informatics research, examined five years of peer-reviewed articles that trained a deep-learning algorithm for a diagnostic task intended to assist with patient care. Among U.S. studies where geographic origin could be characterized, they found the majority (71%) used patient data from California, Massachusetts, or New York to train the algorithms. Some 60%

Health article

Confronting alcohol use disorder and misconceptions as a woman

Forty-one-year-old Christine* started experimenting with alcohol in her early teens.

The first time she tried it, she blacked out and became physically ill. While she instantly regretted that decision, she continued drinking for years, often taking it to the extreme.

“Things fell in place when I started listening and following directions. I didn’t feel like I was a victim anymore.”

“I completely went off the deep end,” she says. “I couldn’t imagine being in a social situation without alcohol. I wanted to feel included, interesting, funny, liked, and loved. Alcohol gave me all those things.”

In her early 20s, Christine started seeing a therapist. She talked about her feelings of guilt and regret over her behavior after a night of drinking. The therapist encouraged her to talk more about it, but she denied she had a problem.

“He told me that people who don’t have a problem with drinking don’t feel the way I always did after drinking,” she says. “That was the start of me realizing that maybe I did have a problem with alcohol.”

Alcohol use disorder (AUD) affects an estimated 15 million Americans. It is a medical condition that can cause you to be unable to control how much you drink. And when you’re not drinking, you may feel anxious, irritable, or stressed.

Seeking help

At the suggestion of her therapist, Christine joined a 12-step program. But she struggled early on to accept why she was there. She felt sorry for herself and regularly cried during meetings. When she was finally able to admit that she had a problem with alcohol, she began to take steps toward recovery.

“Things fell in place when I started listening and following directions,” she says. “I got a sponsor and learned how to cope physically, mentally, and spiritually. This helped me