YOUR HEALTH: Machine learning diagnoses strokes quicker

Every 40 seconds, someone in the U.S. has a stroke.
Published: Feb. 8, 2023 at 4:32 AM CST|Updated: Feb. 8, 2023 at 6:38 AM CST
Email This Link
Share on Pinterest
Share on LinkedIn

LOS ANGELES, Calif. (Ivanhoe Newswire) - Each year, more than 795,000 people in the United States suffer a stroke. Every three and a half minutes someone dies from one. Getting treated quickly is vital for survival and for a better recovery. Now, doctors are using artificial intelligence to diagnose stroke quicker than ever before.

Every 40 seconds, someone in the U.S. has a stroke. Fifty percent of patients who have one are left disabled for life. The mantra in stroke treatment is time is brain.

Neuroradiologists at University of California, Irvine are studying how artificial intelligence and machine learning can detect strokes quicker than ever before.

“Our goal of this research is to actually, hopefully, improve door-to-treatment times by expediting the diagnoses,” neuroradiologist at UCI Health, Dr. Jennifer Soun, MD, explains to Ivanhoe.

AI uses machine learning, instantly comparing a CAT scan, or brain image, of someone in the ER with images of tens of thousands of previous stroke patients.

Dr. Soun says, “These AI tools actually help us by triaging the cases that might demand urgent attention or the positive cases that should go to stroke treatment.”

And just as important as diagnosing a stroke, machine learning will soon be able to rule out problems that mimic a stroke, and it will all be done with an app on the radiologist’s cell phone.

“It can deliver results within seconds,” Dr. Soun adds.

The latest studies also show knowing the early signs of a stroke such as loss of balance, facial paralysis, weakness, and loss of speech, could also help save your life. That’s because patients who arrive at the emergency room within three hours of their first symptoms often have less disability three months after a stroke than those who receive delayed care.

Click here to report a typo.