Choosing a treatment for uterine fibroids can be a difficult decision. You might wonder: will this procedure work for me? A new analysis of existing research suggests that machine learning—a type of artificial intelligence—might one day help answer that question.
The review looked at 14 studies where researchers trained computer models to predict how well women would respond to treatments like focused ultrasound or uterine artery embolization. The models analyzed medical images and sometimes combined that data with clinical information. The best-performing models, which used both types of data, showed strong ability to distinguish between likely good and poor outcomes.
However, it's important to understand where this science stands. The analysis itself notes that many of the original studies had a high risk of bias, meaning their designs might have influenced the results. More crucially, these AI models were rarely tested on new groups of patients outside the original studies, a key step to prove they work in the real world. So while the technology is promising for future personalized care, it's not ready to guide your treatment decisions today.