Artificial intelligence is increasingly being considered as a tool to improve patient care and reduce medical errors in the United States. A 2024 study estimated that approximately 200,000 deaths occur each year due to medical errors, though experts caution that the definition of “medical error” varies. Some analyses focus on point-of-care mistakes, while others include broader systemic failures such as lack of coordinated care or inadequate follow-up.
AI has evolved from a theoretical concept in the mid-20th century to practical applications in modern medicine, particularly since advances in data and computing resources during the 2010s. Pattern-recognition models have been effective at analyzing medical imaging scans with improvements in both speed and accuracy. For example, research conducted at UCLA demonstrated that AI models could identify pancreatic tumors on CT scans months or even years before they were detectable by human eyes. Other successful applications include screening for diabetic retinopathy, detecting heart arrhythmias via ECGs, using early warning systems for sepsis, and reading mammograms for breast cancer screenings.
Despite these advancements, clinicians and researchers advise careful adoption of AI tools. Medicine remains a personal science where context is important; therefore, most experts agree that artificial intelligence should support rather than replace human decision-making.
“Pattern-recognition models, in particular, have already proven effective at analyzing medical imaging scans, often with gains in both speed and accuracy. A study done here at UCLA found that AI models were able to identify pancreatic tumors in CT scans months, and sometimes years, before the human eye could see them,” according to the statement.
Experts believe that when used thoughtfully and under proper oversight, AI can help scan large volumes of data quickly to detect patterns associated with early-stage diseases or conditions that might otherwise go unnoticed. These systems may also enhance accuracy in reviewing images, detect early signs of deteriorating vital signs, spot gaps in follow-up care, and identify breakdowns in coordination before they cause harm.
However, keeping humans involved will be crucial for both successful implementation and broad acceptance of AI technologies within healthcare settings.
Readers can send questions to [email protected] or write to Ask the Doctors at UCLA Health Sciences Media Relations.
