US hospital Mayo Clinic uses AI to analyze patient histories

Mayo Clinic, a renowned U.S. medical center, is implementing an artificial intelligence tool named Record Time to streamline the analysis of patient medical histories. This innovative system processes extensive clinical records from various health centers, organizing documents chronologically and generating automatic summaries. The primary goal is to significantly cut down the time physicians dedicate to reviewing patient files, thereby enhancing their ability to identify relevant information for diagnoses and treatments.

According to CNN, the Record Time platform is designed to handle vast amounts of medical data. Dr. Alexander Ryu, an internist involved in its development, stated that the tool can save between five and 30 minutes of preparation time per consultation, depending on the case's complexity. By organizing records and creating concise summaries, the AI allows healthcare professionals to access critical information more rapidly, aiding decisions on further studies or treatment plans.

"We receive an enormous amount of medical records, tens of millions of pages each year, and we needed a way to find the important information," Ryu explained to CNN. The system not only synthesizes data but also pinpoints relevant details that might be dispersed across hundreds of documents. This ensures that doctors are better prepared for each patient encounter, reducing the risk of overlooking crucial medical history, diagnoses, or study results that could influence subsequent treatment steps.

Mayo Clinic's foray into AI in healthcare extends beyond Record Time. Matthew Callstrom, medical director of Mayo Clinic's generative AI program, revealed to CNN that the institution currently employs approximately 150 AI models across various departments. Some of these systems focus on early disease detection, including a clinical trial exploring AI's potential to identify patients at risk for pancreatic cancer or detect the disease in its initial stages.

“

We receive an enormous amount of medical records, tens of millions of pages each year, and we needed a way to find the important information