Intensive care units (ICUs) are using artificial intelligence (AI) to save patients, by helping staff respond faster to data collected by the ICU.
Following trials of a "smart" electronic medical record system across 15 U.S. ICUs, University of San Francisco researchers in 2016 piloted an AI system for detecting sepsis, which reduced patient death rates, and appeared to accelerate recovery.
Medical AI systems typically function as background tools in hospital computer systems, monitoring vital signs, then flagging trouble to doctors via alerts.
The University of California, San Francisco Medical Center's Christopher Barton said AI systems also ameliorate pressure on ICU staff, by reducing "alarm fatigue"—the tendency to discount alarms, because many turn out to be false.
Most ICU AIs can more accurately predict patient risks via ongoing data feeds from vital-sign monitors and electronic medical records, helping medical personnel to ignore false alarms and concentrate on meaningful patterns.
From NBC News
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA