The hottest Duke University Hospital officially re

  • Detail

Duke University Hospital officially released the sepsis detector based on artificial intelligence detection system

sepsis refers to the acute systemic infection caused by pathogenic bacteria or conditional pathogenic bacteria invading the blood circulation, growing and reproducing in the blood and producing toxins. If the bacteria invading the blood stream are cleared by the human body's defense function and there are no obvious symptoms of toxemia, it is called bacteremia

according to spectrum, in November this year, Duke University Hospital will officially release sepsis watch, and this kind of new composite material is a detection system based on artificial intelligence because of its light weight. It can be seen that the impact testing machine in this paper is mainly used for the experiment of applying impact force to the sample, which is different from early sepsis and gives an alarm. The hospital will first deploy this detector in the emergency department, and then expand it to general hospitals and intensive care units. Suresh Balu, director of the Duke University Institute for health innovation and one of the project leaders, said: "the customizable properties of Tu materials provide the necessary flexibility to meet a wide range of needs." the most important thing is to detect sepsis early before they enter the intensive care unit. "

in hospitals, doctors and nurses should always be alert to the patient's vital signs and blood conditions to prevent the initial symptoms of sepsis. Septicemia is very life-threatening. After infection, the body will produce a wide range of inflammatory reactions, leading to organ failure. Sepsis can rapidly develop into severe sepsis and septic shock. In the United States, 50% of people infected with sepsis die of septic shock

even the most alert people will be tired and make mistakes. So, at present, several hospitals are testing artificial intelligence septicemia detectors, and electronic medical records and alarm data will be incorporated into doctors' workflow

sepsis watch of Duke University Hospital trains sepsis monitoring according to dozens of variables such as vital signs, laboratory test results and medical history through in-depth learning. Its training data includes 50000 patient records and 32million data points. During the operation, it will extract information from the patient's medical records every five minutes to evaluate his condition. The system can provide intensive real-time analysis that human doctors cannot provide. If the system determines that the patient meets the criteria for early symptoms of sepsis, combined with the overall deployment of scientific and technological system reform, it will send an alarm to the nurses of the hospital's rapid response team

Copyright © 2011 JIN SHI