The Rabin Medical Center has found a way to use artificial intelligence to help medical staff decide which coronavirus patients will need to be hooked up to ventilators and at what stage of their hospitalization.
Follow Israel Hayom on Facebook, Twitter, and Instagram
The system was developed through the use of data from 1,700 coronavirus patients hospitalized at the Petah Tikvah hospital, the Sharon Regional Medical Center, and Sheba Medical Center at Tel Hashomer, 160 of whom required artificial respiration. The researchers wrote algorithms that calculate a variety of parameters related to patients' data and their medical condition. These algorithms proved 92% effective at predicting which patients would need to be hooked up to ventilators.
The system, which will be presented at an international intensive care conference in the coming days, is also capable of learning the characteristics of different coronavirus variants.
The use of artificial respiration among COVID-19 patients has changed throughout the various waves of the pandemic. In the first wave, the thought was that every COVID patient should be hooked up to a ventilator, but the results of this approach were not optimal. In the second wave, medical experts tried to hold off on hooking patients up to ventilators, relying instead on alternatives such as high-pressure oxygen. Here, too, there were instances where, by the time the patient had been hooked up to a ventilator, it was already too late.
"Up until this time, the doctors remain undecided as to when to administer artificial resuscitation to coronavirus patients. We didn't have the tools to know when to administer artificial resuscitation in a more precise manner, and it depended on the doctor's sense and the indices that were changing. The default was to wait, but sometimes we waited too long," Professor Pierre Zinger, manager of Rabin Medical Center's intensive care unit, explained.
"There were many dilemmas because this is a new disease that we still today don't know much about, and the last wave was also different from the previous waves," he explained.
"For the first time, intensive care physicians can use the system due to its high levels of reliability and precision. The doctor can also understand and analyze the recommendations. The system is a translation of my old dream not just to use the experience of veteran physicians but to also use the many data collected over the years. The system does not replace the doctor's discretion but supports the decisions," Zinger said.
For example, the system was used in the treatment of a 52-year old coronavirus patient hospitalized at Rabin Medical Center. "The algorithm saw the need to administer artificial respiration before we doctors did," said Zinger.
The system, which was funded by the Innovation Authority at the Economy Ministry, was developed in cooperation with a team of AI experts from TSG IT Advanced Systems Ltd., which specializes in machine learning and big data. The development team was headed by Shmuel Tefler and Dr. Leonid Schwartzer with the Rabin Medical Center's intensive care unit and the hospital's IT team, headed by Chief Information Officer Hagit Handel.
Subscribe to Israel Hayom's daily newsletter and never miss our top stories!
