In a new match of medical science and machine learning, a new Artificial Intelligence platform can successfully predict post-surgery complications in patients.
Post-surgery complications are a major concern for both physicians and patients worldwide, but a new Artificial Intelligence (AI) platform may alleviate their concerns. By automatically acquiring patients’ medical data and delivering it to doctors’ mobile devices, the AI has successfully identified postoperative complications.
The MySurgeryRisk system extracts clinical data in real time, forming a “analytic pipeline” that sends valuable results to surgeons’ mobile devices.
The findings were published on Jama Network Open after a study involving 58,236 adult patients and 74,417 inpatient surgical procedures. “The automated real-time predictions of postoperative complications using mobile device outputs had good performance in clinical settings with prospective validation, matching surgeons’ predictive accuracy,” the researchers wrote in their paper.
The platform is based on machine learning and was built using nearly seven years of data from over 74,000 procedures. The system was able to match surgeons’ accuracy in predicting surgical outcomes on a consistent basis, and it can benefit both doctors and patients by augmenting medical decision-making and reducing complications.
“This is really exciting because we prospectively validated the algorithms that we developed.” “It was critical to demonstrate that we achieved good predictive performance,” said Tezcan Ozrazgat-Baslanti, a research assistant professor of medicine and co-developer of MySurgeryRisk.
Azra Bihorac, senior associate dean for research affairs at the University of Florida College of Medicine and a lead researcher on the MySurgeryRisk project, created the system by analysing massive amounts of anonymized data from patients’ electronic medical records, including medications, lab results, and sociodemographic information, up to one year before surgery.
During the surgical procedures, the machine also collected data such as vital signs and produced quick, accurate predictions of prolonged intensive care unit stays and mortality risk. The AI also calculated the risk of eight major postoperative complications, including sepsis, acute kidney injury, and cardiovascular and neurologic issues, according to the researchers.
“MySurgeryRisk, for example, is designed to supplement surgeons’ skill and experience while also expediting their decision-making. The discovery of higher risks during surgery planning may prompt a doctor-patient discussion about whether surgery is truly necessary “According to a press release issued by Bihorac.
According to the University of Florida Health, the system is also distinct and valuable in how it collects data and distributes real-time information to physicians. Thousands of data points that could be relevant to surgical complications, such as information from ventilator monitors and anaesthesia devices, are now going uncollected in an operating room.