AI Machine Learning Could Lower Death Rates Following Non-Cardiac Surgery, According To A Study

An artificial intelligence approach utilising computer or machine learning may help lessen frequent cardiovascular consequences following non-cardiac surgery, such as heart attacks and lesions to the heart muscle, according to a recent study led by a PhD researcher from the University of Western Australia.

The Vascular Events in Non-cardiac Surgery Patients Cohort Evaluation (VISION) study included more than 24,000 participants, and Dr. Janis Nolde from UWA’s Medical School and the Royal Perth Hospital and an international team of researchers analysed the data. The results were published in Anaesthesia.


In order to more effectively identify and treat susceptible individuals, the team sought to determine whether machine learning and data could predict medical consequences, particularly cardiovascular complications from surgery (other than heart surgery), before they occurred.

The patient, a male, is lying on the bed.

Over 200 million people worldwide have major non-cardiac surgeries each year, and about 10 million of them have a serious cardiovascular event within 30 days, which can increase death rates, negatively impact health, and reduce long-term survival, according to Dr. Nolde.

“Heart attacks and damage to the heart muscle are the most frequent cardiovascular complications after surgery, but they are frequently challenging to detect because symptoms can be subtle and routine tests may miss them.”

The research team discovered that one in six individuals has increased levels in the first three postoperative days using a sensitive laboratory test that evaluates a protein (troponin) released into the circulation when there is damage or injury to the heart muscle.

According to Dr. Nolde, “this condition, known as myocardial injury after non-cardiac surgery, is associated with a greatly increased risk of death and other serious complications in the next few weeks, but predicting it is difficult with variables like age, fitness, any underlying medical disorders, and problems arising during or early after surgery, all of which need to be taken into account.”

“Machine learning, in particular neural networks, presents a promising strategy since these methods can analyse enormous volumes of data and find intricate patterns and relationships that are otherwise challenging to see. They can be used in a variety of settings since they are quite versatile.

Graham Hillis, Professor of Medicine at UWA and Head of Cardiology at Royal Perth Hospital, said the study’s findings point to the possibility of a promising method for more accurately identifying patients at the highest risk and those whose risk might rise over time by combining machine learning techniques with routinely collected data before, during, and after surgery.

According to Professor Hillis, this could help medical practitioners identify issues earlier and take action to prevent complications.

“Additional work is planned to optimise these approaches and incorporate them into routine care.”