Trial uses AI to predict survival of patients requiring ECMO
A research team from Monash University has used artificial intelligence (AI) to develop an algorithm that predicts the survival rate of intensive care patients requiring external heart–lung support (ECMO).
Venoarterial extracorporeal membrane oxygenation (ECMO) is an artificial heart–lung system that operates outside the body and is used to support critically ill patients with cardiac and respiratory failure. The AI developed for this works to determine patient selection for those who will benefit from the system.
Trialled on an international cohort, the Predictive Algorithm (ECMO PAL) predicts the outcome of the ECMO treatment, in particular the risk of patient death in the hospital. This is done using a survival score.
Led by biomedical engineer Dr Andrew Stephens, Deputy Director of the Monash Cardio-Respiratory Engineering And TEchnology laboratory (CREATElab), and Dr Michael Šeman, the multidisciplinary research team set out to develop a predictive measure to improve data-driven decisions about ECMO patient selection, management and resource allocation.
“Using AI offered us an advanced alternative to statistical methods for developing risk and survival scores,” Stephens said.
Using a deep neural network (DNN) to mimic the behaviour of human learning and recognise patterns to identify and solve problems, the AI was trained using data from more than 18,000 patients treated with ECMO in more than 400 centres across Europe, North America, Latin America, Africa and the Asia–Pacific region.
“AI can be applied to large and variable datasets, and it can help uncover patterns and interactions that might otherwise be missed by traditional statistical approaches.”
Monash researchers have designed the ECMO PAL system with an algorithm that can be updated as new data and predictive variables becomes available to healthcare workers.
ECMO PAL will be made available in the coming weeks for free, global and anonymous use.
Patient-specific 3D models to assist in surgery
UNSW engineers have their sights on developing anatomically accurate 3D printed models which...
Alfred Health deploys GE system to optimise operations
The system is designed to enhance situational awareness, communication, and overall operational...
DHCRC project to deliver benchmarking tool for AI in health
The initiative complements efforts by governments, peak organisations, and clinical professional...