AKI-Predict: The use of an artificial intelligence-based prediction device in clinical practice for the early prediction of acute kidney injury

Project title

AKI-Predict: The use of an artificial intelligence-based prediction device in clinical practice for the early prediction of acute kidney injury

Description

Co-Investigators 

Dr Stefan Auer (Chemistry)

Dr Andrew Lewington (Leeds Teaching Hospital)

Acute Kidney Injury

AKI is a sudden, rapid loss of kidney function. In the UK alone, around 615,000 episodes are reported each year, leading to 100,000 deaths. It happens for many different reasons, including dehydration, conditions causing reduced blood flow to the kidneys, infection, and certain medication. It can lead to a build-up of waste products which affect other organs like the brain, heart, and lungs. While kidney health can recover if AKI is diagnosed and treated early, it often means long hospital stays and an increased risk of developing chronic kidney disease (CKD).

It is crucial that AKI is diagnosed early, but this is not simple: no single test can diagnose AKI, and it can happen to any patient in any hospital department. The current NHS AKI alert system uses indicators like changes in blood test results to recognise AKI onset, but there is no way to accurately predict who will develop AKI before it happens.

Project Background

Dr Krivov’s lab developed an artificial intelligence device called ‘AKI-Predict’. It uses an algorithm (calculation) based on the current NHS AKI alert system to rapidly and accurately predict patients’ risk of developing AKI. It sends an alert to medical staff, telling them who is most at risk of AKI, allowing them to treat these patients and prevent them from developing severe illness.

Research Overview 

In this project, researchers will conduct tests to make sure AKI-Predict works in different patients and hospitals, and in real time. They will use this evidence to use AKI-Predict in hospitals across the UK.

What this might mean for patients

AKI-Predict could prevent patients from developing severe forms of AKI, meaning shorter hospital stays, and fewer people developing CKD or needing dialysis or transplant. It could also help medical staff and scientists understand which factors increase or decrease AKI risk.