AI Predicts Acute Kidney Injury 48 Hours In Advance

Artificial intelligence has one again proved itself to be highly beneficial to the medical industry. DeepMind, a company which focuses specifically on artificial intelligence and its applications has unveiled its latest artificial intelligence program. In a research report recently published, DeepMind has managed to successfully develop a technology which can predict acute injury within patients up to 48 hours earlier.

DeepMind collaborated with the US Department of Veteran Affairs in order to access data for its artificial intelligence program. The AI was applied to a record database from several veteran affair medical facilities. This database consisted of approximately 700,000 patients.

The research was published to Nature. It showed that the AI system had an astonishingly high success rate, managing to predict acute kidney injury 48 hours prior in 9 out of 10 patients. These patients would then be treated accordingly based on the type of kidney injury each patient experienced. Interestingly, the program was capable of predicting 60% of inpatient episodes of acute kidney injury and 90% of cases where patients required more intensive treatment.

AI capable of predicting 90.2% of acute kidney injury cases earlier.

AI capable of predicting 90.2% of acute kidney injury cases earlier.

Acute kidney injury is a result of the kidney suddenly failing to function properly. 1 in 5 hospital patients are impacted by ACI within the US and UK. The problem with acute kidney injury is detecting it due to the kidney deteriorating at a fast rate. Although the ‘injury’ may seem harmless it kills 500,000 people in the US and 100,000 in the UK annually due to late detection. With the introduction of DeepMind’s new artificial intelligence program, cases could be reduced by up to 30% as it can be diagnosed earlier.

"Over the last few years, our team at DeepMind has focused on finding an answer to the complex problem of avoidable patient harm, building digital tools that can spot serious conditions earlier and helping doctors and nurses deliver faster, better care to patients in need," said Mustafa Suleyman, co-founder & head of applied AI and Dominic King, health lead, at DeepMind.

With advanced technology capable of predicting the likelihood of a patient having an injury like ACI, we will see a drastic decrease in the number of patients hospitalised for the condition thanks to artificial intelligence. In fact, we may see a day where we will be capable preventing patients from having a life threatening disease thanks to advanced AI algorithms which would predict the likelihood of a patient obtaining it and providing a suitable measure to help combat it.

Zacharia Sharif