Google's DeepMind Can Predict Your Kidney Illness Even Before It Occurs

Every year, Acute Kidney Injury takes away the lives of 100,000 people in the UK and 500,000 in the US. The common reason is that the disease isn't detected early enough, and now researchers want to apply AI technology to change that situation. The Department of Veterans Affairs in the United States has partnered with Alphabet-owned artificial intelligence company DeepMind to create an algorithm to predict AKI.

According to what Nature published, the algorithm is able to detect AKI’s presence 48 hours or less before the illness actually occurs. The model of this prediction algorithm identified correctly nine out of ten patients who needed dialysis with their worsening condition.

Moreover, DeepMind had its own smart assistant called Streams for clinicians. Researchers from University College London help evaluate this mobile assistant. According to the evaluation results, specialists reviewed emergency cases through this app within fifteen minutes instead of several hours. Moreover, rather than 12.4% of Acute Kidney Injury cases missed without the help of the app, only 3.3% were missed. In addition, Streams led to cost savings in health care. This smart assistant, which is combined with the new AKI prediction algorithm of DeepMind, could offer earlier and improved detection.

With the help of Streams and AKI-detecting algorithm, DeepMind expects to improve the medical field from an old reactive model to a new approach based on prevention. The company also does not intend to prevent only AKI but plans to look at other preventable diseases which cause millions of deaths every year. The researchers from the DeepMind aims to carry forward this project as a member of Google Health – the collaboration expected to foster impact globally.

DeepMind wrote: