Google's DeepMind says its AI has acute kidney failure 48 hours ago

Five years after Google's acquisition of DeepMind, the Healthcare and Artificial Intelligence Group is unveiling its biggest breakthrough in the healthcare industry. Its technology can predict whether a patient has life-threatening kidney damage 48 hours before many symptoms can be recognized by doctors.

In an article published Wednesday in the journal Nature, DeepMind researchers said their algorithms correctly predicted 90% of acute kidney damage that would require dialysis. The work was the result of a project led by the US Department of Veterans Affairs to help doctors to take a treatment in advance.

"We are really excited about the potential of AI to help clinicians move from responsive care to preventative and preventative care," said Dominic King, co-founder and clinical lead of DeepMind, during an interview.

According to researchers at the University of Pittsburgh School of Medicine, about 2 million people die each year in the world from acute kidney injury. The disease, which involves a sudden episode of kidney failure or kidney damage, can be tricky to diagnose by doctors because there are not always clear and immediate symptoms. Studies have shown that early detection can reduce the risk of serious injury or death.

In 2014, Google acquired DeepMind for an estimated $ 500 million while it was considering developing AI and calling on the industry's best experts to work on complex issues. involving machine learning. As Alphabet and its various units have embarked on the health care sector in recent years, the focus has been on using its technology to predict serious health consequences before they occur. .

DeepMind's health projects will soon be integrated with Google Health, led by David Feinberg. The group has not said much publicly beyond its website, which says it's studying how AI can be used to help "diagnose cancer, predict outcomes for patients, prevent blindness, and so much more. " Much of his team remains based in the UK, although his health service is expected to move to Google's headquarters in Silicon Valley.

Even in its infancy, the company's work in the health sector has been criticized for not adequately protecting users' privacy. In a recent case, a patient sued Google and the University of Chicago Medical Center for failing to remove doctors' notes and date stamps from personal medical records. A UK government privacy officer said that a hospital had illegally sent 1.6 million records to Google DeepMind for a new health care application.

Kidney injury research comes from two separate joint studies conducted between the VA and the Royal Free Hospital in London. DeepMind said it has badyzed electronically stored data in more than 100 VA hospitals, examining information on hundreds of thousands of patients. Personal data such as names and social security numbers have been removed from the data.

In addition to predicting acute kidney failure two days earlier, the company is also exploring ways to disseminate these alerts in an emergency so that doctors recognize them properly and act accordingly.

DeepMind's King said that there was still work to be done to create a regulatory framework to introduce medical forecasting tools and better understand how they could be administered in real time.

DeepMind's breakthroughs could eventually complement the Streams mobile application, which is primarily used in the UK as a communication tool by doctors and nurses. He does not use AI at the moment, but DeepMind has long emphasized his vision of building an "AI-based badistant for nurses and doctors around the world." ".

WATCH: DeepMind CEO on machine learning