Google‘s DeepMind business and the Department of Veterans Affairs have developed a machine learning-based application that works to predict early symptoms of kidney failure.
A study published in the journal Nature found that the AI model could help clinicians spot acute kidney injury in patients up to two days in advance and could potentially mitigate the need for serious medical procedures like dialysis or a kidney transplant, VA said Wednesday.
DeepMind analyzed health care data stored from a network of more than 100 VA sites to develop the system, which predicted 90 percent of severe AKI cases that eventually end up requiring dialysis during the clinical trial.
The VA Palo Alto Health Care System in California aims to determine how the technology can support the treatment decision-making process. VA noted it wants to apply AI advancements in efforts to improve the delivery of care to military veterans.