Researchers found that using machine learning algorithms pinpointed 25 percent more diabetics at risk of kidney damage than clinical tools and human judgment.

Source: AI mines EHR data to predict diabetic patients at risk for kidney damage, study finds | Healthcare IT News