The machine learning algorithm works by flagging abnormal, noncontrast exams for intracranial hemorrhage, experts wrote in Radiology: AI. 

Kaiser Permanente’s system works by dividing patients with suspicious nodules into eight separate categories, similar to screening mammography. 

Healthcare systems can use the tool to merge clinical health information with imaging files and perform tasks that are difficult and expensive to complete with current approaches.

The gains were particularly notable in MRI and conventional radiography and at smaller hospitals, experts reported in Insights into Imaging. 

Queen Elizabeth Hospital is now in the process of purchasing a new PACS system to reduce some unnecessary manual steps. 

Innovation is crucial to the evolution of radiology, but trailblazers who take shortcuts will be left disappointed, experts advised. 

This more uniform method of reporting has now become the "backbone of oncological imaging" at one high-volume cancer center. 

As recently as eight months ago, cardiologists sitting down to work with medical images at Novant Health had plenty of choices on where and how to go about that part of their jobs.