Researchers at NYU Langone Health created an artificial intelligence algorithm for routine computed tomography scans that automatically measures calcium buildup in the aortic valve and bone density, allowing doctors to screen for bone loss in patients that have yet to be diagnosed.

The algorithm analyzes CT scans for osteoporosis, which weakens bones and makes individuals more prone to fractures. While it is traditionally diagnosed using an X-ray, the disease often goes unidentified until a bone is fractured due to a lack of physical symptoms. 

Miriam Bredella, professor of radiology and director of the Clinical and Translational Science Institute at NYU’s Grossman School of Medicine, told WSN that the technology “could save $2.5 billion a year,” including patients’ hospital expenses if a bone is fractured.

The researchers ran tests on more than 1,500 subjects, building an algorithm that can analyze hundreds of thousands of patients without supervision within seconds and account for a variety of demographic factors.

“We now have normal values by age, sex, race, ethnicity for each vertebral body, so we can do precision medicine by focusing the treatment on exactly the patient’s specific bone health,” Bredella said, adding that the tool is particularly useful to treat NYU Langone’s diverse patient population.

The algorithm can also be used to detect heart disease and screen patients for risk of cardiovascular events, such as heart attacks or strokes, even if they had been admitted to the hospital for separate reasons. 

“Radiologists are very busy people who read these scans,” Jeffrey Berger, director of the NYU Langone Center for the Prevention of Cardiovascular Disease, said in an interview with WSN. “When somebody comes into the hospital at 2 in the morning because they’re having abdominal pain, they’re not looking to think about how much calcium there is in their aorta. The beauty of this AI algorithm is that it just does it routinely.”

Contact Ashlie West at [email protected].