Researchers at NYU Langone Health have developed an artificial intelligence tool that can analyze existing CT scans—originally taken for other medical reasons—to identify signs of bone loss. The technology, created in partnership with Visage Imaging, is expected to be introduced soon as part of a clinical trial at NYU Langone hospitals, aiming to detect osteoporosis through “opportunistic screening.”
The study, published online in Radiology on November 11, examined 538,946 CT scans from 283,499 patients. These scans were performed using various machines and protocols and included a diverse patient population by age, sex, and race. The AI tool assessed bone mineral density for each major lumbar and thoracic vertebra based on demographic factors.
Findings indicated that young women under age 50 generally had higher bone density than men of the same age. However, this difference lessened with age due to steeper declines among postmenopausal women; after age 50, men had higher bone density than women. Bone density was highest among Black patients, followed by Asian patients, and lowest among White patients.
“Our study offers proof that existing medical images done for other reasons can be repurposed and used to reliably identify bone loss, such as in osteoporosis,” said Miriam A. Bredella, MD, MBA, senior investigator of the study and professor at NYU Grossman School of Medicine.
Osteoporosis affects over 10 million Americans—mostly women over 50—with another 40 million showing early signs of low bone mass. Many cases remain undiagnosed until fractures occur.
“Our goal is to use the large quantity of imaging data that we already have on-site at scale to potentially resolve the problem of underdiagnosis of osteoporosis and help people with the disease live healthier lives with stronger bones,” said Soterios Gyftopoulos, MD, MBA, co-investigator and professor at NYU Grossman School of Medicine.
Past research led by Dr. Gyftopoulos suggested that opportunistic screening could more than double annual testing rates for osteoporosis while saving Medicare over $2.5 billion each year.
Dr. Bredella added that broader adoption of AI tools like those developed at NYU Langone could significantly impact detection rates for osteoporosis since many people are unaware they have it until after a fracture.
The team also plans to adapt their approach to develop AI programs capable of diagnosing other conditions using similar scanning datasets—including heart disease and muscle loss—as previous investigations found abdominal CT scans could reveal cardiovascular risk factors as well.
Funding for this research was provided by NYU Langone Health. Other contributors from NYU Langone include Bari Dane, MD; Emilio Vega, RT; Michael P. Recht, MD; along with Malte Westerhoff, PhD (lead investigator), Norbert Lindow, PhD; Felix Herter MSc; and Khaled Bousabarah PhD from Visage Imaging in Berlin.
NYU Langone Health operates seven inpatient locations as well as more than 320 outpatient centers in New York and Florida. It has received national recognition for quality care outcomes from organizations including Vizient Inc., which ranked it No. 1 among comprehensive academic medical centers nationwide for four consecutive years.
