Bone density scans can now quickly identify indicators of cardiovascular health risk.
Thanks to artificial intelligence, we may soon be able to predict our risk of developing a serious health condition in the future with the push of a button.
Abdominal aortic calcification (AAC) refers to the accumulation of calcium deposits in the walls of the abdominal aorta. This may indicate an increased risk of cardiovascular events such as heart attack or stroke.
It also predicts the risk of falls, fractures, and late-life dementia. Conveniently, common bone densitometry scans used to detect osteoporosis can also detect AAC.
However, analyzing the images requires a highly trained professional reader, and this process can take 5 to 15 minutes per image.
But researchers from Edith Cowan University’s (ECU) School of Science and School of Medical and Health Sciences have teamed up to develop software that can make scanning much faster – around 60,000 images in a day.
Researcher and Heart Foundation Future Leader Fellow Associate Professor Joshua Lewis said this significant increase in efficiency was essential to the widespread use of AAC in research and to helping people avoid developing future health problems. Ta.
“These images and automated scores can be quickly and easily obtained during bone density testing, which could lead to new approaches for early-stage patients in the future.” cardiovascular disease “We can detect and monitor diseases during daily clinical practice,” he said.
Save a lot of time
The results are the result of an international collaboration between ECU, the University of Washington, the University of Minnesota Southampton, the University of Manitoba, the Marcus Institute on Aging, and Hebrew Senior Life Harvard Medical School. It is truly an interdisciplinary, global effort.
Although not the first algorithm developed to assess AAC from these images, this study is the largest of its kind and is based on the most commonly used bone density mechanical model. , is the first to be tested in a real-world setting using images. It is done as part of a routine bone density test.
More than 5,000 images were identified that were analyzed by experts and the team’s software.
After comparing the results, the experts and the software reached the same conclusion about the degree of AAC (low, moderate, or high) 80% of the time. An impressive number considering this is the first version of the software.
Importantly, only 3% of those deemed to have high AAC levels were incorrectly diagnosed by the software as having low AAC levels.
“This is noteworthy because these are the people with the greatest degree of disease and the highest risk of fatal and non-fatal cardiovascular events and all-cause mortality,” Professor Lewis said.
“While there is still work to be done to improve the functionality of the software, Accuracy When compared to human measurements, these results are from the version 1.0 algorithm, and the latest version has already significantly improved results.
“Automated assessment of the presence and extent of AAC with imaging expert-like accuracy enables large-scale screening for cardiovascular disease and other conditions even before symptoms appear.”
“This allows people at risk to make the necessary lifestyle changes much earlier, putting them in a position to be healthier later in life.”
Reference: “Machine learning for abdominal aortic calcification assessment from bone density machine-derived lateral spine images” Naeha Sharif, Syed Zulqarnain Gilani, David Suter, Siobhan Reid, Pawel Szulc, Douglas Kimelman, Barret A. Monchka, Mohammad Jafari Jozani, Jonathan M. Hodgson, Mark Sim, Kun Zhu, Nicholas C. Harvey, Douglas P. Keel, Richard L. Prince, John T. Scarbaugh, William D. Leslie, Joshua R. Lewis, e-biomedicine.
DOI: 10.1016/j.ebiom.2023.104676
The Heart Foundation funded this project thanks to three years of research support from Professor Lewis’ 2019 Future Leadership Fellowship.