April 30, 2025

New machine algorithm can identify heart, fracture risks with routine bone scans

Researchers from Edith Cowan University and University of Manitoba have developed a groundbreaking machine learning algorithm that can efficiently analyze bone density scans for hidden health risks. The innovative technology can detect abdominal aortic calcification, a critical marker for heart disease, strokes, and potential falls, in less than a minute. By examining routine bone scans, the algorithm provides insights into cardiovascular health that would typically go unnoticed during standard screenings. This breakthrough could significantly improve early diagnosis and preventive care for millions of older adults, particularly women who are often under-screened for cardiovascular conditions.

"Women are recognised as being under-screened and under-treated for cardiovascular disease" - Cassandra Smith, ECU Research Fellow

Australian and Canadian researchers have developed a cutting-edge machine learning algorithm capable of rapidly identifying heart disease and fracture risks using routine bone density scans.

Key Points

1 Machine learning algorithm rapidly identifies heart disease risks

2 Detects abdominal aortic calcification in under one minute

3 58% of older women show moderate to high cardiovascular risk

4 Predicts fall and fracture likelihood more accurately than traditional methods

The innovation, developed by researchers from Australia's Edith Cowan University (ECU) in conjunction with Canada's University of Manitoba, could pave the way for more comprehensive and earlier diagnoses during routine osteoporosis screenings, improving outcomes for millions of older adults, Xinhua news agency reported.

The automated system analyses vertebral fracture assessment (VFA) images to detect abdominal aortic calcification (AAC) -- a key marker linked to heart attacks, strokes, and falls.

Traditionally, assessing AAC requires around five to six minutes per image by a trained expert. The new algorithm slashes that time to under a minute for thousands of images, making large-scale screening far more efficient, it said.

About 58 per cent of older women undergoing routine bone scans showed moderate to high levels of AAC, many of them unaware of the elevated cardiovascular risk, ECU research fellow Cassandra Smith said.

"Women are recognised as being under-screened and under-treated for cardiovascular disease," Smith said.

"People who have AAC don't present any symptoms, and without doing specific screening for AAC, this prognosis would often go unnoticed. By applying this algorithm during bone density scans, women have a much better chance of a diagnosis," Smith added.

Further research by ECU's Marc Sim revealed that AAC is not only a cardiovascular risk indicator but also a strong predictor of falls and fractures. In fact, AAC outperformed traditional fall risk factors like bone mineral density and past fall history.

"The higher the calcification in your arteries, the higher the risk of falls and fractures," Sim said, adding clinicians typically overlook vascular health in fall assessments, and this algorithm changes that.

"Our analysis uncovered that AAC was a very strong contributor to fall risks and was actually more significant than other factors that are clinically identified as fall risk factors."

Sim said that the new machine algorithm, when applied to bone density scans, could give clinicians more information about the vascular health of patients, which is an under-recognised risk factor for falls and fractures.

No comments:

Post a Comment