AI Repurposes Routine Bone Scans to Detect Hidden Heart Disease Years Before Symptoms

Heart disease remains the leading cause of death worldwide, yet it frequently operates in the shadows. For millions of people, the first noticeable symptom of cardiovascular trouble is a catastrophic event, such as a heart attack or stroke. Alarmingly, roughly one in five heart attacks are considered "silent," meaning the damage occurs before the patient even realizes they are at risk. Traditional risk assessments, which rely heavily on indirect indicators like cholesterol levels and blood pressure, often fail to provide a complete picture of a patient's actual arterial health. Now, a major breakthrough in artificial intelligence is poised to change that dynamic, offering a way to peer inside the body's blood vessels long before a crisis strikes.[1]

An international team of researchers, led by Professor Joshua Lewis at Edith Cowan University in Western Australia, has successfully trained an artificial intelligence algorithm to detect early markers of heart disease using routine bone scans. The breakthrough demonstrates that AI can accurately identify dangerous calcium build-ups in the body's largest artery. By repurposing existing medical imaging, the technology offers a cost-effective, highly scalable method for preventative screening.[2][3]

The innovation centers on dual-energy x-ray absorptiometry, commonly known as DEXA scans. These low-dose x-ray machines are a staple of modern medicine, primarily used to assess bone mineral density and diagnose osteoporosis in older adults. Because the scans capture clear lateral images of the spine and surrounding tissues, they inadvertently photograph the abdominal aorta. For years, researchers have known that these images contained valuable cardiovascular data, but extracting that information was a logistical challenge.[1][5]

The specific marker the AI looks for is abdominal aortic calcification, or AAC. This condition involves the gradual accumulation of calcium deposits within the walls of the aorta. Medical professionals recognize AAC as a powerful, independent predictor of advanced blood vessel disease and future cardiovascular events. Crucially, this calcification often appears in the abdominal aorta years before similar blockages develop in the arteries directly supplying the heart.[1]

Catching AAC early provides a critical window for intervention, but doing so has historically been difficult. Detecting the calcification on a standard DEXA scan requires a highly trained specialist to manually review the lateral spine images. This process is painstakingly slow, expensive, and generally cost-prohibitive for widespread population screening. As a result, the cardiovascular data captured during millions of routine bone scans has largely been ignored, leaving patients unaware of their hidden risks.[3][4]

The newly developed AI algorithm eliminates this bottleneck entirely. The software automatically analyzes the lateral spine images captured by the DEXA machine, detecting and quantifying the exact level of abdominal aortic calcification. What once required a specialist's time-consuming review is now accomplished by the algorithm in a matter of seconds. The system is designed to complement existing diagnostic workflows, delivering fast, reliable results without requiring any additional effort from the radiology technician.[1][2]

The public health implications of this technology are massive, largely because it piggybacks on existing medical infrastructure. In Australia alone, approximately 700,000 people undergo DEXA bone density testing every year. In the United States, that number exceeds three million. By integrating the AI software into the machines already sitting in hospitals and clinics worldwide, healthcare systems can instantly transform a single-purpose osteoporosis check into a dual-diagnostic tool, screening millions for heart disease at zero additional cost and with zero extra radiation exposure.[1][5]

The project has garnered significant support from the medical community, including backing from the Heart Foundation's Catalyst Partnership program. Professor Garry Jennings, the Heart Foundation's Chief Medical Advisor, praised the discovery as a vital development in preventative cardiology. The research team is now actively seeking further backing from private and philanthropic funders to accelerate the software's regulatory approval and bring the intervention to clinical settings as quickly as possible.[1][3]

For patients, the transition from invisible risk to actionable data can be life-altering. Pippa Luke, a patient who participated in the research, discovered she had advanced blood vessel disease only after the AI analyzed her scan. She noted that without the technology, her condition would have remained entirely silent. The stark results served as an immediate wake-up call, prompting her to take control of her cardiovascular health before the calcification could trigger a heart attack.[3][4]

Armed with the knowledge of her AAC risk, Luke fundamentally overhauled her lifestyle. She adopted a Mediterranean diet, incorporated regular cardiovascular and resistance training into her routine, and began scheduling frequent heart health checks with her physician. Clinical data shows that providing patients with their AAC results can drive measurable improvements in cardiovascular risk factors within just twelve weeks, proving that early awareness directly translates into better health outcomes.[1][3][4]

This breakthrough represents a broader paradigm shift in how artificial intelligence is being deployed in cardiovascular medicine. Beyond bone scans, researchers are increasingly using AI to extract hidden data from other routine tests, such as measuring pericardial fat from standard CT scans or quantifying arterial calcification in mammograms. By training algorithms to see what the human eye might miss—or what human specialists simply do not have the time to look for—technology is closing the gap between medical discovery and everyday care delivery.[2][6]

Professor Lewis, who first began exploring the link between DEXA scans and heart disease in 2009 at Sir Charles Gairdner Hospital, views the AI as an empowering tool rather than a replacement for doctors. The goal is to provide clinicians with better, faster data so they can intervene earlier. Whether through dietary changes, exercise, or preventative medications, giving doctors a multi-year head start on heart disease dramatically increases the odds of patient survival.[2][4]

The research team is currently focused on the next phase of deployment, which involves securing the necessary regulatory approvals to roll the software out to commercial radiology clinics. Because the AI utilizes machines that are already universally available, the barrier to entry is significantly lower than it would be for a completely new diagnostic device. Researchers hope to see the technology available in everyday clinical practice within the next few years.[3][4]

Ultimately, the DEXA scan breakthrough highlights the immense potential of looking at old data through a new lens. By applying cutting-edge artificial intelligence to routine, decades-old imaging techniques, medical science has unlocked a powerful new defense against the world's deadliest disease. It is a testament to the idea that the next great leap in healthcare might not require inventing a new machine, but simply teaching the machines we already have to look a little closer.[1][2]