The AI tool examines routine heart CT scans to identify textural changes in the fat around the heart, which can indicate the heart muscle underneath is inflamed and unhealthy. This detection cannot be spotted by doctors through any routine medical imaging tests. The tool was used on 72,000 patients across nine NHS Trusts in England who had cardiac CT scans between 2007 and 2022.
The study found that those deemed to be in the highest risk group were 20 times more likely to develop heart failure than those in the lowest risk group, with high-risk patients having a one in four chance of developing heart failure within five years. The algorithm predicts the risk that a person developed heart failure in the next five years with 86% accuracy. Researchers are looking to roll the tool out across the NHS, where around 350,000 patients are referred for a cardiac CT scan each year in the UK.
In a separate development, the risk of serious or fatal heart disease can be predicted by AI analysis of mammograms, according to a study published in the European Heart Journal. The study involved 123,762 women who had undergone preventive mammography screening and had no known history of cardiovascular disease. Researchers used AI to analyze the amount of calcium accumulated in the breast arteries, categorized as severe, moderate, mild, or absent.
Women with mild calcification had about 30% higher risk of serious cardiovascular disease compared to those with no calcification. This finding applied even to women under 50, a group often considered low-risk, and remained strong even when other risk factors like diabetes and smoking were considered. Researchers, led by Emory University in the US, highlight that this technique could help identify many women with undiagnosed or untreated cardiovascular disease.
Policy makers could consider integrating this technology into existing mammography screening programs, potentially reaching tens of millions of women each year without additional infrastructure. The main steps required are integrating the AI tool into existing imaging workflows and establishing clear guidelines for informing doctors and patients. However, the specific timeline or steps for rolling out the Oxford AI tool across the NHS remain unclear, as do the cost implications and integration details for the mammogram AI tool.
Additionally, the potential false positive or false negative rates for these AI tools, and how they might impact patient care, have not been detailed, nor have regulatory approvals or clinical trials needed before widespread adoption been specified. It is also unknown how these AI tools compare to or complement existing risk assessment methods for heart disease.