人工智能可在两秒内识别心脏病

研究人员开发的一种人工智能工具能够在不到两秒内从心电图中识别心脏病迹象1。这项由帝国理工学院领导、英国心脏基金会资助的研究成果在慕尼黑欧洲心脏学会年会上发布1。

在涉及67000名美国患者的试验中，该工具识别心力衰竭的准确率达到81%，识别心脏瓣膜病的准确率达到90%1。患者通常需要等待数月才能进行心脏超声扫描1，而这项技术的快速诊断能力有望加快心脏病的早期发现。英国心脏基金会的医学主任表示，看到人工智能能够在瞬间从心电图中读出诊断结果令人兴奋1。全球每年进行约十亿次心电图检测1，这一工具的应用潜力巨大。

Researchers have developed an artificial intelligence system capable of detecting signs of heart disease from electrocardiograms in less than two seconds 1. The tool can identify both heart failure and valvular heart disease, two conditions that typically require time-consuming diagnostic procedures 1.

In a trial involving 67,000 American patients, the AI achieved an 81% accuracy rate for detecting heart failure and a 90% accuracy rate for identifying valvular heart disease 1. The technology was developed with funding from the British Heart Foundation and led by researchers at Imperial College London, with results presented at the European Heart Rhythm Association annual conference in Munich, described as the world's largest cardiology meeting 1. According to Professor Fu Siong Ng, patients normally face waits of several months to access cardiac ultrasound scans, a delay the new tool could help reduce 1. Dr Sonya Babu-Narayan noted the significance of the breakthrough, stating: "It is exciting to see AI now able to read results from an electrocardiogram in the blink of an eye" 1. With approximately one billion electrocardiograms performed globally each year, the rapid diagnostic capability could accelerate early detection of heart disease across healthcare systems 1.