AI“扫一眼”就能救命？阿里达摩院开发AI预警系统——iAorta，只需普通CT快速精准揪出致命主动脉疾病

Core Viewpoint - Acute Aortic Syndrome (AAS) remains a severe cardiovascular emergency with a high mortality rate, necessitating rapid and accurate diagnosis to improve patient outcomes [3][4][7] Group 1: AAS Overview - AAS has a mortality rate of approximately 40%-50% within 48 hours of onset, with an hourly increase of 1%-2% if untreated [3] - Clinical symptoms of AAS are often non-specific and variable, complicating timely diagnosis [3][4] - Traditional imaging methods like CTA are costly and have associated risks, while the actual incidence of AAS in patients undergoing CTA is only 2.7% [3][4] Group 2: AI-Based Diagnosis - Alibaba's Damo Academy developed an AI-based warning system called iAorta, which utilizes non-contrast CT scans to identify AAS with high accuracy [5][8] - iAorta demonstrated an area under the curve (AUC) of 0.958 in a multi-center retrospective study and showed sensitivity of 0.913-0.942 and specificity of 0.991-0.993 in large-scale real-world studies [10][20] - The AI system significantly reduced the time for correct diagnosis from an average of 219.7 minutes to 61.6 minutes [10] Group 3: Real-World Applications - iAorta has been tested in various hospitals, successfully identifying AAS in patients who were misdiagnosed or refused further testing [15][19] - In a pilot at Shanghai Chang Hai Hospital, iAorta accurately identified 21 out of 22 AAS patients, with an average diagnosis time of 102.1 minutes [19][20] - The AI system enhances diagnostic accuracy for doctors of all experience levels, with significant improvements noted in interns and general practitioners [20][23] Group 4: Conclusion - The integration of AI in diagnosing AAS represents a significant advancement in emergency medicine, potentially saving lives by preventing diagnostic delays and errors [24][23]