Ke Ji Ri Bao·2025-09-04 23:38Core Insights - The AI model iAorta, developed by Zhejiang University First Affiliated Hospital and Alibaba Damo Academy, can quickly identify acute aortic syndrome risks in routine CT scans within seconds [1][3] - iAorta has been deployed in the first 10 hospitals in Zhejiang and aims for nationwide expansion [1][4] Group 1: AI Model Development - The iAorta model was developed to address the challenge of diagnosing acute aortic syndrome, which often goes undetected due to atypical symptoms [2][3] - The model integrates with hospital PACs systems, allowing simultaneous analysis of CT scans as doctors review them [3][4] Group 2: Performance Metrics - iAorta demonstrated a sensitivity of 97% and specificity of 94% in identifying acute aortic syndrome [4] - In a retrospective study involving 130,000 patients, iAorta achieved a sensitivity of 92.6% and specificity of 99.2%, significantly reducing the misdiagnosis rate from 48.8% to 4.8% [4][5] Group 3: Clinical Impact - The average diagnosis time for acute aortic syndrome was reduced from 11.4 hours to 1.1 hours with the use of iAorta [5] - In a prospective study with over 15,000 patients, iAorta identified 22 cases of acute aortic syndrome, achieving a specificity of 99.4% and sensitivity of 95.5% [6]