The relationship between different degrees of COVID-19 pneumonia and arterial stiffness based on artificial intelligence analysis

Lingheng Wu^1,2Lin Jin³Xinyi Li^4,5Mengjiao Zhang^4,6Jianxiong Chen^1,2Xiaobo Tang⁷Lianfang Du⁴Xifu Wang⁷Zhaojun Li^1,4*Xianghong Luo⁴

• ¹Shanghai General Hospital of Nanjing Medical University, Shanghai, China
• ²Mindong Hospital Affiliated to Fujian Medical University, Ningde, Fujian, China
• ³Department of Ultrasound, Guanghua Hospital Affiliated to Shanghai University of Traditional Chinese Medicine, Shanghai, China
• ⁴Shanghai General Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China
• ⁵Hubei University, Wuhan, Hubei Province, China
• ⁶Department of Medical Imaging, Weifang Medical University, Weifang, Shandong Province, China
• ⁷Department of Radiology, Jiading Branch of Shanghai General Hospital, Shanghai Jiaotong University School of Medicine, Shanghai, China

Rationale and Objectives: This study aimed to investigate the correlation between the severity of pulmonary infection and arterial stiffness pulse in Coronavirus Disease 2019 patients based on artificial intelligence (AI) quantitative analysis.A total of 100 COVID-19 patients (mean age 76 years) were enrolled in this study and were stratified into three groups (mild, moderate, or severe).An AI imaging diagnostic system was used for automatic identification and quantitative analysis of infected lesions. Arterial stiffness was evaluated by arterial velocity pulse index (AVI). The multiple linear regression analyses were used to investigate the independent associations between AVI and inflammatory markers and radiographic parameters. Hazard ratios and Kaplan-Meier curves were produced to assess the association between arterial stiffness and radiographic parameters with a composite outcome of all-cause mortality.The AVI were elevated in the moderate and severe groups compared to the mild group of COVID-19 patients (P<0.001). The multiple linear regression analyses showed that AVI was linked with the highest lobe infection percentage (β=0.813, 95%CI, 0.056-0.394, P=0.011). In multivariable Cox regression, AVI ≥ 33 was associated with all-cause death {hazard ratio, 16.201 [95% confidence interval (CI), 1.601, 163.987].As the degree of pneumonia infection increased in COVID-19 patients, vascular endothelial function was impaired and arterial stiffness increased. AVI was associated with most lobe infection percentage, and the degree of pneumonia was an independent risk factor for increased arterial stiffness. Worsening arterial stiffness poses an increased risk of death in COVID-19.