Comparing Clinical Efficacy of Artificial Intelligence-powered Ultrasound with Pulse Index Continuous Cardiac Output for Monitoring Critically ill Patients

Yi Fang,Di Lv,Haidong Zhou,Xiaoxiao Sun,Yuanzhuo Chen,Yugang Zhuang
DOI: https://doi.org/10.21203/rs.3.rs-61575/v1
2020-08-20
Abstract:Abstract Background: To compare safety and clinical efficacy of artificial intelligence (AI)-powered ultrasound with pulse index continuous cardiac output (PiCCO) for monitoring critically ill patients. Methods: Patients who were admitted to our hospital from April to June 2020 were recruited. PiCCO was employed to monitor cardiac output (CO) and extravascular lung water index (ELWI). Simultaneously, an AI-powered ultrasound was used to automatically monitor CO and the left ventricular outflow tract velocity time integral (LVOT-VTI), and Lung ultrasound B line. Results: A total of 41 patients were enrolled, the male/female ratio was 26:15, and the patients’ median age was 73.6±8.85 years old. There was no significant difference between PiCCO and AI-powered ultrasound in monitoring of CO (t = 1.01, P = 0.316), and the correlation between these two techniques was significant (r=0.911; 95% confidence interval (CI): [0.82; 0.96]; P < 0.001). Similarly, the correlation between lung ultrasound and ELWI was significant r=0.770 (95%CI: [0.58; 0.88]; P < 0.001). Conclusions: Clinical management of critically ill patients can be achieved via monitoring techniques, such as PICCO and AI-powered ultrasound.
What problem does this paper attempt to address?