Development of artificial intelligence-driven biosignal-sensitive cardiopulmonary resuscitation robot

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Cardiovascular Research

Background:

• Cardiopulmonary resuscitation (CPR) is critical for cardiac arrest survival.
• Optimizing CPR delivery remains a challenge in clinical practice.
• Current mechanical CPR devices may have limitations in adapting to individual patient needs.

Purpose of the Study:

• To evaluate an artificial intelligence (AI)-driven robot for cardiopulmonary resuscitation (CPR).
• To assess if AI-driven CPR can enhance hemodynamic parameters and clinical outcomes.
• To compare the AI-driven CPR robot against a standard mechanical CPR device.

Main Methods:

• Development of an AI-driven CPR robot with a feedback system predicting carotid blood flow (CBF).
• Study conducted on 12 pigs undergoing ventricular fibrillation and CPR.
• Comparison between the AI robot group and the LUCAS 3 device group, measuring CBF, coronary perfusion pressure (CPP), end-tidal carbon dioxide (ETCO2), and return of spontaneous circulation (ROSC).

Main Results:

• The AI model demonstrated excellent prediction performance (Pearson correlation coefficient = 0.98).
• No significant differences were observed in CBF, CPP, ETCO2 levels, or ROSC rates between the AI robot and LUCAS 3 groups.
• The AI-driven CPR robot achieved comparable hemodynamic and clinical outcomes.

Conclusions:

• This study provides proof of concept for the feasibility of an AI-driven CPR robot in a porcine cardiac arrest model.
• The AI-driven CPR robot shows potential as an alternative to existing mechanical CPR devices.
• Further research is warranted to explore the clinical utility of AI-driven CPR technology.