Research on Intelligent Intensive Care System Based on Micro-Expression Tracking and Automated RASS Scoring

In the intensive care environment, patients' verbal communication is often limited by various factors. These challenges complicate patient management and increase uncertainties in the treatment process. Moreover, they make it difficult for traditional pain assessment methods to provide timely and accurate results, thus posing significant risks and challenges to clinical care. Therefore, to effectively enhance the efficiency and quality of intensive care, this paper introduces an innovative contactless monitoring technology that automatically assesses emotions and psychological states by real-time tracking of patients' micro-expressions. This system integrates advanced image processing techniques, artificial intelligence algorithms, and an automated Richmond Agitation-Sedation Scale (RASS) scoring tool. Utilizing a convolutional neural network (CNN) as the deep learning model, the system is designed as an Intelligent ICU—Micro-Expression Tracking and Automated RASS Scoring System, capable of monitoring emotional fluctuations in real-time and issuing alerts during emergencies. This study explores how technological innovation can optimize pain assessment based on micro-expression tracking and enhance the system's overall intelligence and accuracy through the integration of the RASS scoring tool. This study not only enhances the objectivity and accuracy of pain assessment but also achieves more personalized and efficient pain management in intensive care, significantly improving patient treatment outcomes.