Automated Multimodal Performance Evaluation in Simulation-based Medical Education using Natural Language Processing

Simulation-based medical education (SBME) is an effective method leveraging cyber-physical-human system functions for training health care professionals (HCPs). Comprehensive assessment of learners during SBME requires multimodal assessment of factors at the learner, instructor, and training environment levels. Currently instructors often assess performance of learners via checklists or paper-based assessment scales and there is little insight into learner performance as it relates to temporal changes occurring at these three levels. Often instructor assessments are either summative or delayed with respect to occurrence of key events during training. In this research, we aim to address these limitations by incorporating natural language processing (NLP) to automate performance assessment as well as synchronize assessments with changes occurring at the learner, instructor, and training environment levels.