摘要
当前医护人员在使用人工智能(Artificial Intelligence,AI)辅助网络教学工作后易产生心理问题,为实时监测医护人员在远程学习过程中的情绪状态,研究提出融合多模态情感识别模型(Multimodal Fusion Recognition Model,MFRM)与卷积神经网络-类激活映射(Convolutional Neural Network-Class Activation Mapping,CNN-CAM)模型的技术方案。研究结果中MFRM模型的平均准确率达到90.23%,平均精度为89.45%,平均召回率为90.12%,平均F1分数为88.54%;CNN-CAM模型的平均准确率为87.9%,平均精度为86.7%,平均召回率为88.9%,平均F1分数为87.4%;在医护人员的个体状态预警中,研究方法计算出的心理波动值均契合预先设定的阈值。结果表明该技术方案能够有效监测网络教育个体的心理状态,有助于提升网络教育的互动性和个体心理健康水平。研究为网络教育平台提供了可靠的技术支持。
At present,medical staff are prone to psychological problems after using AI-assisted online teaching work.To monitor the emotional state of medical staff in the process of remote learning in real time,a Multimodal Fusion Recognition Model is proposed.The technical solution of MFRM and the Convolutional Neural Network-Class Activation Mapping(CNN-CAM)model.In the research results,the average accuracy rate of the MFRM model reached 90.23%,the average precision was 89.45%,the average recall rate was 90.12%,and the average F1 score was 88.54%.The average accuracy rate of the CNN-CAM model is 87.9%,the average precision is 86.7%,the average recall rate is 88.9%,and the average F1 score is 87.4%.In the individual state early warning of medical staff,the psychological fluctuation values calculated by the research method all conform to the pre-set thresholds.The results show that this technical solution can effectively monitor the psychological state of individuals in online education,which is conducive to improving the interactivity of online education and the mental health level of individuals.The research provides reliable technical support for online education platforms.
作者
王凌姗
吴虹仪
WANG Lingshan;WU Hongyi(Tianjin Medical University General Hospital Airport Hospital,Tianjin 300308,China;Tianjin Medical University Cancer Institute&Hospital,Tianjin 300060,China)
出处
《自动化与仪器仪表》
2025年第12期315-319,共5页
Automation & Instrumentation
基金
天津市医院协会医院管理研究项目《人工智能在医疗实践中的角色——医生的接受度与工作变革》(2024zc118)。
