Biomedical Engineering Group, Department of Applied Mechanics, Indian Institute of Technology Madras, Chennai, Tamil Nadu, India.
Peter L. Reichertz Institute for Medical Informatics of TU Braunschweig and Hannover Medical School, Germany.
Stud Health Technol Inform. 2022 May 25;294:941-942. doi: 10.3233/SHTI220631.
In this work, an analysis based on complex demodulation is proposed to classify dichotomous emotional states using Electrodermal activity (EDA) signals. For this, annotated happy and sad EDA is obtained from an online public database. The sympathetic activity indices, namely Time-varying (TVSymp) and Modified TVSymp, are computed from the reconstructed EDA signal. Further, the derivative of phasic EDA is calculated from the phasic component obtained using the convex optimization (cvxEDA) based EDA decomposition method. Five statistical features are computed from each index and used for the classification. The results of the classification indicate that these features are capable of differentiating happy and sad emotional states with 75% accuracy. This technique could be effective in the identification of clinical disorders associated with happy and sad emotional states.
在这项工作中，提出了一种基于复解调的分析方法，利用皮肤电活动（EDA）信号对二分情绪状态进行分类。为此，从在线公共数据库中获得了标注的快乐和悲伤 EDA。从重构的 EDA 信号中计算出交感活动指数，即时变（TVSymp）和改进的 TVSymp。此外，从使用凸优化（cvxEDA）基于 EDA 分解方法获得的瞬态分量中计算出瞬态 EDA 的导数。从每个指数中计算出五个统计特征，并用于分类。分类结果表明，这些特征能够以 75%的准确率区分快乐和悲伤的情绪状态。该技术在识别与快乐和悲伤情绪状态相关的临床障碍方面可能是有效的。
