This study reports on how velocity and force levels of caress-like haptic stimuli can elicit different emotional responses, which can be identified through the analysis of Autonomic Nervous System (ANS) dynamics. Affective stimuli were administered on the forearm of 32 healthy volunteers (16 women) through a haptic device with two levels of force, 2 N and 6 N, and two levels of velocity, 9.4 mm/s and 37 mm/s. ANS dynamics was estimated through Heart Rate Variability (HRV) linear and nonlinear analysis on recordings gathered before and after each stimulus. To this extent, we here propose and assess novel features from HRV symbolic analysis and Lagged Poincaré Plot. Classification was performed following a leave-one-subject-out procedure on nonlinear support vector machines. Pattern classification was split according to gender, significantly improving accuracies of recognition with respect to a “all-subjects” classification. Caressing force and velocity levels were recognized with up to 80 percent accuracy for men, and up to 84.38 percent for women. Our results demonstrate that changes in ANS control on cardiovascular dynamics, following emotional changes induced by caress-like haptic stimuli, can be effectively recognized by the proposed computational approach, considering that they occur in a gender-specific and nonlinear manner.

中文翻译：

---

通过性别特定的心率变异非线性分析对情感触觉刺激进行分类

这项研究报告了类似爱抚的触觉刺激的速度和力量水平如何引起不同的情绪反应，这可以通过对自主神经系统 (ANS) 动力学的分析来确定。通过触觉设备对 32 名健康志愿者（16 名女性）的前臂施加情感刺激，该触觉设备具有两个水平的力，2 N 和 6 N，以及两个水平的速度，9.4 mm/s 和 37 mm/s。通过对每次刺激前后收集的记录的心率变异性 (HRV) 线性和非线性分析来估计 ANS 动态。在这方面，我们在这里提出并评估 HRV 符号分析和滞后庞加莱图的新特征。在非线性支持向量机上遵循留一主题程序进行分类。模式分类按性别划分，显着提高了对“所有受试者”分类的识别准确度。男性的爱抚力和速度水平的识别准确率高达 80%，女性高达 84.38%。我们的结果表明，考虑到它们以性别特异性和非线性方式发生，所提出的计算方法可以有效识别 ANS 对心血管动力学的控制，在由类似爱抚的触觉刺激引起的情绪变化之后。