BCI Lab, Institute for Knowledge Discovery, Graz University of Technology, Krenngasse 37, 8010 Graz, Austria.
Med Biol Eng Comput. 2012 Apr;50(4):347-57. doi: 10.1007/s11517-012-0877-9. Epub 2012 Mar 8.
Steady-state somatosensory evoked potentials (SSSEPs) have been elicited applying vibro-tactile stimulation to all fingertips of the right hand. Nine healthy subjects participated in two sessions within this study. All fingers were stimulated 40 times with a 200-Hz carrier frequency modulated with a rectangular signal. The frequencies of the rectangular signal ranged between 17 and 35 Hz in 2 Hz steps. Relative band power tuning curves were calculated, introducing two different methods. Person-specific resonance-like frequencies were selected based on the data from the first session. The selected resonance-like frequencies were compared with the second session using an ANOVA for repeated measures to investigate the stability of SSSEPs over time. To determine, if SSSEPs can be classified with a classifier based on unseen data, an LDA classifier was trained with data from the first and applied to data from the second session. Person-specific resonance-like frequencies within a range from 19 to 29 Hz were found. The relative band power of the resonance-like frequencies did not differ significantly between the two sessions. Significant differences were found for the two methods and the used channels. SSSEPs were classified with a hit rate from 51 to 96 %.
稳态体感诱发电位(SSSEP)通过对右手所有指尖施加振动触觉刺激来诱发。本研究共有 9 名健康受试者参加了两次实验。所有手指均以 200Hz 的载波频率进行刺激,载波频率由矩形信号调制,其矩形信号的频率以 2Hz 的步长在 17 到 35Hz 之间变化。通过两种不同的方法计算了相对频带功率调谐曲线。基于第一次实验的数据选择了特定个体的共振样频率。使用重复测量的方差分析比较了选定的共振样频率与第二次实验之间的稳定性。为了确定 SSSEP 是否可以基于未见数据进行分类,使用来自第一组的 LDA 分类器进行训练,并应用于第二组的实验数据。在 19 到 29Hz 的范围内发现了特定个体的共振样频率。两个实验之间的共振样频率的相对频带功率没有显著差异。两种方法和使用的通道之间存在显著差异。SSSEP 的分类准确率为 51%到 96%。