Saarland University Hospital, Saarland, 66421 Homburg/Saar, Germany.
IEEE Trans Neural Syst Rehabil Eng. 2009 Dec;17(6):530-6. doi: 10.1109/TNSRE.2009.2032627. Epub 2009 Sep 22.
We have recently shown that click evoked auditory brainstem responses (ABRs) can be efficiently processed using a novelty detection paradigm. Here, ABRs as a large-scale reflection of a stimulus locked neuronal group synchronization at the brainstem level are detected as novel instance-novel as compared to the spontaneous activity which does not exhibit a regular stimulus locked synchronization. In this paper we propose for the first time Gabor frame operators as an efficient feature extraction technique for ABR single sweep sequences that is in line with this paradigm. In particular, we use this decomposition technique to derive the Gabor frame phase stability (GFPS) of sweep sequences of click and chirp evoked ABRs. We show that the GFPS of chirp evoked ABRs provides a stable discrimination of the spontaneous activity from stimulations above the hearing threshold with a small number of sweeps, even at low stimulation intensities. It is concluded that the GFPS analysis represents a robust feature extraction method for ABR single sweep sequences. Further studies are necessary to evaluate the value of the presented approach for clinical applications.
我们最近表明，使用新颖性检测范式可以有效地处理点击诱发的听觉脑干反应 (ABR)。在这里，ABR 作为大脑脑干水平上受刺激锁定神经元群体同步的大规模反射，被检测为新颖实例，与不表现出规则刺激锁定同步的自发活动相比是新颖的。在本文中，我们首次提出使用伽柏框架算子作为 ABR 单扫描序列的有效特征提取技术，这与该范式是一致的。具体来说，我们使用这种分解技术来推导点击和啁啾诱发 ABR 单扫描序列的伽柏框架相位稳定性 (GFPS)。我们表明，啁啾诱发 ABR 的 GFPS 即使在低刺激强度下，也可以用少量的扫描来稳定地区分自发活动和听力阈值以上的刺激。因此，GFPS 分析代表了 ABR 单扫描序列的稳健特征提取方法。需要进一步的研究来评估所提出方法在临床应用中的价值。
