Medical & Biological Engineering & Computing ( IF 2.6 ) Pub Date : 2021-01-25 , DOI: 10.1007/s11517-021-02323-z Felipe Antunes 1 , Tiago Zanotelli 2 , David Martin Simpson 3 , Leonardo Bonato Felix 4
|
Auditory steady-state response (ASSR) is useful for hearing threshold estimation. The ASSR is usually detected with objective response detectors (ORD). The performance of these detectors depends on the signal-to-noise ratio (SNR) as well as the signal length. Since it is undesirable to increase the signal length, then, this work provides a multivariate technique for improving the SNR and consequently the detection power. We propose the insertion of a short calibration step before the detection protocol, in order to perform a search among the available electroencephalogram (EEG) derivations and select the derivation with the highest SNR. The ORD used in this work was the magnitude-squared coherence (MSC). The standard detection protocol is to use the same EEG derivation in all exams. Using 22-scalp positions, the new technique achieved a detection rate higher than that obtained in 99.13% of the standard detection protocol. When restrictions were applied to the search, a superior performance was achieved. Thus, the technique proposed was able to track the best EEG derivations before exams and seems to be able to deal with the variability between individuals and between sessions.
Graphical abstract
中文翻译：
用于改进听觉稳态响应检测的多通道搜索策略
听觉稳态响应 (ASSR) 可用于估计听力阈值。ASSR 通常用客观响应检测器 (ORD) 检测。这些检测器的性能取决于信噪比 (SNR) 以及信号长度。由于增加信号长度是不可取的，因此，这项工作提供了一种多变量技术来提高 SNR 并因此提高检测功率。我们建议在检测协议之前插入一个简短的校准步骤，以便在可用的脑电图 (EEG) 推导中进行搜索并选择具有最高 SNR 的推导。这项工作中使用的 ORD 是幅度平方相干 (MSC)。标准检测协议是在所有检查中使用相同的 EEG 推导。使用 22 个头皮位置，新技术的检测率高于标准检测协议的99.13%。当对搜索应用限制时，实现了卓越的性能。因此，所提出的技术能够在考试前跟踪最佳 EEG 推导，并且似乎能够处理个体之间和会话之间的可变性。
京公网安备 11010802027423号