Coherence between one random and one periodic signal-application to the EEG during sensory stimulation

de, A.M.F.L. Miranda; Simpson, David M.; Fernando, A

doi:10.1109/iembs.2000.901139

Cited by 1 publication

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, for the case of periodic, deterministic stimulation, coherence may be estimated using only the EEG signal [1], which simplifies its estimation as well as reduces random errors due to noise in data acquisition. In order to evaluate the coherence-based detectors' performance, the sampling distribution of coherence between one random and one periodic signal was derived in [2] with the aim of obtaining its confidence limits and the probability of detecting a response (PD) if such is present. The latter is shown in Fig.…”

Section: Introductionmentioning

confidence: 99%

“…In the present work such methodology is further investigated with Monte Carlo simulations and applied to the EEG of 12 normal subjects during photic stimulation. [k], with the first being a periodic, deterministic signal and the latter, a random one may be obtained using the well known approach of dividing the signals into M segments as [3]: [4] and some further manipulations as [2]:…”

Section: Introductionmentioning

confidence: 99%

“…As it can be seen, for a fixed SNR-value in y 1 [k], an infinite number of PD-values may be found, depending on the SNR in y 2 [k]. On the other hand, if we fix the probability of detection, the infinite pairs of SNR in y 1 [k] and in y 2 [k] (SNR 1 and SNR 2 ) leading to the same PD may be found as the contour maps in the xy-plane of Fig.3. Such information is very important in order to obtain the minimum SNR in y 2 [k] that improves the detection rate when using both y 1 In Fig.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

On the detection of visual evoked responses by using multiple coherence

Félix

2001 Conference Proceedings of the 23rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society

Self Cite

View full text Add to dashboard Cite

Abstract-The coherence between the stimulation signal and the EEG has been used in the detection of evoked responses. However, the detector's performance depends on both the signal-to-noise ratio of the responses and the number of data segments used in the coherence estimation. Recently, the multiple coherence between the stimulation signal and the EEG has been suggested as a way of improving such detection. In this work such methodology is further investigated using Monte Carlo simulations, and the effectiveness of the method is illustrated on data recorded from 12 young normal subjects during rhythmic photic stimulation. Keywords-EEG, photic stimulation, detection of responses, coherence, multiple coherence, Monte Carlo simulations I. INTRODUCTIONThe coherence between the stimulation signal and the EEG has been proposed as a way of detecting evoked responses embedded in the background EEG [1]. Thus, the detection task is achieved by comparing the estimated coherence with a threshold, which is obtained based on its well known sampling distribution under the null hypothesis of zero coherence (absence of evoked responses).The main advantage claimed for the technique lies on the fact that the detector is very robust, since the threshold is independent of both the shape of the response and the signal-to-noise ratio (SNR). Thus, the probability of mistakenly detecting a response (probability of false alarm) will be constant and equal to the significance level of the test. In addition, for the case of periodic, deterministic stimulation, coherence may be estimated using only the EEG signal [1], which simplifies its estimation as well as reduces random errors due to noise in data acquisition. In order to evaluate the coherence-based detectors' performance, the sampling distribution of coherence between one random and one periodic signal was derived in [2] with the aim of obtaining its confidence limits and the probability of detecting a response (PD) if such is present. The latter is shown in Fig. 1 for different SNRvalues (in dB) and number of segments used in coherence estimation (M). As it can be seen, for a given SNR, the detection can only be improved by increasing M, which results in a larger stretch of EEG signal to be processed. This may constitute a serious limitation for low SNR-values, since a suitable number of segments could become large, leading to a whole data length greater than the period during the which the EEG may be considered stationary. Furthermore, in clinical applications of evoked responses such as in monitoring surgeries, one is often interest in avoiding injuries in the nervous fibbers due to surgical procedures, and therefore in detecting responses as fast as possible.Recently, the use of multiple coherence between the stimulation and two EEG signals has been proposed as an alternative to improve the detection rate without increasing the number of segments used [3]. In the present work such methodology is further investigated with Monte Carlo simulations and applied to the EEG of 12 normal...

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%