John M. O'Toole

Time-Frequency Processing of Nonstationary Signals: Advanced TFD Design to Aid Diagnosis with Highlights from Medical Applications

This article presents a methodical approach for improving quadratic time-frequency distribution (QTFD) methods by designing adapted time-frequency (T-F) kernels for diagnosis applications with illustrations on three selected medical applications using the electroencephalogram (EEG), heart rate variability (HRV), and pathological speech signals. Manual and visual inspection of such nonstationary multicomponent signals is laborious especially for long recordings, requiring skilled interpreters with possible subjective judgments and errors. Automated assessment is therefore preferred for objective diagnosis by using T-F distributions (TFDs) to extract more information. This requires designing advanced high-resolution TFDs for automating classification and interpretation. As QTFD methods are general and their coverage is very broad, this article concentrates on methodologies using only a few selected medical problems studied by the authors.

Measuring Time-Varying Information Flow in Scalp EEG Signals: Orthogonalized Partial Directed Coherence

This study aimed to develop a time-frequency method for measuring directional interactions over time and frequency from scalp-recorded electroencephalographic (EEG) signals in a way that is less affected by volume conduction and amplitude scaling. We modified the time-varying generalized partial directed coherence (tv-gPDC) method, by orthogonalization of the strictly causal multivariate autoregressive model coefficients, to minimize the effect of mutual sources. The novel measure, generalized orthogonalized PDC (gOPDC), was tested first using two simulated models with feature dimensions relevant to EEG activities. We then used the method for assessing event-related directional information flow from flash-evoked responses in neonatal EEG. For testing statistical significance of the findings, we followed a thresholding procedure driven by baseline periods in the same EEG activity. The results suggest that the gOPDC method 1) is able to remove common components akin to volume conduction effect in the scalp EEG, 2) handles the potential challenge with different amplitude scaling within multichannel signals, and 3) can detect directed information flow within a subsecond time scale in nonstationary multichannel EEG datasets. This method holds promise for estimating directed interactions between scalp EEG channels that are commonly affected by the confounding impact of mutual cortical sources.

A nonparametric feature for neonatal EEG seizure detection based on a representation of pseudo-periodicity ✩

Automated methods of neonatal EEG seizure detection attempt to highlight the evolving, stereotypical, pseudo-periodic, nature of EEG seizure while rejecting the nonstationary, modulated, coloured stochastic background in the presence of various EEG artefacts. An important aspect of neonatal seizure detection is, therefore, the accurate representation and detection of pseudo-periodicity in the neonatal EEG. This paper describes a method of detecting pseudo-periodic components associated with neonatal EEG seizure based on a novel signal representation; the nonstationary frequency marginal (NFM). The NFM can be considered as an alternative time-frequency distribution (TFD) frequency marginal. This method integrates the TFD along data-dependent, time-frequency paths that are automatically extracted from the TFD using an edge linking procedure and has the advantage of reducing the dimension of a TFD. The reduction in dimension simplifies the process of estimating a decision statistic designed for the detection of the pseudo-periodicity associated with neonatal EEG seizure. The use of the NFM resulted in a significant detection improvement compared to existing stationary and nonstationary methods. The decision statistic estimated using the NFM was then combined with a measurement of EEG amplitude and nominal pre- and post-processing stages to form a seizure detection algorithm. This algorithm was tested on a neonatal EEG database of 18 neonates, 826 h in length with 1389 seizures, and achieved comparable performance to existing second generation algorithms (a median receiver operating characteristic area of 0.902; IQR 0.835-0.943 across 18 neonates).

A new discrete-time analytic signal for reducing aliasing in discrete time-frequency distributions

The commonly used discrete-time analytic signal for discrete time-frequency distributions (DTFDs) contains spectral energy at negative frequencies which results in aliasing in the DTFD. A new discrete-time analytic signal is proposed that approximately halves this spectral energy at the appropriate discrete negative frequencies. An empirical comparison shows that aliasing is reduced in the DTFD using the proposed analytic signal rather than the conventional analytic signal. The time domain characteristics of the two analytic signals are compared using an impulse signal as an example, where the DTFD of the conventional signal produces more artefacts than the DTFD of the proposed analytic signal. Furthermore, the proposed discrete signal satisfies two important properties, namely the real part of the analytic signal is equal to the original real signal and the real and imaginary parts are orthogonal.It is not possible to generate an alias-free discrete Wigner-Ville distribution (DWVD) from a discrete analytic signal. This is because the discrete analytic signal must satisfy two mutually exclusive constraints. We present, in this paper, a new discrete analytic signal that improves on the commonly used discrete analytic signals approximation of these two constraints. Our analysis shows that - relative to the commonly used signal - the proposed signal reduces aliasing in the DWVD by approximately 50%. Furthermore, the proposed signal has a simple implementation and satisfies two important properties, namely, that its real component is equal to the original real signal and that its real and imaginary components are orthogonal.

Time-frequency detection of slowly varying periodic signals with harmonics: methods and performance evaluation

We consider the problem of detecting an unknown signal from an unknown noise type. We restrict the signal type to a class of slowly varying periodic signals with harmonic components, a class which includes real signals such as the electroencephalogram or speech signals. This paper presents two methods designed to detect these signal types: the ambiguity filter and the time-frequency correlator. Both methods are based on different modifications of the time-frequency-matched filter and both methods attempt to overcome the problem of predefining the template set for the matched filter. The ambiguity filter method reduces the number of required templates by one half; the time-frequency correlator method does not require a predefined template set at all. To evaluate their detection performance, we test the methods using simulated and real data sets. Experiential results show that the two proposed methods, relative to the time-frequency-matched filter, can more accurately detect speech signals and other simulated signals in the presence of coloured Gaussian noise. Results also show that all time-frequency methods outperform the classical time-domain-matched filter for both simulated and real signals, thus demonstrating the utility of the time-frequency detection approach.

Analysis of the time-varying cortical neural connectivity in the newborn EEG: A time-frequency approach

Relationships between cortical neural recordings as a representation of functional connectivity between cortical brain regions were quantified using different time-frequency criteria. Among these, Partial Directed Coherence (PDC) and Directed Transfer Function (DTF) and their extensions have found wide acceptance. This paper aims to assess and compare the performance of these two connectivity measures that are based on time-varying multivariate AR modeling. The time-varying parameters of the AR model are estimated using an Adaptive AR modeling (AAR) approach and a short-time based stationary approach. The performance of these two approaches is compared using both simulated signal and a multichannel newborn EEG recording. The results show that the time-varying PDC outperforms the time-varying DTF measure. The results also point to the limitation of the AAR algorithm in tracking rapid parameter changes and the drawback of the short-time approach in providing high resolution time-frequency coherence functions. However, it can be demonstrated that time-varying MVAR representations of the cortical connectivity will potentially lead to better understanding of non-symmetric relations between EEG channels.

Improved Discrete Definition of Quadratic Time-Frequency Distributions

Computation of a time-frequency distribution (TFD) requires a discrete version of the continuous distribution. This discrete TFD (DTFD) should be free from aliasing and conserve all the important mathematical properties of the continuous distribution. Existing DTFD definitions, however, poorly approximate this ideal. One popular definition, the generalized DTFD (GDTFD), is alias free but does not retain all the desirable properties from the continuous distribution. Another definition, the so-called alias-free GDTFD (AF-GDTFD), retains most properties yet is not always alias free. We propose a new DTFD definition, based on the GDTFD, that retains all desirable properties and is always alias free.

Perfusion index in the preterm infant immediately after birth

AIM To evaluate PI in preterm infants during the first 10 min of life. DESIGN/METHODS An observational study was conducted in the delivery room on preterm infants (less than 32 week gestation). PI values were obtained from a pre ductal saturation probe placed on the right wrist. Analysis was performed on the first 10 min of data to investigate the correlation of PI with gestational age, heart rate, blood pressure, and lactate values. RESULTS 33 infants with a median gestational age of 29 wks (IQR, 26-30 wks) and median birth weight of 1205 g (IQR, 925-1520 g) were included for analysis. The overall median PI value for the first 10 min was 1.3 (IQR, 0.86-1.68). There was no significant correlation found between delivery room PI and gestational age(r=0.28, 95% CI: -0.09, 0.59), lactate levels (r=-0.25, 95% CI: -0.62, 0.18) and blood pressure values (r=-0.18, 95% CI: -0.46, 0.20). An average correlation value of r=-0.417 (95% CI: - 0.531, -0.253) was found between PI and heart rate values. There was no statistical difference between the median of the median PI value over the first 5 min of life compared to the second 5 min (p=0.22). Variability, as quantified by the IQR, was higher in the first 5 min compared to the second 5 min: median of 0.5(IQR, 0.27, 0.92) vs 0.2(IQR, 0.10, 0.30) (p<0.00). CONCLUSIONS Delivery room PI values are easily obtained, however, have significant variability over the first 5 min of life and may add little to delivery room assessment.

Kalman filter-based time-varying cortical connectivity analysis of newborn EEG

Multivariate Granger causality in the time-frequency domain as a representation of time-varying cortical connectivity in the brain has been investigated for the adult case. This is, however, not the case in newborns as the nature of the transient changes in the newborn EEG is different from that of adults. This paper aims to evaluate the performance of the time-varying versions of the two popular Granger causality measures, namely Partial Directed Coherence (PDC) and direct Directed Transfer Function (dDTF). The parameters of the time-varying AR, that models the inter-channel interactions, are estimated using Dual Extended Kalman Filter (DEKF) as it accounts for both non-stationarity and non-linearity behaviors of the EEG. Using simulated data, we show that fast changing cortical connectivity between channels can be measured more accurately using the time-varying PDC. The performance of the time-varying PDC is also tested on a neonatal EEG exhibiting seizure.

Electrographic Seizures during the Early Postnatal Period in Preterm Infants

OBJECTIVE To investigate the frequency and characteristics of electrographic seizures in preterm infants in the early postnatal period. STUDY DESIGN Infants <32 weeks gestational age (GA) (n = 120) were enrolled for continuous multichannel electroencephalography (EEG) recording initiated as soon as possible after birth and continued for approximately up to 72 hours of age. Electrographic seizures were identified visually, annotated, and analyzed. Quantitative descriptors of the temporal evolution of seizures, including total seizure burden, seizure duration, and maximum seizure burden, were calculated. RESULTS Median GA was 28.9 weeks (IQR, 26.6-30.3 weeks) and median birth weight was 1125 g (IQR, 848-1440 g). Six infants (5%; 95% CI, 1.9-10.6%) had electrographic seizures. Median total seizure burden, seizure duration, and maximum seizure burden were 40.3 minutes (IQR, 5.0-117.5 minutes), 49.6 seconds (IQR, 43.4-76.6 seconds), and 10.8 minutes/hour (IQR, 1.6-20.2 minutes/hour), respectively. Seizure burden was highest in 2 infants with significant abnormalities on neuroimaging. CONCLUSION Electrographic seizures are infrequent within the first few days of birth in very preterm infants. Seizures in this population are difficult to detect accurately without continuous multichannel EEG monitoring.