Mohamed Salah Khlif

Passive detection of accelerometer-recorded fetal movements using a time-frequency signal processing approach

This paper describes a multi-sensor fetal movement (FetMov) detection system based on a time-frequency (TF) signal processing approach. Fetal motor activity is clinically useful as a core aspect of fetal screening for well-being to reduce the current high incidence of fetal deaths in the world. FetMov are present in early gestation but become more complex and sustained as the fetus progresses through gestation. A decrease in FetMov is an important element to consider for the detection of fetal compromise. Current methods of FetMov detection include maternal perception, which is known to be inaccurate, and ultrasound imaging which is intrusive and costly. An alternative passive method for the detection of FetMov uses solid-state accelerometers, which are safe and inexpensive. This paper describes a digital signal processing (DSP) based experimental approach to the detection of FetMov from recorded accelerometer signals. The paper provides an overview of the significant measurement and signal processing challenges, followed by an approach that uses quadratic time-frequency distributions (TFDs) to appropriately deal with the non-stationary nature of the signals. The paper then describes a proof-of-concept with a solution consisting of a detection method that includes (1) a new experimental set-up, (2) an improved data acquisition procedure, and (3) a TF approach for the detection of FetMov including TF matching pursuit (TFMP) decomposition and TF matched filter (TFMF) based on high-resolution quadratic TFDs. Detailed suggestions for further refinement are provided with preliminary results to establish feasibility, and considerations for application to clinical practice are reviewed.

Accelerometer-based fetal movement detection

Monitoring fetal wellbeing is a compelling problem in modern obstetrics. Clinicians have become increasingly aware of the link between fetal activity (movement), well-being, and later developmental outcome. We have recently developed an ambulatory accelerometer-based fetal activity monitor (AFAM) to record 24-hour fetal movement. Using this system, we aim at developing signal processing methods to automatically detect and quantitatively characterize fetal movements. The first step in this direction is to test the performance of the accelerometer in detecting fetal movement against real-time ultrasound imaging (taken as the gold standard). This paper reports first results of this performance analysis.

Effective implementation of time–frequency matched filter with adapted pre and postprocessing for data-dependent detection of newborn seizures

Neonatal EEG seizures often manifest as nonstationary and multicomponent signals, necessitating analysis in the time-frequency (TF) domain. This paper presents a novel neonatal seizure detector based on effective implementation of the TF matched filter. In the detection process, the TF signatures of EEG seizure are extracted to construct the TF templates used by the matched filter. Matching pursuit (MP) decomposition and narrowband filtering are proposed for the reduction of artifacts prior to seizure detection. Geometrical correlation is used to consolidate the multichannel detections and to reduce the number of false detections due to remnant artifacts. A data-dependent threshold is defined for the classification of EEG. Using 30 newborn EEG records with seizures, the classification process yielded an overall detection accuracy of 92.4% with good detection rate (GDR) of 84.8% and false detection rate of 0.36FD/h. Better detection performance (accuracy >95%) was recorded for relatively long EEG records with short seizure events.

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.

Multichannel-Based Newborn EEG Seizure Detection using Time-Frequency Matched Filter

In recent years, much effort has been made toward developing computerized methods to detect seizures. In adults, the clinical signs of seizures are well defined and easily recognizable. But in newborns, these signs are either subtle or completely absent. For this reason, the electroencephalogram (EEG) has been the most dependable tool used for detecting seizures in newborns. Considering the non-stationary and multicomponent nature of the EEG signals, time- frequency (TF) based methods were found to be very suitable for the analysis of such signals. Using TF representation of EEG signals allows extracting TF signatures that are characteristic of EEG seizures. In this paper we present a TF method for newborn EEG seizure detection using a TF matched filter. The threshold used to distinguish between seizure and non- seizure is data-dependent and is set using the EEG background. Multichannel geometrical correlation, based on a concept of incidence matrix, was utilized to further enhance the performance of the detector.

Time-frequency characterization of tri-axial accelerometer data for fetal movement detection

Monitoring fetal wellbeing is a significant problem in modern obstetrics. Clinicians have become increasingly aware of the link between fetal activity and its well-being. Using data acquired by accelerometry sensors, we use TFDs such as the spectrogram and modified B distribution (MBD) to characterize fetal movements in the time-frequency (TF) domain. This paper reports a fetal activity detection method based on the root-mean-square (RMS) of time series and evaluates its performance against real-time ultrasound imaging, taken as the gold standard. The evaluation showed better performance with the RMS-based detector as compared to maternal perception. The evaluation also showed that the detector performance is age-dependent and that fetal movement is characterized by different TF morphology. Time-frequency distributions (TFDs) with better resolution such as MBD are investigated for TF-based techniques for the detection of fetal movements.

Detection of neonatal EEG seizure using multichannel matching pursuit

It is unusual for a newborn to have the classic “tonic-clonic” seizure experienced by adults and older children. Signs of seizure in newborns are either subtle or may become clinically silent. Therefore, the electroencephalogram (EEG) is becoming the most reliable tool for detecting neonatal seizure. Being non-stationary and multicomponent, EEG signals are suitably analyzed using time-frequency (TF) based methods. In this paper, we present a seizure detection method using a new measure based on the matching pursuit (MP) decomposition of EEG data. Signals are represented in the TF domain where seizure structural characteristics are extracted to form a new coherent TF dictionary to be used in the MP decomposition. A new approach to set data-dependent thresholds, used in the seizure detection process, is proposed. To enhance the performance of the detector, the concept of areas of incidence is utilized to determine the geometrical correlation between EEG recording channels.

A passive DSP approach to fetal movement detection for monitoring fetal health

Fetal movement can help clinicians understand fetal functional development. Active methods for fetal monitoring such as ultrasound are expensive and there are objections to their long term usage. This paper presents a passive approach for fetal monitoring which uses solid state accelerometers placed on the mothers abdomen for the collection of fetal movements. The proposed fetal movement detection is based on the root-mean-square (RMS) of time series. The detection performance is evaluated against real-time ultrasound imaging. A good detection rate of 80% and a positive predictive value of 77% were achieved based on the analysis of 4 subjects. Time-frequency (TF) analysis of fetal movement signals, using a number of quadratic TF distributions, has shown that some fetal movements are spectrally characterized by nonstationary and nonlinear behavior and that fetal activity is generally below 20 Hz. More data are needed for further TF analysis and future detections will depend on the outcome of this analysis.

Detection of neonatal seizure using multiple filters

It is often impossible to accurately differentiate between seizure and non-seizure related activities in infants based on clinical manifestations alone. The electroencephalogram (EEG) is therefore the best tool available for the recognition, management, and prognosis of neonatal seizures. The EEG signal is known to change structural characteristics between seizure and non-seizure states. In this work, matching pursuit (MP) decomposition, based on a coherent time-frequency (TF) dictionary, has provided us with a measure for quantifying changes in the structure of the neonatal EEG signal as it alternates between the various states. The quantification of state changes served as the basis for detecting seizures in 35 newborn patients. For each record, a patient-dependent threshold that marks the transition to seizure state is established. The use of multiple filters reduced the amount of artifacts and enhanced the detector performance. Overall, 93.4% detection accuracy and 0.26 false alarms per hour were achieved.

Newborn EEG seizure detection using optimized time-frequency matched filter

In recent years, much effort has been made toward developing computerized methods to detect seizures. In adults, the clinical signs of seizures are well defined and easily recognizable. This is, however, not the case for newborns where the clinical signs are either subtle or completely absent. For this reason, the electroencephalogram (EEG) has been the most dependable tool used for detecting seizures in newborns. Considering the non-stationary and multicomponent nature of the EEG signals, time-frequency (TF) based methods were found to be very suitable for the analysis of such signals. Using TF representation of EEG signals allows extracting TF signatures that are characteristic of EEG seizures. In this paper we present a TF method for newborn EEG seizure detection using a TF matched filter. The TF signatures of EEG seizures are used to construct time-frequency templates that are used by the matched filter to detect EEG seizures. The results obtained so far are very promising.