Arnaldo Batista

ARMA Modelling of Sleep Spindles

Differences in EEG sleep spindles constitute a promising indicator of neurodegenerative disorders. In this paper an ARMA modelling to sleep spindles is proposed and tested. The primary objective is to distinguish, via poles and zeros location, between regular, elderly and dementia subjects. In order to achieve this goal, a model validation has been done.

A Contribution for the Automatic Sleep Classification Based on the Itakura-Saito Spectral Distance

Sleep staging is a crucial step before the scoring the sleep apnoea, in subjects that are tested for this condition. These patients undergo a whole night polysomnography recording that includes EEG, EOG, ECG, EMG and respiratory signals. Sleep staging refers to the quantification of its depth. Despite the commercial sleep software being able to stage the sleep, there is a general lack of confidence amongst health practitioners of these machine results. Generally the sleep scoring is done over the visual inspection of the overnight patient EEG recording, which takes the attention of an expert medical practitioner over a couple of hours. This contributes to a waiting list of two years for patients of the Portuguese Health Service. In this work we have used a spectral comparison method called Itakura distance to be able to make a distinction between sleepy and awake epochs in a night EEG recording, therefore automatically doing the staging. We have used the data from 20 patients of Hospital Pulido Valente, which had been previously visually expert scored. Our technique results were promising, in a way that Itakura distance can, by itself, distinguish with a good degree of certainty the N2, N3 and awake states. Pre-processing stages for artefact reduction and baseline removal using Wavelets were applied.

Railscan: A Tool for the Detection and Quantification of Rail Corrugation

Rail corrugation is a phenomenon that leads to a waving in the rails with wavelengths typically between 3 cm and 100 cm and amplitude levels of several microns. The genesis of this waving is complex. Rail corrugation is a recognized problem that leads to excess vibration on the rails and vehicles to a point of reducing their life span and compromising safety. In urban areas excess vibration noise is also a problem. A software tool was developed to analyze accelerometer signals acquired in the boggies of rail vehicles in order to quantify the rail corrugation according to their frequency and amplitude. A wavelet packet methodology was used in this work and compared with the One Third Octave Filter (OTOF) power representations, which is currently used in the industry. It is shown that the former produces better results.

A multichannel time-frequency and multi-wavelet toolbox for uterine electromyography processing and visualisation

The uterine electromyogram, also called electrohysterogram (EHG), is an electrical signal generated by the uterine contractile activity. The EHG has been considered a promising biomarker for labour and preterm labour prediction, for which there is a demand for accurate estimation methods. Preterm labour is a significant public health concern and one of the major causes of neonatal mortality and morbidity [1]. Given the non-stationary properties of the EHG signal, time-frequency domain analysis can be used. For real life signals it is not generally possible to determine a priori the suitable quadratic time-frequency kernel or the appropriate wavelet family and relative parameters, regarding, for instance, the adequate detection of the signal frequency variation in time. There has been a lack of a comprehensive software tool for the selection of the appropriate time frequency representation of a multichannel EHG signal and extraction of relevant spectral and temporal information. The presented toolbox (Uterine Explorer) has been specifically designed for the EHG analysis and exploration in view of the characterisation of its components. The starting point is the multichannel scalogram or spectrogram representation from which frequency and time marginals, instantaneous frequency and bandwidth are obtained as EHG features. From this point the detected components undergo parametric and non-parametric spectral estimation and wavelet packet analysis. Intrauterine pressure estimation (IUP) is obtained using the Teager, RMS, wavelet marginal and Hilbert operators over the EHG. This toolbox has been tested to build up a dictionary of 288 EHG components [2], useful for research in preterm labour prediction.

ARMA Modelling for Sleep Disorders Diagnose

Differences in EEG sleep spindles constitute a promising indicator of sleep disorders. In this paper Sleep Spindles are extracted from real EEG data using a triple (Short Time Fourier Transform-STFT; Wavelet Transform-WT; Wave Morphology for Spindle Detection-WMSD) algorithm. After the detection, an Autoregressive–moving-average (ARMA) model is applied to each Spindle and finally the ARMA’s coefficients’ mean is computed in order to find a model for each patient. Regarding only the position of real poles and zeros, it is possible to distinguish normal from Parasomnia REM subjects.

Short Time Fourier Transform and Automatic Visual Scoring for the Detection of Sleep Spindles

Sleep spindles are the most interesting hallmark of stage 2 sleep EEG. Their accurate identification in a polysomnographic signal is essential for sleep professionals to help them mark Stage 2 sleep. Visual spindle scoring however is a tedious workload. In this paper two different approaches are used for the automatic detection of sleep spindles: Short Time Fourier Transform and Automatic Visual Scoring. The results obtained using both methods are compared with human expert scorers.

Sleep spindles: Decomposition, parameterization and applications

A new methodology for sleep spindles characterization based on the amplitude/frequency modulation (AM/FM) model is introduced after an empirical mode decomposition (EMD) signal procedure. A set of intrinsic mode functions (IMF) is obtained which are basically AM/FM signals that we can demodulate to obtain sets of parameters characterizing the amplitudes and phases. The demodulation leads to the instantaneous envelope (IE) and the instantaneous frequency (IF) estimation and subsequent modelling as AM/FM signals. Three AM/FM models are compared for the analysis of the signals under scrutiny. For the calculation of the model parameters two different methods were also used, namely, the least squares method and genetic algorithms. Finally, the parameters of the AM/FM model are used to compare healthy and Alzheimer or dementia patients. It was concluded that, based only on analysis of the IMF in the range of 11–15 Hz, five parameters of the model can quantify significantly changes on the structure of sleep spindles in patients with dementia. In (Ktonas, 2007) it was carried out a similar study, but based on “complete” sleep spindles and, for the same AM/FM model, only two of the parameters showed significant variations. This led us to believe that the IMFs, not only can quantify the variations in the structure of sleep spindles of demented patients, as they do it in a more expressive form.

MicroECG: An Integrated Platform for the Cardiac Arrythmia Detection and Characterization

A software tool for the analysis of the High-Resolution Electrocardiogram (HR-ECG) for Arrhythmia detection is introduced. New algorithms based on Wavelet analysis are presented and compared with the classic Simson protocol over the P and QRS segments of the Electrocardiogram (EEG). A novel procedure based on a two step wavelet analysis and synthesis is performed in order to obtain a frequency description of the P, T or QRS segments. This frequency “signature” is useful for the detection of otherwise asymptomatic Arrhythmia patients. The tool has been developed in Matlab, and deployed for a standalone C application.