New Cardiac MRI Gating Method Using Event-Synchronous Adaptive Digital Filter

When imaging the heart using MRI, an artefact-free electrocardiograph (ECG) signal is not only important for monitoring the patient’s heart activity but also essential for cardiac gating to reduce noise in MR images induced by moving organs. The fundamental problem in conventional ECG is the distortion induced by electromagnetic interference. Here, we propose an adaptive algorithm for the suppression of MR gradient artefacts (MRGAs) in ECG leads of a cardiac MRI gating system. We have modeled MRGAs by assuming a source of strong pulses used for dephasing the MR signal. The modeled MRGAs are rectangular pulse-like signals. We used an event-synchronous adaptive digital filter whose reference signal is synchronous to the gradient peaks of MRI. The event detection processor for the event-synchronous adaptive digital filter was implemented using the phase space method—a sort of topology mapping method—and least-squares acceleration filter. For evaluating the efficiency of the proposed method, the filter was tested using simulation and actual data. The proposed method requires a simple experimental setup that does not require extra hardware connections to obtain the reference signals of adaptive digital filter. The proposed algorithm was more effective than the multichannel approach.     

Application of a wavelet adaptive filter to minimise distortion of the ST-segment

A wavelet adaptive filter (WAF) for the removal of baseline wandering in ECG signals is described. The WAF consists of two parts. The first part is a wavelet transform that decomposes the ECG signal into seven frequency bands using Vaidyanathan-Hoang wavelets. The second part is an adaptive filter that uses the signal of the seventh lowest-frequency band among the wavelet transformed signals as primary input and a constant as reference input. To evaluate the performance of the WAF, two baseline wandering elimination filters are used, a commercial standard filter with a cutoff frequency of 0.5 Hz and a general adaptive filter. The MIT/BIH database and the European ST-T database are used for the evaluation. The WAF performs better in the average power of eliminated noise than the standard filter and adaptive filter. Furthermore, it shows a lower ST-segment distortion than the standard filter and the adaptive filter.     

High-pass filtering of the electrogastrogram

The recording and processing of an electrogastrogram require adequate bandpass filtering, to suppress unwanted artefacts but preserve the original signal waveform. High-pass filtering of various types of different time constants τ, filter order, analogue and digital implementation, have been used to obtain higher baseline stability and faster signal recovery after strong artefacts. Special attention should be given to possible signal amplitude and phase distortions due to high-pass filtering, which can strongly influence accurate amplitude measurements or studies of signal propagation from multichannel recordings. Synthesised and original signals are used to demonstrate the effect of high-pass filtering. The use of a first-order filter with τ=5 s is recommended for EGG studies if not especially directed to investigation of bradygastria. In the opposite case, τ=15 s should be used, and with backward filtering a full restoration of the original signal can be obtained. The same is valid for recording the electrical activity of the colon. Lower time constants (τ=5 s or less) can be applied to acquire signals from the small intestines. A radical solution is the use of a DC amplifier with controllable subtraction of the DC component.     

Cardiac and Respiratory MRI Gating Using Combined Wavelet Sub-Band Decomposition and Adaptive Filtering

Cardiac Magnetic Resonance Imaging (MRI) requires synchronization to overcome motion related artifacts caused by the heart’s contractions and the chest wall movements during respiration. Achieving good image quality necessitates combining cardiac and respiratory gating to produce, in real time, a trigger signal that sets off the consecutive image acquisitions. This guarantees that the data collection always starts at the same point of the cardiac cycle during the exhalation phase. In this paper, we present a real time algorithm for extracting a cardiac-respiratory trigger signal using only one, adequately placed, ECG sensor. First, an off-line calculation phase, based on wavelet decomposition, is run to compute an optimal QRS filter. This filter is used, afterwards, to accomplish R peak detection, while a low pass filtering process allows the retrieval of the respiration cycle. The algorithm’s synchronization capabilities were assessed during mice cardiac MRI sessions employing three different imaging sequences, and three specific wavelet functions. The prominent image enhancement gave a good proof of correct triggering. QRS detection was almost flawless for all signals. As for the respiration cycle retrieval it was evaluated on contaminated simulated signals, which were artificially modulated to imitate respiration. The results were quite satisfactory.     

Development of a computerized EEG imaging system with a personal computer

The authors developed a computerized electroencephalography imaging system with an IBM PC AT. The EEG signals amplified with a 16 channel EEG machine were digitized at 51.2 Hz (512 samples per epoch). The shifted DC potential and 60Hz artificats were removed by a high pass filter and 60Hz notch filter. A window function consisting of a 10% cosine taper was obtained by weighting the points at either end of the epoch by a cosine bell. A fast Fourier transform was applied to every epoch and the power spectrum estimates were computed in 0.39 Hz steps. The activity estimates for the delta, theta, alpha and beta bands were computed by summimg adjacent values. The outline of the top-down maps was formed from a series of sagittal cuts, then 32 electrodes were placed on the map. A file was created which contained a table of weighting parameters for calculating the interpolated values for every point within the outline. Each weight was in inverse linear proportion to the distance of the pixel to the nearest four electrodes. The map was finally generated with computation of the spectral EEG in each pixel according to the weighting parameter. The functioning of this system was tested with a functional generator and a human subject.     

Artefact cancellation in motor-sensory evoked potentials: two approaches using adaptive filtration and exponential approximation

Adaptive filtering for artefact cancellation in motor-sensory evoked potentials using signals obtained by subtraction methods (double-stimulus, off-nerve and subthreshold) is proposed. This is advantageous as inherent non-linear distortions can be overcome in an easier way by adaptive filtering. Efficiency is assessed with reference signals synthesised by varying the shape and reducing the amplitude of a ‘pure’ evoked potential in the range from 10% to 50%. The experiments show virtually identical shapes of the ‘pure’ and the filtered signal. The time shift between them is insignificant if a causal filter and small number of Widrow coefficients, e.g. N=8, are used. Further, two-exponential artefact approximation is applied with subsequent direct subtraction from the contaminated signal by a specially designed PC-controlled system for data acquisition and processing. For a fast procedure convergence, one-parametric optimisation of the time-constant τ is used, starting with τ=0.5 ms. The results obtained with artefact-corrupted evoked potentials from several subjects prove the efficiency of the approach. It has the substantial advantage of avoiding the need for reference signals. Both methods have advantages compared with other known software techniques.     

Cortical activation by monaural speech sound stimulation demonstrated by positron emission tomography

To investigate how auditory input from each ear contributes to spoken language processing, cortical activation by monaural speech sound stimulation was examined in 12 normal subjects using¹⁵O-labeled water positron emission tomography. Regional cerebral blood flow (rCBF) was measured under four different sound stimulation conditions: (1) silence, (2) white noise, (3) sequential Japanese sentences (“speech”), and (4) Japanese sentences played backward (“reversed speech”), and the results were evaluated by statistical parametric mapping (SPM). Noise induced significant rCBF increase in the contralateral Heschl’s gyrus. Speech and reversed speech stimuli caused significant rCBF increase in the contralateral Heschl’s gyrus and the bilateral superior temporal gyri, with contralateral activation broader than that in the ipsilateral hemisphere. Monaurally input speech sound signals that reach the contralateral Heschl’s gyrus may be processed chiefly and phonologically in the surrounding superior temporal gyrus in the same hemisphere. Comparison of speech activation with reversed speech activation failed to demonstrate a significant difference, which made it difficult to identify the area for lexical and semantic processing.     

Removal of ocular artifacts from electro-encephalogram by adaptive filtering

The electro-encephalogram (EEG) is useful for clinical diagnosts and in biomedical research. EEG signals, however, especially those recorded from frontal channels, often contain strong electro-oculogram (EOG) artifacts produced by eye movements. Existing regression-based methods for removing EOG artifacts require various procedures for preprocessing and calibration that are inconvenient and timeconsuming. The paper describes a method for removing ocular artifacts based on adaptive filtering. The method uses separately recorded vertical EOG and horizontal EOG signals as two reference inputs. Each reference input is first processed by a finite impulse response filter of length M (M=3 in this application) and then subtracted from the original EEG. The method is implemented by a recursive leastsquares algorithm that includes a forgetting factor (λ=0.9999 in this application) to track the non-stationary portion of the EOG signals. Results from experimental data demonstrate that the method is easy to implement and stable, converges fast and is suitable for on-line removal of EOG artifacts. The first three coefficients (up to M=3) were significantly larger than any remaining coefficients.     

Perception based method for the investigation of audiovisual integration of speech

Speech comprehension is significantly improved by visual input on the speaker's mouth movements. Audiovisual integration underlying this phenomenon is often studied in EEG experiments in which the event related brain potential (ERP) elicited by a bimodal stimulus is compared to the sum of ERPs triggered by auditory and visual signals of the same source. However, this method leads to spurious results in time ranges when ERP components common to all these stimulus types are present. A method that aims to filter out such common early anticipatory potentials is data high-pass filtering. In the present study, first, we demonstrated that subtle changes in filter cut-off frequency lead to remarkably different results on the interaction effect so that no reliable conclusion on the spatial distribution of the interaction could be drawn. Second, we suggested a different approach for the investigation of ERP correlates of audiovisual integration: bimodal syllables modified by light temporal asynchrony were presented to subjects and ERPs correlating with the fused and unfused perceptions were compared. We found that components corresponding to both auditory N1 and P2 waves were smaller in case of the fused perception, supporting the view that N1 and P2 generator activities are suppressed during multimodal speech perception. The N1 effect showed a clearly right hemisphere dominance while the effect around the P2 peak was most pronounced on centroparietal electrodes and dominated over the left hemisphere.     

Reduction of noise from magnetoencephalography data

A noise reduction method for magnetoencephalography (MEG) data is proposed. The method is a combination of Kalman filtering and factor analysis. A statespace model for a Kalman filter was constructed using the forward problem in MEG measurement. Factor analysis provide estimations of noise covariances required by the Kalman filter to eliminate independent additive sensor noise. The proposed method supports independent component analysis (ICA), which is difficult to use in MEG analysis owing to the sensor noise. Numerical experiments were conducted to investigate the performance of the proposed method. In a single dipole case where the maximum signal-to-noise ratio (SNR) was — 10 dB, approximately equivalent to raw MEG data, noise-free signals were successfully estimated from noisy data; a 0.02 s delay of the peak latency and 15–40% of attenuation of the peak amplitude were observed. Moreover, in a multiple dipole case, independent components preprocessed with the proposed method had high correlation, 0.88 at the lowest, with correlation of 0.69 and 0.52 for those preprocessed with conventional bandpass filters. The results show that the noise reduction method reduces sensor noise effectively. High SNR-independent components are obtained by the proposed method. Real MEG data analysis was also demonstrated. The proposed method extracted auditory evoked responses from unaveraged single-trial data.     

Adaptive filters for analysis of intra-cardiac signals

Implantable cardioverters and defibrillators must discriminate malignant ventricular arrhythmias from other supraventricular tachycardias. Electrogram signals are sensed by a single floating endocardial catheter with multiple sensors. The sensors acquire composite atrial and ventricular complexes from atrial and ventricular chambers. Two adaptive filters are utilised to discriminate atrial signals from ventricular signals. The adaptive impulse correlated filter utilises an impulse sequence correlated with the ventricular depolarisations in order to filter the composite signal. This filter should be useful in the analysis of atrial arrhythmias. The adaptive series feedback filter utilises two back-to-back coupled filters to simultaneously extract ventricular depolarisations from one input channel and atrial depolarisations from another channel. This filter should be useful in the analysis of ventricular arrhythmias. Theoretical analysis of the adaptation capabilities of these filters and criteria for convergence are presented. Experimental electrogram recordings are analysed to demonstrate the performance of the two filters. Computational simplicity of these filters makes them particularly suitable for programmable implantable devices.     

Noise and selectivity of velocity-selective multi-electrode nerve cuffs

Using a multi-electrode nerve-signal recording cuff and a method of signal processing described previously, activity in axons with different propagation velocities can be distinguished, and the relative amplitude of the small-fibre signals increased. This paper is, largely, an analysis of the selectivity and noise of this system though impedance measurements from an actual cuff are included. The signal processor includes narrow band-pass filters. It is shown that the selectivity and noise both increase with the centre frequencies of these filters. A convenient approach is to make the filter frequencies inversely related to the artificial time delays so that the filter ‘Q’s are approximately constant and the noise densities are equal for all velocity filters. Numerical calculations, using formulae for this system and for the conventional tripole, based on a fixed cuff size and tissue resistivity, find the number of action potentials per second that must pass through the cuff so that the signal power equals the noise power. For slow fibres (20 m/s), the rate is 14 times lower for the multi-electrode cuff than the tripole, a significant advantage for recording from these fibres.     

Simple system for analog data transmission from the physiological research laboratory to a digital computer

An unconventional analog-to-digital converter and multiplexer is described. The instrument is based on conversion of the analog signals to time intervals between narrow-width pulses by an encoder unit. The coded analog signals may then be transferred on a single line to a decoder unit and/or stored on a single track of an AM magnetic tape recorder for later processing. The signals are digitized in the decoder unit by measuring the time intervals between the pulses by means of a binary counter driven by a 19.8-MHz oscillator. The decoder is directly communicating with the interface of a digital computer, and a software package has been written for the control of data transmission. The performance of the instrument in our version is: eight separate analog inputs, sampling rate 200 Hz, input voltage −1.6 to 1.6 V, resolution capacity 3.2 V/2¹³, noise ≈0.1%, signal bandwidth de to 50 Hz ac.     

Formation of synchronous pulses from an electrocardiogram

Conclusions In forming synchronous pulses from an EKG it is advisable to use a linear (optimal filter for the R-wave. We recommend using an oscillation stage for the optimal filter with resonant frequencyf ₀ = 1/2τ_u, where τ_u is the length of the R-wave. The selective properties of the filtering stage on the basis of the length of the input signal are shown. Special Design and Construction Office for Electronic Medical Instruments, L'vov. Translated from Meditsinskaya Tekhnika, No.4, pp. 29–33, July–August, 1970.     

提升小波和平滑滤波在心电信号快速滤波中的研究

利用提升小波将原始心电信号(ECG)按频率特性分解为低频的逼近信号和高频的细节信号,舍弃前几层细节信号,采用自适应平滑滤波器选择合适的阈值对逼近信号进行滤波,从而避免平滑滤波对QRS波的损伤,再用提升小波逆变换重建保留的逼近信号,进而实现对心电信号中三种主要噪声的抑制.实验表明该方法计算量小,实时性强.     

Automatic correction for drift in an integrator for phasic signals

An electronic integrator for phasic signals has been designed that does not undergo slow changes in output voltage (drift). Output is stabilised by means of negative feedback using an active 5-pole lowpass filter. Accurate integration of phasic signals with frequencies above 0·1 Hz is obtained. The integrator has been used for continuous determination of cyclincal volume changes, with a pneumotachograph as an input source, in the plotting of respiratory pressure-volume loops.     

Detection of brainstem auditory evoked potential by adaptive filtering

A method of detecting brainstem auditory evoked potential (BAEP) using adaptive signal enhancement (ASE) is proposed and tested in humans and cats. The ASE in this system estimates the signal component of the primary input, which is correlated with the reference input to the adaptive filter. The reference input is carefully designed to make an optimal and rapid estimation of the signal corrupted with noise, such as ongoing EEG. With a good choice of reference input, it is possible to track the variability of BAEP efficiently and rapidly. Moreover, the number of repetitions required could be markedly reduced and the result of the system is superior to that of ensemble averaging (EA). To detect BAEP in cats, only 30 ensemble averages are needed to obtain a reasonable reference input to the adaptive filter, and, for humans, 350–750 ensemble averages are sufficient for a satisfactory result. Using the LMS adaptive algorithm, individual BAEP can be obtained in real-time.     

A chaos-based visual encryption mechanism for clinical EEG signals

In this study, we have developed a chaos-based visual encryption mechanism that can be applied for clinical electroencephalography (EEG) signals. In comparison with other types of random sequences, chaos sequences were mainly used to increase unpredictability. We used a 1D chaotic scrambler and a permutation scheme to achieve EEG visual encryption. One approach of realizing the visual encryption mechanism is to scramble the signal values of the input EEG signal by multiplying a 1D chaotic signal to randomize the EEG signal values. We then applied a chaotic address scanning order encryption to the randomized reference values. Simulation results show that when the correct deciphering parameters are entered, the signal is completely recovered, and the percent root-mean-square difference (PRD) values for control and alcoholic clinical EEG signals are 4.33 × 10⁻¹⁵ and 4.11 × 10⁻¹⁵%, respectively. As long as there is an input parameter error, with an initial point error of 0.00000001% as an example, thereby making these clinical EEG signals unrecoverable.     

Segmentation and tracking of the electro-encephalogram signal using an adaptive recursive bandpass filter

An adaptive filtering approach for the segmentation and tracking of electro-encephalogram (EEG) signal waves is described. In this approach, an adaptive recursive bandpass filter is employed for estimating and tracking the centre frequency associated with each EEG wave. The main advantage inherent in the approach is that the employed adaptive filter has only one unknown coefficient to be updated. This coefficient, having an absolute value less than 1, represents an efficient distinct feature for each EEG specific wave, and its time function reflects the non-stationarity behaviour of the EEG signal. Therefore the proposed approach is simple and accurarate in comparison with existing multivariate adaptive approaches. The approach is examined using extensive computer simulations. It is applied to computer-generated EEG signals composed of different waves. The adaptive filter coefficient (i.e. the segmentation parameter) is −0.492 for the delta wave, −0.360 for the theta wave, −0.191 for the alpha wave, −0.027 for the sigma wave, 0.138 for the beta wave and 0.605 for the gamma wave. This implies that the segmentation parameter increases with the increase in the centre frequency of the EEG waves, which provides fast on-line information about the behaviour of the EEG signal. The approach is also applied to real-world EEG data for the detection of sleep spindles.     

Prediction of blood glucose level of type 1 diabetics using response surface methodology and data mining

In order to improve the accuracy of predicting blood glucose levels, it is necessary to obtain details about the lifestyle and to optimize the input variables dependent on diabetics. In this study, using four subjects who are type 1 diabetics, the fasting blood glucose level (FBG), metabolic rate, food intake, and physical condition are recorded for more than 5 months as a preliminary study. Then, using data mining, an estimation model of FBG is obtained, and subsequently, the trend in fluctuations in the next morning’s glucose level is predicted. The subject’s physical condition is self-assessed on a scale from positive (1) to negative (5), and the values are set as the physical condition variable. By adding the physical condition variable to the input variables for the data mining, the accuracy of the FBG prediction is improved. In order to determine more appropriate input variables from the biological information reflecting on the subject’s glucose metabolism, response surface methodology (RSM) is employed. As a result, using the variables exhibiting positive correlations with the FBG in the RSM, the accuracy of the FBG prediction improved. Conditions could be found such that the accuracy of the predicting trends in fluctuations in blood glucose level reached around 80%. The prediction method of the trend in fluctuations in the next morning’s glucose levels might be useful to improve the quality of life of type 1 diabetics through insulin treatment, and to prevent hypoglycemia.