Showing papers in "IEEE Transactions on Biomedical Engineering in 1992"
TL;DR: The authors present general descriptive models for spatiotemporal MEG (magnetoencephalogram) data and show the separability of the linear moment parameters and nonlinear location parameters in the MEG problem and present a subspace methodology and computational approach to solving the conventional least-squares problem.
Abstract: The authors present general descriptive models for spatiotemporal MEG (magnetoencephalogram) data and show the separability of the linear moment parameters and nonlinear location parameters in the MEG problem. A forward model with current dipoles in a spherically symmetric conductor is used as an example: however, other more advanced MEG models, as well as many EEG (electroencephalogram) models, can also be formulated in a similar linear algebra framework. A subspace methodology and computational approach to solving the conventional least-squares problem is presented. A new scanning approach, equivalent to the statistical MUSIC method, is also developed. This subspace method scans three-dimensional space with a one-dipole model, making it computationally feasible to scan the complete head volume. >
1,080 citations
TL;DR: Computerized sampling of the system output and subsequent calculation of the approximate inverse (sound-to-image) mapping provided the first convincing experimental evidence for the preservation of visual information in sound representations of complicated images.
Abstract: An experimental system for the conversion of images into sound patterns was designed to provide auditory image representations within some of the known limitations of the human hearing systems possibly as a step towards the development of a vision substitution device for the blind. The application of an invertible (one-to-one) image-to-sound mapping ensures the preservation of visual information. The system implementation involves a pipelined special-purpose computer connected to a standard television camera. A novel design and the use of standard components have made for a low-cost portable prototype conversion system with a power dissipation suitable for battery operation. Computerized sampling of the system output and subsequent calculation of the approximate inverse (sound-to-image) mapping provided the first convincing experimental evidence for the preservation of visual information in sound representations of complicated images. >
822 citations
TL;DR: It was concluded that silicon microshafts of appropriate shaft and tip design were biocompatible along the sides of the shaft, but that relatively severe reactions could be anticipated at the tips.
Abstract: Issues that determine the biocompatibility of insertable microelectrode arrays were investigated. Results from a limited number of tests indicated that there was minimal tissue response along the sides of the shafts when shafts were well sharpened, had sufficiently small tip angles, and were clean. Tissue was usually more reactive at the tips of the shafts. It was concluded that silicon microshafts of appropriate shaft and tip design were biocompatible along the sides of the shaft, but that relatively severe reactions could be anticipated at the tips. >
501 citations
TL;DR: It is proposed that limits on levels of electrical stimulation take into account the location of the electrode relative to the stimulated tissue; these limits can be computed algorithmically from the model.
Abstract: A model that represents a large body of data on safety and damage levels of electrical stimulation is presented. The predictions of the model are consistent with known principals of current flow and known mechanisms of damage around stimulating electrodes. It is proposed that limits on levels of electrical stimulation take into account the location of the electrode relative to the stimulated tissue; these limits can be computed algorithmically from the model. >
495 citations
TL;DR: The authors have developed an adaptive matched filtering algorithm based upon an artificial neural network (ANN) for QRS detection that is very effective at removing the time-varying, nonlinear noise characteristic of ECG signals.
Abstract: The authors have developed an adaptive matched filtering algorithm based upon an artificial neural network (ANN) for QRS detection. They use an ANN adaptive whitening filter to model the lower frequencies of the electrocardiogram (ECG) which are inherently nonlinear and nonstationary. The residual signal which contains mostly higher frequency QRS complex energy is then passed through a linear matched filter to detect the location of the QRS complex. The authors developed an algorithm to adaptively update the matched filter template from the detected QRS complex in the ECG signal itself so that the template can be customized to an individual subject. This ANN whitening filter is very effective at removing the time-varying, nonlinear noise characteristic of ECG signals. The detection rate for a very noisy patient record in the MIT/BIH arrhythmia database is 99.5% with this approach, which compares favorably to the 97.5% obtained using a linear adaptive whitening filter and the 96.5% achieved with a bandpass filtering method. >
467 citations
TL;DR: A method for predicting excitation of axons based on the response of passive models is proposed and a total equivalent driving function including both terms can be used to provide predictions of excitation thresholds for any applied field.
Abstract: A method for predicting excitation of axons based on the response of passive models is proposed. An expression describing the transmembrane potential induced in passive models to an applied electric field is presented. Two terms drive the polarization of each node: a source term described by the activating function at the node, and an ohmic term resulting from redistribution of current from sources at other nodes. It is shown that a total equivalent driving function including both terms can be used to provide predictions of excitation thresholds for any applied field. The method requires only knowledge of the intracellular strength-duration relationship of the axon, the passive step response of the axon to an intracellular current, and the values of the extracellular potentials. Excitation thresholds for any given applied field can then be calculated using a simple algebraic expression. This method eliminates the errors associated with use of the activating function alone, and greatly reduces the computation required. >
350 citations
TL;DR: In model simulations of cortical activity of the human brain, the magnetic field pattern across the scalp is interpreted with prior knowledge of anatomy to yield a unique magnetic source image across a portion of cerebral cortex, without resort to an explicit source model.
Abstract: The minimum norm least-squares approach based on lead field theory provides a unique inverse solution for a magnetic source image that is the best estimate in the least-squares sense. This has been applied to determine the source current distribution when the primary current is confined to a surface or set of surfaces. In model simulations of cortical activity of the human brain, the magnetic field pattern across the scalp is interpreted with prior knowledge of anatomy to yield a unique magnetic source image across a portion of cerebral cortex, without resort to an explicit source model. >
300 citations
TL;DR: In this paper, the use of 2450-MHz microwave energy applied via a miniature coaxial cable-mounted helical coil antenna was investigated as a means to increase the treated volume of cardiac tissue in a saline-perfused, tissue equivalent manner during ablation.
Abstract: The use of 2450-MHz microwave energy applied via a miniature coaxial cable-mounted helical coil antenna was investigated as a means to increase the treated volume of cardiac tissue in a saline-perfused, tissue-equivalent manner during ablation. Using an array of fiber-optic temperature probes implanted in a saline-perfused, tissue-equivalent gel phantom model designed to simulate the myocardium during ablation, the heating pattern from the microwave antenna was characterized and compared to that induced by a commercial RF electrode catheter at 550 kHz. Effects of variable contact angle between the heat source and heart wall were assessed. Heating patterns from the RF electrodes dropped off much more abruptly both radially and axially than the microwave antenna. The volume of effectively heated tissue was more than ten times larger for the microwave antenna when the heat sources were well-coupled to the tissue, and more than four times larger for the microwave antenna when the sources were angled 30 degrees away from the tissue surface. >
224 citations
TL;DR: The development of a high-Q approximation, which simplifies the design procedure is presented, and the closed-loop class E circuit shows great promise, especially for circuits with unusually low coefficients of coupling.
Abstract: The use of a multifrequency transmitter coil driver based on the class E topology is described. The development of a high-Q approximation, which simplifies the design procedure is presented. A closed-loop controller to compensate for transmitter and receiver variations, and a method of data modulation using synchronous frequency shifting are described. The closed-loop class E circuit shows great promise, especially for circuits with unusually low coefficients of coupling. Currents of several amperes, at radio frequencies, can easily and efficiently be obtained. >
213 citations
TL;DR: A microelectrode array capable of recording from and stimulating peripheral nerves at prolonged intervals after surgical implantation has been demonstrated.
Abstract: A microelectrode array capable of recording from and stimulating peripheral nerves at prolonged intervals after surgical implantation has been demonstrated. The microelectrode array, fabricated on a silicon substrate perforated by multiple holes (referred to as via holes), is implanted between the ends of a surgically severed nerve. Regenerating tissue fixes the device in place to provide a stable mapping between the microelectrodes and the axons in the nerve. Processes were developed for the fabrication of thin-film iridium microelectrodes, micromachined via holes, and silicon nitride passivation layers. All fabrication methods were designed to be compatible with standard CMOS/BiCMOS processes to allow for on-chip signal processing circuits in future designs. Such arrays, implanted in the peroneal nerves of rats, were used to record from and stimulate the nerves at up to 13 months postoperatively. >
202 citations
TL;DR: It was found that transmembrane potentials, due to external monopolar stimulation, at the node at which a collateral is attached, is significantly influenced by the presence of the collateral.
Abstract: An electrical network model of myelinated dorsal column nerve fibers is presented. The effect of electrical simulation was investigated using both a homogeneous volume conductor and a more realistic model of the spinal cord. An important feature of dorsal column nerve fibers is the presence of myelinated collaterals perpendicular to the rostro-caudal fibers. It was found that transmembrane potentials, due to external monopolar stimulation, at the node at which a collateral is attached, is significantly influenced by the presence of the collateral. It is concluded that both excitation threshold and blocking threshold of dorsal column fibers are decreased up to 50% compared to unbranched fibers. >
TL;DR: In this article, a method is presented to compute the potential distribution on the surface of a homogeneous isolated conductor of arbitrary shape based on an approximation of a boundary integral equation as a set of linear algebraic equations.
Abstract: A method is presented to compute the potential distribution on the surface of a homogeneous isolated conductor of arbitrary shape. The method is based on an approximation of a boundary integral equation as a set of linear algebraic equations. The potential is described as a piecewise linear or quadratic function. The matrix elements of the discretized equation are expressed as analytical formulas. >
TL;DR: The development and testing of a noninvasive true phase optical polarimetry sensing system to monitor in vivo glucose concentrations is described and the applicability of this optical sensor for glucose movement is demonstrated.
Abstract: The development and testing of a noninvasive true phase optical polarimetry sensing system to monitor in vivo glucose concentrations is described. To demonstrate the applicability of this optical sensor for glucose movement, the authors calibrate the system and then test it in vitro using both a glass test cell filled with glucose solution in the physiologic range, with a path length of 0.9 cm to approximate the 1-cm path length present in the anterior chamber of the eye, and then on an excised human eye. The technique used helium neon laser light which was coupled through a rotating linear polarizer along with two stationary linear polarizers and two detectors to produce reference and signal outputs whose amplitudes varied sinusoidally with a frequency of twice the angular velocity of the rotating polarizer, and whose phase was proportional to the rotation of the linear polarization vector passing through the glucose solution. >
TL;DR: A multiple-model adaptive predictive controller has been designed to simultaneously regulate mean arterial pressure and cardiac output in congestive heart failure subjects by adjusting the infusion rates of nitroprusside and dopamine.
Abstract: A multiple-model adaptive predictive controller has been designed to simultaneously regulate mean arterial pressure and cardiac output in congestive heart failure subjects by adjusting the infusion rates of nitroprusside and dopamine. The algorithm is based on the multiple-model adaptive controller and utilizes model predictive controllers to provide reliable control in each model subspace. A total of 36 linear small-signal models were needed to span the entire space of anticipated responses. To reduce computation time, only the six models with the highest probabilities were used in the control calculations. The controller was evaluated on laboratory animals that were either surgically or pharmacologically altered to exhibit symptoms of congestive heart failure. During trials, the controller performance was robust with respect to excessive switching between models and nonconvergence to a single dominant model. A comparison with a previous multiple-drug controller design is made. >
TL;DR: Computer simulations demonstrate the ability of the Laplacian map compared with the potential map to image spatially distributed dipole sources embedded in a semi-infinite volume conductor.
Abstract: A noninvasive approach has been developed to resolve spatially distributed cardiac electrical activity by measuring the surface Laplacian of the body surface potential. Computer simulations demonstrate the ability of the Laplacian map compared with the potential map to image spatially distributed dipole sources embedded in a semi-infinite volume conductor. Body surface Laplacian mapping has been implemented in human subjects utilizing dry bipolar Laplacian electrodes and was compared with potential maps obtained using the central terminal of each bipolar Laplacian electrode. The body surface Laplacian map appears to resolve depolarization and repolarization of different regions of the heart. Further improvements of the body surface Laplacian mapping may permit noninvasive mapping of spatially distributed intracardiac events. >
TL;DR: One important conclusion of this study is that the assumption that the Doppler spectral power density at a frequency in the Dooppler spectrum is linearly proportional to the number of red cells flowing at that velocity used in many theoretical models developed to explain the doppler phenomenon may be erroneous.
Abstract: Previous studies of ultrasonic scattering properties of blood using a pulse-echo experimental arrangement show that ultrasonic backscatter from blood is influenced by a number of factors, including hematocrit, shear rate, and the nature of flow. Since the Doppler frequency spectrum from a Doppler flowmeter is derived from echoes backscattered by red blood cells in the flowing blood, it is also undoubtedly a function of these parameters. The effects of these parameters on Doppler spectrum from blood have been investigated using a pulsed Doppler flowmeter. The results agree well with those obtained in previous studies. One important conclusion of this study is that the assumption that the Doppler spectral power density at a frequency in the Doppler spectrum is linearly proportional to the number of red cells flowing at that velocity used in many theoretical models developed to explain the Doppler phenomenon may be erroneous. An alternative is proposed. It is shown that conclusions derived from these theoretical models would remain valid by making this assumption. >
TL;DR: In this paper, an adaptive impulse correlated filter (AICF) was proposed to estimate the deterministic component of the signal and remove the noise uncorrelated with the stimulus even if this noise is colored, as in the case of evoked potentials.
Abstract: An adaptive impulse correlated filter (AICF) for event-related signals that are time-locked to a stimulus is presented. This filter estimates the deterministic component of the signal and removes the noise uncorrelated with the stimulus, even if this noise is colored, as in the case of evoked potentials. The filter needs two inputs: the signal (primary input) and an impulse correlated with the deterministic component (reference input). The LMS algorithm is used to adjust the weights in the adaptive process. It is shown that the AICF is equivalent to exponentially weighted averaging (FWA) when using the LMS algorithm. A quantitative analysis of the signal-to-noise ratio improvement, convergence, and misadjustment error is presented. A comparison of the AICF with ensemble averaging (EA) and moving window averaging (MWA) techniques is also presented. The adaptive filter is applied to real high-resolution ECG signals and time-varying somatosensory evoked potentials. >
TL;DR: It was shown that the Doppler signal is a Gaussian random process that arises from fluctuation scattering, which implies that the backscattered power is proportional to the variance of local RBC concentrations, which means some common misconceptions about the relationship between theBackscattering coefficient and hematocrit can be readily resolved.
Abstract: A unified approach to modeling the backscattered Doppler ultrasound signal from blood is presented. The approach consists of summing the contributions from elemental acoustic voxels, each containing many red blood cells (RBCs). For an insonified region that is large compared to a wavelength, it is shown that the Doppler signal is a Gaussian random process that arises from fluctuation scattering, which implies that the backscattered power is proportional to the variance of local RBC concentrations. As a result, some common misconceptions about the relationship between the backscattering coefficient and hematocrit can be readily resolved. The unified approach was also used to derive a Doppler signal simulation model which shows that, regardless of flow condition, the power in the Doppler frequency spectrum is governed by the exponential distribution. For finite beamwidth and paraxial flow, it is further shown that the digitized Doppler signal can be modeled by a moving average random process whose order is determined by the signal sampling rate as well as the flow velocity profile. >
TL;DR: A flexible and perforated 32-element planar microelectrode array has been fabricated and used to measure evoked potentials in brain slices and showed an average increase of 10 h to the viability of the slice while using the perforations versus nonperforated arrays.
Abstract: A flexible and perforated 32-element planar microelectrode array has been fabricated and used to measure evoked potentials in brain slices. Electrodes are spaced 200 mu m apart in a 4*8 array and are sandwiched between layers of insulating polyimide. The polyimide sandwich is lifted off its substrate, making it flexible so that it could be shaped to contoured tissues. Prior to liftoff, holes are etched to expose recording sites 15 mu m in diameter and to create perforations which allow increased circulation of artificial cerebrospinal fluid to the recording surface of the tissue and, hence, increased viability. Comparisons of evoked potentials measured over tie showed an average increase of 10 h to the viability of the slice while using the perforated versus nonperforated arrays. >
TL;DR: A detailed discussion of several ways to measure the ability of anImpedance imaging system to distinguish between two different conductivity distributions is given.
Abstract: Impedance imaging systems apply currents to the surface of a body, measure the induced voltages on the surface, and from this information reconstruct an approximation to the electrical conductivity in the interior. A detailed discussion of several ways to measure the ability of such a system to distinguish between two different conductivity distributions is given. The subtle differences between these related measures are discussed, and examples are provided to show that these different measures can give rise to different answers to various practical questions about system design. >
TL;DR: Simulation studies show that the theory underlying a new method for the identification of time-varying systems can track rapid changes in system dynamics accurately and is robust in the presence of output noise.
Abstract: The theory underlying a new method for the identification of time-varying systems is described. The method uses singular value decomposition to obtain least-squares estimates of time-varying impulse response functions from an ensemble of input-output realizations. No a priori assumptions regarding the system structure or form of the time-variation are required and there are few restrictions on the input signal. Simulation studies, using a model of time-varying joint dynamics, show that the method can track rapid changes in system dynamics accurately and is robust in the presence of output noise. An application of the method is demonstrated by using it to track dynamic ankle stiffness during a rapid, voluntary, isometric contraction. During the transient phase of the contraction, low-frequency ankle stiffness gain decreased in a manner which could not be described with the second-order model of joint dynamics often used under stationary conditions. >
TL;DR: In this article, the inverse problem of electrocardiography is solved in order to reconstruct electrical events within the heart from information measured noninvasively on the body surface; therefore, a noninvasive method of recovering epicardial potentials from body surface data is useful in clinical and experimental work.
Abstract: The inverse problem of electrocardiography is solved in order to reconstruct electrical events within the heart from information measured noninvasively on the body surface. These electrical events can be deduced from measured epicardial potentials; therefore, a noninvasive method of recovering epicardial potentials from body surface data is useful in clinical and experimental work. The ill-posed nature of this problem necessitates the use of regularization in the solution procedure. Inversion using Tihonov zero-order regularization, a quasi-static method, had been employed previously and was able to reconstruct, with relatively good accuracy, important events in cardiac excitation (maxima, minima, etc.). Taking advantage of the fact that the process of cardiac excitation is continuous in time, one can incorporate information from the time progression of excitation in the regularization procedure using the Twomey technique. Methods of this type were tested on data obtained from a human-torso tank in which a beating canine heart was placed. The results show a marked improvement in the inverse solution when these temporal methods are used. >
TL;DR: Designs of both open- and closed-loop controllers of electrically stimulated muscle that explicitly depend on a nonlinear mathematical model of muscle input-output properties are presented and evaluated and an open-loop nonlinear compensator was tested in an animal model.
Abstract: Designs of both open- and closed-loop controllers of electrically stimulated muscle that explicitly depend on a nonlinear mathematical model of muscle input-output properties are presented and evaluated. The muscle model consists of three factors: a muscle activation dynamics factor, an angle-torque relationship factor, and an angular velocity torque relationship factor. These factors are multiplied to relate output torque to input simulation and joint angle. An experimental method for the determination of the parameters of this model was designed, implemented, and evaluated. An open-loop nonlinear compensator, based upon this model, was tested in an animal model. Its performance in the control of joint angle in the presence of a known load was compared with a PID (proportional-integral-derivative) controller, and with a combination of the PID controller and the nonlinear compensator. The results are presented. >
TL;DR: In this paper, the authors derived an expression for the covariance between the magnetic field and a differentially measured EEG signal, which was used to obtain a (spatial) covariance matrix of the background noise, which occurs in evoked potential measurements.
Abstract: The statistical properties of the EEG and the MEG can be described mathematically as the result of randomly distributed dipoles representing the interactions of cortical neurons. If the dipoles are in a spherical volume conductor and have no preference for any direction, the variance of a differentially measured EEG signal is only a function of the electrode distance. The theoretically derived variance function is compared with EEG and MEG measurements. It is shown that a dipole with a fixed position and a randomly fluctuating amplitude is an adequate model for the alpha -rhythm. An expression for the covariance between the magnetic field and a differentially measured EEG signal is derived. This covariance is considered as a function of the magnetometer position. The theory can be used to obtain a (spatial) covariance matrix of the background noise, which occurs in evoked potential measurements. Such a covariance matrix can be used to obtain a maximum likelihood estimator (MLE) of the dipole parameters in evoked potential studies. >
TL;DR: The binomial joint time-frequency transform is used to test the hypothesis that first heart sound frequency rises during the isovolumic contraction period and provides much better resolution than the spectrograph or spectrogram.
Abstract: The binomial joint time-frequency transform is used to test the hypothesis that first heart sound frequency rises during the isovolumic contraction period. Cardiac vibrations were recorded from eight open-chest dogs using an ultralight accelerometer cemented directly to the epicardium of the anterior left ventricle. Three characteristic time-frequency spectral patterns were evident in the animals investigated: (1) a frequency component that rose from approximately 40-140 Hz in a 30-50-ms interval immediately following the ECG R-wave, (2) a slowly varying or static frequency of 60-100 Hz beginning midway through the isovolumic contraction period, and (3) broadband peaks occurring at the time of the Ia and Ib high frequency components. The binomial transform provided much better resolution than the spectrograph or spectrogram. By revealing the onset and dynamics of first heart sound frequencies, time-frequency transforms may allow mechanical assessment of individual cardiac structures. >
TL;DR: The authors present a feature-based detection approach using neural networks that provides good agreement with visual K-complex recognition: a sensitivity of 90% is obtained with about 8% false positives.
Abstract: The main difficulties in reliable automated detection of the K-complex wave in EEG are its close similarity to other waves and the lack of specific characterization criteria. The authors present a feature-based detection approach using neural networks that provides good agreement with visual K-complex recognition: a sensitivity of 90% is obtained with about 8% false positives. The respective contribution of the features and that of the neural network is demonstrated by comparing the results to those obtained with (i) raw EEG data presented to neural networks, and (ii) features presented to Fisher's linear discriminant. >
TL;DR: A fuzzy control system to provide closed-loop control of mean arterial pressure (MAP) in postsurgical patients in a cardiac surgical intensive care unit setting by regulating sodium nitroprusside (SNP) infusion is discussed.
Abstract: A fuzzy control system to provide closed-loop control of mean arterial pressure (MAP) in postsurgical patients in a cardiac surgical intensive care unit setting by regulating sodium nitroprusside (SNP) infusion is discussed. The fuzzy controller, originally expert-system-based, was analytically converted to ten nonfuzzy control algorithms, which reduced execution time dramatically. The core of the control algorithms was a nonlinear proportional-integral (PI) controller whose proportional gain and integral gain adjusted continuously according to error and rate change of error of the process output. The gains become larger when process output was far from desired setpoint and smaller when process output was close to desired setpoint, resulting in more dynamic and stable control performance than the regular PI controller, especially when a linear process with time-delay or a nonlinear process was involved. The control algorithms, encoded in C programming language, were implemented to control MAP in patients. Preliminary clinical results showed that the average percentage of time in which MAP stayed between 90% and 110% of the MAP setpoint was 89.31%, with a standard deviation of 4.96%. These were calculated based on 12 patient trials, with total trial time of 95 and 13 min. >
TL;DR: A family of pseudorandom signals which are especially suited for the estimation of Z and gamma /sup 2/ in mechanical measurement of physiological systems at low frequencies is proposed and is supported by simulation results and experimental examples.
Abstract: For pseudorandom (PRN) input stimuli, general expressions are derived for the apparent transfer (Z) and coherence ( gamma /sup 2/) functions of nonlinear systems that can be represented by a Volterra series. To avoid the problems that are shown here to be associated with harmonic distortions and to minimize the influence of crosstalk, a family of pseudorandom signals which are especially suited for the estimation of Z and gamma /sup 2/ in mechanical measurement of physiological systems at low frequencies is proposed. The components in the signals cannot be reproduced as linear combinations of two or more frequency components of the input. In a second-order system, this completely eliminates the bias, while in higher order but not strongly nonlinear systems, the interactions among the components are reduced to such a level that the response can be considered as if it were measured with independent sine waves of an equivalent amplitude. It is also shown that the values of gamma /sup 2/ are not appropriate for assessing linearity of the system. The theory is supported by simulation results and experimental examples. >
TL;DR: The simplicity of the solution and a very short computation time make this method particularly attractive for gaining a physical insight into the spatial behavior of the stimulating parameter and for the optimization of coils.
Abstract: Spatial distribution of the derivative of the electric field induced in a planar semi-infinite tissue model by various current-carrying coils and their utility in neural stimulation are evaluated. Analytical expressions are obtained for the electric field and its spatial derivatives produced by an infinitely short current element. Fields and their derivatives for an arbitrarily shaped coil are then obtained by numerical summation of contributions from all the elements forming the coil. The simplicity of the solution and a very short computation time make this method particularly attractive for gaining a physical insight into the spatial behavior of the stimulating parameter and for the optimization of coils. Such analysis is useful as the first step before undertaking a more complex numerical analysis of a model more closely representing the tissue geometry and heterogeneity. >
TL;DR: Weighted averages of brain evoked potentials are obtained by weighting each single EP sweep prior to averaging to maximize the signal-to-noise ratio (SNR) of the resulting average if they satisfy a generalized eigenvalue problem involving the correlation matrices of the underlying signal and noise components.
Abstract: Weighted averages of brain evoked potentials (EPs) are obtained by weighting each single EP sweep prior to averaging. These weights are shown to maximize the signal-to-noise ratio (SNR) of the resulting average if they satisfy a generalized eigenvalue problem involving the correlation matrices of the underlying signal and noise components. The signal and noise correlation matrices are difficult to estimate and the solution of the generalized eigenvalue problem is often computationally impractical for real-time processing. Correspondingly, a number of simplifying assumptions about the signal and noise correlation matrices are made which allow an efficient method of approximating the maximum SNR weights. Experimental results are given using actual auditory EP data which demonstrate that the resulting weighted average has estimated SNRs that are up to 21% greater than the conventional ensemble average SNR. >