Transformation of multichannel magnetocardiographic signals tostandard grid form

Multichannel magnetocardiographic (MCG) recordings with fixed sensor arrays are not directly comparable with single-channel measurements carried out at standard grid locations. In addition, comparison of data obtained with different types of magnetometers is difficult. The authors present a method for transforming multichannel measurements to the standard-grid format. The minimum-norm estimate (MNE) of the source current distribution in the body is calculated, and the desired field components in standard grid points are then computed from the MNE. The authors measured three subjects with both a 24-channel and a single-channel instrument. The signals extrapolated from the multichannel measurements corresponded quite well to the single-channel data registered at the standard grid locations, especially in those grid points that were covered by the 24-channel device. The signal-amplitude-weighted correlations between the extrapolated and directly measured signals were 0.73-0.87. In simulations with ideal measurement geometry but with a realistic amount of random noise in the signals, the authors obtained a 0.99 correlation. It was also found that the method is relatively tolerant to errors in the location and orientation of the multichannel magnetometer. For example, a simulated 20-mm displacement in the location of the sensor array caused only a 3% decrease in the correlation, and when it was rotated and tilted by 10°, the correlation decreased by 5%. The basic advantage of the authors' extrapolation method is its physiologic nature: the method is based on the mathematical modeling of the source current distribution, rather than on direct constraints applied to the magnetic field     

Applications of the Signal Space Separation Method

The reliability of biomagnetic measurements is traditionally challenged by external interference signals, movement artifacts, and comparison problems caused by different positions of the subjects or different sensor configurations. The Signal Space Separation method (SSS) idealizes magnetic multichannel signals by transforming them into device-independent idealized channels representing the measured data in uncorrelated form. The transformation has separate components for the biomagnetic and external interference signals, and thus, the biomagnetic signals can be reconstructed simply by leaving out the contribution of the external interference. The foundation of SSS is a basis spanning all multichannel signals of magnetic origin. It is based on Maxwell's equations and the geometry of the sensor array only, with the assumption that the sensors are located in a current free volume. SSS is demonstrated to provide suppression of external interference signals, standardization of different positions of the subject, standardization of different sensor configurations, compensation for distortions caused by movement of the subject (even a subject containing magnetic impurities), suppression of sporadic sensor artifacts, a tool for fine calibration of the device, extraction of biomagnetic DC fields, and an aid for realizing an active compensation system. Thus, SSS removes many limitations of traditional biomagnetic measurements.     

Parametric sensor array calibration using measured steering vectorsof uncertain locations

We consider the problem of sensor array calibration using a set of unique measured steering vectors of uncertain locations to estimate the unknown deterministic array perturbation parameters in a maximum likelihood framework. The array perturbations are parameterized by the sensor locations, mutual coupling coefficients, and receiver channel mismatch. We introduce a hybrid optimizer based on the amalgamation of gradient-based algorithms and the genetic algorithm. This optimizer is capable of coping with the problem of local extrema attractors, particularly initial estimates with large deviations from their true values. Numerical examples are provided to demonstrate the effectiveness and behavior of the proposed algorithms     

Monitoring of thermo-mechanical stress via CMOS sensor array: Effects of warpage and tilt in flip chip thermo-compression bonding

Flip chip thermo-compression bonding (TCB) involves the use of heat and pressure to simultaneously form interconnections for microelectronic packaging. In-situ measurements of thermo-mechanical stresses that arise during this bonding process could provide unique insight to help better understand the TCB process. A 4 mm × 3 mm × 500 μm complementary metal-oxide-semiconductor (CMOS) sensor chip with an 8 × 8 array of Au-bumped sensor pads was developed for this purpose. It was designed to record XYZ force and temperature signals from bump locations, during a simulated flip chip process similar to TCB.In-situ measurements during simulated TCB events proved useful for tilt detection, thermal gradient characterization, and thermal expansion measurements. Further interpretation of the signals proved tilt and other thermo-mechanical effects were induced by thermal expansion mismatches. The most thermo-mechanically stressful stage of bonding was found to occur during thermal transients, specifically during bond head ramping. Further analysis concluded the actual time necessary to heat the bumps was less than 0.5 s. Finally, the lateral thermal gradient across the sensor chip was calculated to be smallest in the central bump locations, and largest in the bump array corners due to warpage, tilt, and heat sink effects of the digital logic region.     

Direction finding in the presence of mutual coupling

An eigenstructure-based method for direction finding in the presence of sensor mutual coupling, gain, and phase uncertainties is presented. The method provides estimates of the directions-of-arrival (DOA) of all the radiating sources as well as calibration of the gain and phase of each sensor and the mutual coupling in the receiving array. The proposed algorithm is able to calibrate the array parameters without prior knowledge of the array manifold, using only signals of opportunity and avoiding the need for deploying auxiliary sources at known locations. The algorithm is described in detail, and its behavior is illustrated by numerical examples     

Directional ocean wave measurements in a coastal setting using afocused array imaging radar

A unique focused array imaging Doppler radar was used to measure directional spectra of ocean surface waves in a nearshore experiment performed on the North Carolina Outer Banks. Radar images of the ocean surface's Doppler velocity were used to generate two dimensional spectra of the radial component of the ocean surface velocity field. These are compared to simultaneous in-situ measurements made by a nearby array of submerged pressure sensors. Analysis of the resulting two-dimensional spectra include comparisons of dominant wave lengths, wave directions, and wave energy accounting for relative differences in water depth at the measurement locations. Limited estimates of the two-dimensional surface displacement spectrum are derived from the radar data. The radar measurements are analagous to those of interferometric synthetic aperture radars (INSAR), and the equivalent INSAR parameters are shown. The agreement between the remote and in-situ measurements suggests that an imaging Doppler radar is effective for these wave measurements at near grazing incidence angles     

双重不变ESPRIT

ESPRIT利用了阵列在一个方向上的平移不变性,从而可估计信号的一个方向。该文利用3个相互位移的子阵列所具有的双重不变性质,从而估计出了信号的两个方向,并且计算量较小,两个方向角自然配对,对子阵列的内部结构也没有要求,因此可以选择适当的子阵列结构以获得良好的估计效果。     

The effect of geometric and topologic differences in boundary element models on magnetocardiographic localization accuracy

This study was performed to evaluate the changes in magnetocardiographic (MCG) source localization results when the geometry and the topology of the volume conductor model were altered. Boundary element volume conductor models of three patients were first constructed. These so-called reference torso models were then manipulated to mimic various sources of error in the measurement and analysis procedures. Next, equivalent current dipole localizations were calculated from simulated and measured multichannel MCG data. The localizations obtained with the reference models were regarded as the "gold standard." The effect of each modification was investigated by calculating three-dimensional distances from the gold standard localizations to the locations obtained with the modified model. The results show that the effect of the lungs and the intra-ventricular blood masses is significant for deep source locations and, therefore, the torso model should preferably contain internal inhomogeneities. However, superficial sources could be localized within a few millimeters even with nonindividual, so called standard torso models. In addition, the torso model should extend long enough in the pelvic region, and the positions of the lungs and the ventricles inside the model should be known in order to obtain accurate localizations.     

MEG source localization using an MLP with a distributed output representation

We present a system that takes realistic magnetoencephalographic (MEG) signals and localizes a single dipole to reasonable accuracy in real time. At its heart is a multilayer perceptron (MLP) which takes the sensor measurements as inputs, uses one hidden layer, and generates as outputs the amplitudes of receptive fields holding a distributed representation of the dipole location. We trained this Soft-MLP on dipolar sources with real brain noise and converted the network's output into an explicit Cartesian coordinate representation of the dipole location using two different decoding strategies. The proposed Soft-MLPs are much more accurate than previous networks which output source locations in Cartesian coordinates. Hybrid Soft-MLP-start-LM systems, in which the Soft-MLP output initializes Levenberg-Marquardt, retained their accuracy of 0.28 cm with a decrease in computation time from 36 ms to 30 ms. We apply the Soft-MLP localizer to real MEG data separated by a blind source separation algorithm, and compare the Soft-MLP dipole locations to those of a conventional system.     

Inference of Complete Tissue Temperature Fields from a Few Measured Temperatures: An Unconstrained Optimization Method

In clinical applications of hyperthermia, tissue temperature measurements are made at only a few selected locations because of patient tolerance and practical clinical limitations. Since it is necessary to know the complete tumor temperature field in order to effectively evaluate a treatment, methods of interpolating and extrapolating must be developed to estimate the unmeasured tumor temperatures. The difficulty of making such estimates from only a few data points is compounded by a lack of knowledge of the tumor blood perfusion characteristics. To solve this problem we have developed an iterative state and parameter estimation algorithm to attempt to estimate complete tissue temperature fields from temperatures measured at selected locations when tissue perfusion values are unknown. This approach uses either a conjugate gradient or a relaxation method to minimize the difference between the measured temperatures and the temperatures predicted at those same locations by the bioheat transfer equation. To investigate the mathematical capabilities and limitations of this technique a sensitivity analysis has been performed by applying it to a large number of simulated one-dimensional hyperthermia treatments. To illustrate its applicability to clinical situations, the estimation algorithm is also applied to temperature measurements from an animal experiment. The simulation results show that the technique is promising for estimating the complete tumor temperature distribution from measurements at a small number of sampled locations, if some knowledge of the blood perfusion pattern is available.     

Linear dependence of steering vectors associated with tripolearrays

We are concerned with the linear independence of steering vectors associated with tripoles, each of which provides measurements of the three components of electric field induced by electromagnetic signals. We first establish that for a single tripole, any steering vector is linearly dependent on at least one other steering vector corresponding to a different direction-of-arrival (DOA) for a general problem where signals may arrive from anywhere in a three-dimensional (3-D) space, but every two steering vectors with distinct DOAs are linearly independent if the signals are nonlinearly polarized and arrive from a strictly hemispherical space. We then obtain a series of upper bounds for the number of linearly independent steering vectors associated with a tripole array with general sensor configurations. We also show that for applications where signals are known to be linearly polarized in the same direction, the ability to estimate DOAs using a tripole array is identical to that using a scalar-sensor array if both of them have identical sensor configurations     

Spatial evolutionary spectrum for DOA estimation and blind signalseparation

We combine the concepts of evolutionary spectrum and array processing. We present a cross-power stationary periodogram for both direction-of-arrival (DOA) estimation and blind separation of nonstationary signals. We model the nonstationary signals received by each array sensor as a sum of complex sinusoids with time-varying amplitudes. These magnitudes carry information about the DOA that may also be time-varying. We first estimate the time-varying amplitudes using estimators obtained by minimizing the mean-squared error. Then using the estimated time-varying amplitudes, we estimate the evolutionary cross-power distributions of the sensor. Next, using cross-power estimates at time-frequency points interest, we estimate the DOAs using one of the existing methods. If the directions are time varying, we choose time-frequency points around the time of interest to estimate spontaneous source locations. If the sources are stationary, time-frequency points of interest can be combined for the estimation of fixed directions. Whitening and subspace methods used to find the mixing matrix and separate nonstationary signals received by the array. We present examples illustrating the performance of the proposed algorithms     

Noncausal 2-D spectrum estimation for direction finding

A two-dimensional noncausal autoregressive (NCAR) plus additive noise model-based spectrum estimation method is presented for planar array data typical of signals encountered in array processing applications. Since the likelihood function for NCAR plus noise data is nonlinear in the model parameters and is further complicated by the unknown variance of the additive noise, computationally intensive gradient search algorithms are required for computing the estimates. If a doubly periodic lattice is assumed, the complexity of the approximate maximum likelihood (ML) equation is significantly reduced without destroying the theoretical asymptotic properties of the estimates and degrading the observed accuracy of the estimated spectra. Initial conditions for starting the approximate ML computation are suggested. Experimental results that can be used to evaluate the signal-plus-noise approach and compare its performance to those of signal-only methods are presented for Gaussian and simulated planar array data. Statistics of estimated spectrum parameters are given, and estimated spectra for signals with close spatial frequencies are shown. The approximate ML parameter estimate's asymptotic properties, such as consistency and normality, are established, and lower bounds for the estimate's errors are derived, assuming that the data are Gaussian     

Laguerre-model blind system identification: cardiovascular dynamics estimated from multiple peripheral circulatory signals

This paper presents a method for comparing multiple circulatory waveforms measured at different locations to improve cardiovascular parameter estimation from these signals. The method identifies the distinct vascular dynamics that shape each waveform signal, and estimates the common cardiac flow input shared by them. This signal-processing algorithm uses the Laguerre function series expansion for modeling the hemodynamics of each arterial branch, and identifies unknown parameters in these models from peripheral waveforms using multichannel blind system identification. An effective technique for determining the Laguerre base pole is developed, so that the Laguerre expansion captures and quickly converges to the intrinsic arterial dynamics observed in the two circulatory signals. Furthermore, a novel deconvolution method is developed in order to stably invert the identified dynamic models for estimating the cardiac output (CO) waveform from peripheral pressure waveforms. The method is applied to experimental swine data. A mean error of less than 5% with the measured peripheral pressure waveforms has been achieved using the models and excellent agreement between the estimated CO waveforms and the gold standard measurements have been obtained.     

Experimental and theoretical comparison of some algorithms forbeamforming in single receiver adaptive arrays

Adaptive null steering in single receiver adaptive arrays is discussed. The single receiver structure allows only output power for a given set of weights to be measured. The problem, then, is to adaptively adjust the weights of the antenna array based on output power measurements only so as to reject interference signals while maintaining a fixed response in a given look direction. The authors determine the optimal beamformer weights in a single step, by estimating the covariance matrix of the array sensor outputs using a weight perturbation technique. Based on this covariance matrix estimate, three different approaches for finding the beamformer weights are studied. The first corresponds to a sample matrix inversion scheme, with the sample covariance matrix replaced by the one obtained from the perturbation estimation method, while in the second approach the weights are determined using an eigenvalue decomposition of the covariance matrix estimate. In the third approach the directions-of-arrival (DOAs) of the incoming wavefronts are first estimated, and then, in a second step, the beamformer weights are calculated from the DOA estimates. The advantage of the third approach is that this method is not affected by correlation between the different sources     

Refining inaccurate sensor positions using target at unknown location

This paper addresses the problem of improving the receiver positions in a sensor array using the positioning measurements of targets at unknown locations. Using an efficient estimate of the target locations, a computationally efficient estimator is proposed to refine the sensor positions using the same set of measurements from the targets. The considered measurements include TOA and TDOA. The proposed estimator has closed-form and is able to reach the CRLB accuracy under small Gaussian noise which is supported by theoretical analysis and simulation studies.     

A method to improve the estimation of conduction velocitydistributions over a short segment of nerve

Accurate, noninvasive determination of the distribution of conduction velocities (DCV) among fibers of a peripheral nerve has the potential to improve both clinical diagnoses of pathology and longitudinal studies of the progress of disease or the efficacy of treatments. Current techniques rely on long distances of propagation to increase the amount of temporal dispersion in the compound signals and reduce the relative effect of errors in the forward model. The method described in this paper attempts to reduce errors in DCV estimation through transfer function normalization and, thereby, eliminate the need for long segments of nerve. Compound action potential (CAP) signals are recorded from several, equally spaced electrodes in an array spanning only a 10-cm length of nerve. Relative nerve-to-electrode transfer functions (NETF's) between the nerve and each of the array electrodes are estimated by comparing discrete Fourier transforms of the array signals. NETF's are normalized along the array so that waveform differences can be attributed to the effects of temporal dispersion between recordings, and more accurate DCV estimates can be calculated from the short nerve segment. The method is tested using simulated and real CAP data. DCV estimates are improved for simulated signals. The normalization procedure results in DCV's that qualitatively match those from the literature when used on actual CAP recordings.     

Color reproduction from noisy CFA data of single sensor digital cameras.

Single sensor digital color still/video cameras capture images using a color filter array (CFA) and require color interpolation (demosaicking) to reconstruct full color images. The color reproduction has to combat sensor noises which are channel dependent. If untreated in demosaicking, sensor noises can cause color artifacts that are hard to remove later by a separate denoising process, because the demosaicking process complicates the noise characteristics by blending noises of different color channels. This paper presents a joint demosaicking-denoising approach to overcome this difficulty. The color image is restored from noisy mosaic data in two steps. First, the difference signals of color channels are estimated by linear minimum mean square-error estimation. This process exploits both spectral and spatial correlations to simultaneously suppress sensor noise and interpolation error. With the estimated difference signals, the full resolution green channel is recovered. The second step involves in a wavelet-based denoising process to remove the CFA channel-dependent noises from the reconstructed green channel. The red and blue channels are subsequently recovered. Simulated and real CFA mosaic data are used to evaluate the performance of the proposed joint demosaicking-denoising scheme and compare it with many recently developed sophisticated demosaicking and denoising schemes.     

Nested planar array: configuration design,optimal array and DOA estimation

Nested array enables to enhance localisation resolution and achieve under-determined direction of arrival (DOA) estimation. In this paper, we improve the traditional nested planar array to achieve more degrees of freedom (DOFs) and better angle estimation performance. The closed-form expressions for sensor positions of the improved array are given and the optimal array configuration for largest available DOFs is derived. Meanwhile, a computationally efficient DOA estimation algorithm is proposed. Specifically, we utilise two dimensional Discrete Fourier Transform (2D DFT) method to obtain the coarse DOA estimates; Subsequently, we achieve the fine DOA estimates by 2D spatial smoothing multiple signals classification (SS-MUSIC) algorithm. The proposed algorithm enjoys the same estimation accuracy as SS-MUSIC algorithm but with lower complexity because the coarse DOA estimates enable to shrink the range of spectral search. In addition, estimation of the number of signals is not required by 2D DFT method. Extensive simulation results testify the effectiveness of the proposed algorithm.     

船舶海上超视距无线自定位技术研究

针对卫星拒止环境下,海上船舶及其编队对导航定位性能的最低保障需求,开展了一系列研究。首先分析了线阵综合测角的基本理论,用遗传算法对线阵进行了综合仿真,用谱估计对线阵测角进行了研究;其次对圆阵综合测角进行了理论分析,对圆阵、圆柱阵进行了综合计算和谱估计测角仿真,分析了不同阵元数、信噪比等情况下的测向夹角情况,并对多信号同时测角情况进行了仿真;最后阐述了夹角定位的基本思想,提出了船舶传感器构阵测角的可能形式,并对定位数据进行了分析及滤波优化。研究结果表明,相关技术对船舶海上导航定位有积极的参考意义。