Inverse Filtering Technique Research Articles

Multimodal coded excitation in medical ultrasonic imaging.

Coded excitation has been used successfully in medical ultrasound to improve the signal to noise ratio and penetration depth. Furthermore, inverse filtering techniques have been shown to improve axial resolution without significant loss in contrast due to the range sidelobe phenomenon. With the next generation ultrasound scanners, the opportunity exist for transmitting multiple codes simultaneously, with or without focusing, into the region of interest. Using appropriate 2D inverse filtering on receive, it is possible to control the axial and lateral resolutions of the imaging system while maintaining an acceptable tradeoff with image contrast. In this paper, we review recent results on 1D and 2D inverse filtering of beamformed echo data resulting from insonification of speckle‐generating cyst phantoms with multiple codes. We demonstrate that (regularized) inverse filtering can be derived from the space‐time array patterns. Furthermore, we demonstrate the feasibility of reconstructing multiple image lines in parallel to increase the frame rate in 2D and 3D imaging.

New method for measuring sound absorption coefficients in an industrial hall

Predicting the sound pressure level at a workplace requires in‐situ characterization of the facings. This work describes a new method for the measurement of the sound absorption coefficient of flat panels present in industrial halls. In such room, it is necessary to separate the echo coming from the studied panel from the others due to the entire reverberation. This separation has been achieved in space by an acoustic array and in time by an impulse sound source. The array processing uses a multipolar weighting to achieve a directivity constant with frequency and with attenuated side lobes. This weighting requires a limited number of microphones. The impulse source has been designed using the inverse impulse response from the emission system (equalizer, amplifier, loudspeaker). This inverse filtering technique allows equalizing the response of the emission system in order to radiate very short pulses. Sound absorption coefficient of several flat facings have been evaluated by mean of this new device in a ...

Synthesis and analysis of linear and nonlinear acoustical vortices

Acoustical screw dislocations are synthesized in various configurations with a versatile experimental setup. The experimental setup is based on the inverse filter technique and allows one to synthesize one or more acoustical vortices with a chosen width, position, and topological charge. An interesting feature of this experimental facility to study screw dislocation behavior is the direct measurement in amplitude and phase. This characteristic is used to develop an original method of decomposition of an acoustical vortex field in order to analyze the acoustical vortices. Moreover, the behaviors of two acoustical vortices of the same or opposite charge have been studied experimentally and compared to theoretical laws.

Nonlinear focusing of acoustic shock waves at a caustic cusp

The present study investigates the focusing of acoustical weak shock waves incoming on a cusped caustic. The theoretical model is based on the Khokhlov-Zabolotskaya equation and its specific boundary conditions. Based on the so-called Guiraud's similitude law for a step shock, a new explanation about the wavefront unfolding due to nonlinear self-refraction is proposed. This effect is shown to be associated not only to nonlinearities, as expected by previous authors, but also to the nonlocal geometry of the wavefront. Numerical simulations confirm the sensitivity of the process to wavefront geometry. Theoretical modeling and numerical simulations are substantiated by an original experiment. This one is carried out in two steps. First, the canonical Pearcey function is synthesized in linear regime by the inverse filter technique. In the second step, the same wavefront is emitted but with a high amplitude to generate shock waves during the propagation. The experimental results are compared with remarkable agreement to the numerical ones. Finally, applications to sonic boom are briefly discussed.

Vibrato in Singing Voice: The Link between Source-Filter and Sinusoidal Models

The application of inverse filtering techniques for high-quality singing voice analysis/synthesis is discussed. In the context of source-filter models, inverse filtering provides a noninvasive method to extract the voice source, and thus to study voice quality. Although this approach is widely used in speech synthesis, this is not the case in singing voice. Several studies have proved that inverse filtering techniques fail in the case of singing voice, the reasons being unclear. In order to shed light on this problem, we will consider here an additional feature of singing voice, not present in speech: the vibrato. Vibrato has been traditionally studied by sinusoidal modeling. As an alternative, we will introduce here a novel noninteractive source filter model that incorporates the mechanisms of vibrato generation. This model will also allow the comparison of the results produced by inverse filtering techniques and by sinusoidal modeling, as they apply to singing voice and not to speech. In this way, the limitations of these conventional techniques, described in previous literature, will be explained. Both synthetic signals and singer recordings are used to validate and compare the techniques presented in the paper.

Revisiting the Stokes relations in a time-reversal cavity: Suppression of intraplate echoes induced by a titanium plate

When focusing through plates or tubes the presence of multiple interfaces induces reflected wavefronts that follow the main wavefront. Adaptive focusing techniques can be used to cancel the echoes. For that purpose, two linear arrays of transducers have been placed on each side of a titanium plate. Three propagation operators have been acquired: transmission from one array to the other, and two reflection operators acquired by each array. In this work, two adaptative focusing methods have been used to cancel the echoes: first, they have been suppressed with a time-reversal mirror, using the two arrays’ cavity surrounding the plate. Second, the echoes have been canceled by using the inverse filter technique, inverting the transmission operator. Thus, the inverse filter achieves echo cancellation by using only the transmitted fields, whereas time reversal also requires the reflected fields. It is shown how transmission and reflection operators are related by the Stokes relations in a matrix formalism. These relations clearly exhibit how the inverse filter takes advantage of the reflections in the medium. An iterative mathematical resolution of these equations yields a new way to invert the transmission operator.

High-resolution ultrasonic brain imaging: Noninvasive adaptive focusing based on twin arrays

Ultrasonic imaging systems assume a constant acoustic velocity in human tissues in the beamforming process. However, in the case of brain imaging, strong skull aberrations induce a displacement of the focal stain, a spreading in the main lobe, and an increase in the sidelobes level. This limits considerably ultrasonic brain imaging applications. It has been shown that a very accurate focusing could be achieved through a human skull by using an inverse filter technique. This method, based on a set of acoustic sensors located inside the brain, demonstrated invasively that it was possible to achieve high-resolution brain imaging. A noninvasive ultrasonic method is presented. It is based on two arrays located on each side of the head. Each array focuses on the other one to deduce the aberrations induced by the skull bone. Once the effect of the skull bone in front of each array is extracted, the wavefronts are corrected and emitted. This noninvasive technique restores the position of the focal stain and lowers the secondary lobe level up to 15 dB compared to a cylindrical focusing.

Real time inverse filter focusing through iterative time reversal

In order to achieve an optimal focusing through heterogeneous media we need to build the inverse filter of the propagation operator. Time reversal is an easy and robust way to achieve such an inverse filter in nondissipative media. However, as soon as losses appear in the medium, time reversal is not equivalent to the inverse filter anymore. Consequently, it does not produce the optimal focusing and beam degradations may appear. In such cases, we showed in previous works that the optimal focusing can be recovered by using the so-called spatiotemporal inverse filter technique. This process requires the presence of a complete set of receivers inside the medium. It allows one to reach the optimal focusing even in extreme situations such as ultrasonic focusing through human skull or audible sound focusing in strongly reverberant rooms. But, this technique is time consuming and implied fastidious numerical calculations. In this paper we propose a new way to process this inverse filter focusing technique in real time and without any calculation. The new process is based on iterative time reversal process. Contrary to the classical inverse filter technique, this iteration does not require any computation and achieves the inverse filter in an experimental way using wave propagation instead of computational power. The convergence from time reversal to inverse filter during the iterative process is theoretically explained. Finally, the feasibility of this iterative technique is experimentally demonstrated for ultrasound applications.

Experimental simulation of supersonic superboom in a water tank: nonlinear focusing of weak shock waves at a fold caustic.

An accelerating supersonic aircraft produces noisy superboom due to acoustical shock wave focusing at a fold caustic. This phenomenon is modeled by the mixed-type nonlinear Tricomi equation. An innovative experimental simulation in a water tank has been carried out, with perfect similitude to sonic boom in air. In the linear regime, the canonical Airy function is reproduced using the inverse filter technique. In the nonlinear regime (weak shock waves), the experiment demonstrates the key role of nonlinear effects: they limit the field amplitude, distort the sonic line, and strongly alter the phase of the signal, in agreement with numerical simulations.

Real time inverse filter focusing by iterative time reversal

In order to achieve an optimal focusing through heterogeneous media, we need to build an inverse filter of the propagation operator. Time-reversal achieves such an inverse filter in nondissipative media. However, as soon as losses appear in the medium, time-reversal is no more equivalent to an inverse filter. Consequently, it does no more produce the optimal focusing and beam degradations may appear. In such cases, previous works have shown that the optimal focusing can be recovered by using a spatiotemporal inverse filter technique. This process requires the presence of a set of receivers inside the medium but allows us to reach the optimal focusing even in extreme situations such as ultrasonic focusing through a human skull. To date, it remained time consuming and implied fastidious numerical calculations. We propose here a new way to process this inverse filtering in real time and without any calculation. The new process is based on iterative time-reversal processing. Contrary to the classical inversion, this iteration does not require any computation as it achieves the inverse filter experimentally using waves propagation instead of computational power. The feasibility of this iterative technique is theoretically and experimentally demonstrated for ultrasound applications and its potential for telecommunications is discussed.

SEC OF SIMPLE POLYMERS WITH MOLAR MASS DETECTION IN PRESENCE OF INSTRUMENTAL BROADENING. COMPUTER SIMULATION STUDY ON THE CALCULATION OF UNBIASED MOLECULAR WEIGHT DISTRIBUTIONS

This theoretical work presents a correction procedure for calculating an unbiased MWD when analyzing a linear homopolymer by size exclusion chromatography with molar mass detection. The fractionation and the measurements are ideal, but the chromatograms are distorted by a nonuniform and skewed instrumental broadening (IB), and are contaminated with a zero-mean random noise. The main assumption is that the molecular weight calibration is linear in the chromatogram range. The following phenomenological procedure was proposed: (i) independently correct the mass- and molar mass chromatograms for IB through a robust inverse filtering technique; (ii) estimate the unbiased calibration from the ratio of the corrected chromatograms; (iii) adjust a straight line to the mid-values of that calibration; and (iv) estimate the unbiased MWD from the linear calibration and the corrected mass chromatogram. For the chromatogram inversions, a minimum sized broadening matrix was selected and a singular value decomposition technique was applied. The following effects were evaluated: of an increased measurement noise, of uncertainties in the range of the measured chromatograms, and of systematic errors in the IB function. The method fails for very narrow MWDs.

Objective assessment of hoarseness by measuring jitter.

The objective measurement of hoarseness by measuring 'jitter' (the average percentage pitch-period variation between consecutive pitch-cycles) using an inverse filtering technique is described. Twenty-five patients with a variety of causes of hoarseness were studied, together with five individuals who had mild hoarseness induced by histamine challenge and 12 normal individuals. The mean severity of jitter in the patient group (9.8%) was significantly different from the normals. (1.04%) In addition, there was a significant correlation (R2 = 0.53; P < 0.0001) between jitter and subjective assessment of hoarseness. The mean values of jitter with histamine challenge before and after recovery (1.03%, and 1.18%) were significantly different (P < 0.0001) to the mean maximum value during the challenge (2.64%). These data suggest that jitter is an objective and repeatable measurement of hoarseness-even small changes in hoarseness in individual patients. It is likely to prove most effective for monitoring treatment response.

1P1-F2 Improvement of the Depth From Focus Criteria's performance by using Inverse Filtering technique Focal Distance proberties.

1P1-F2 Improvement of the Depth From Focus Criteria's performance by using Inverse Filtering technique Focal Distance proberties.

Reconstruction of cone-beam projections from Compton scattered data

The problem of reconstructing a 3D source distribution from Compton scattered data can be separated into two tasks. First, the angular distribution of line projections at different observation points within the detector volume are reconstructed. Then, reconstruction techniques are applied to the resulting cone-beam projections to synthesize the 3D source distribution. This paper describes an analytic solution for the first, yet unsolved, task. Building on the convolution theorem in spherical coordinates, a back-projection and inverse filtering technique in terms of spherical harmonics is formulated, The rotation invariance of the point response of the back-projection in spherical coordinates is proved; and the corresponding inverse filter function is derived. The resulting filtered back-projection algorithm then consists of a summation over all detected events of fixed and known event response functions. Measurement errors, which for Compton scatter detectors are typically different for each detected event, can easily be accounted for in the proposed algorithm. The computational cost of the algorithm is O(NT/sup 2/), where N is the number of detected events and /spl pi//T is the desired angular resolution.

AN IMPROVED FINGERPRINT CLASSIFICATION TECHNIQUE

This paper presents an automatic fingerprint classification technique similar to that reported in [2] but, an inverse filtering technique was introduced to restore the distorted parts of the images prior to the feature extraction stage. The results have shown that introducing the inverse filter stage has improved the percentage of correct classification. It reaches 97.5% compared to the 95% correct classification obtained using the previously reported technique.

A method for generating natural-sounding speech stimuli for cognitive brain research

Objective: In response to the rapidly increasing interest in using human voice in cognitive brain research, a new method, semisynthetic speech generation (SSG), is presented for generation of speech stimuli.Methods: The method synthesizes speech stimuli as a combination of purely artificial processes and processes that originate from the natural human speech production mechanism. SSG first estimates the source of speech, the glottal flow, from a natural utterance using an inverse filtering technique. The glottal flow obtained is then used as an excitation to an artificial digital filter that models the formant structure of speech.Results: SSG is superior to commercial voice synthesizers because it yields speech stimuli of a highly natural quality due to the contribution of the man-originating glottal excitation.Conclusion: The artificial modelling of the vocal tract enables one to adjust the formant frequencies of the stimuli as desired, thus making SSG suitable for cognitive experiments using speech sounds as stimuli.

The nature of recorded magnetisation transitions in thin film media

The longitudinal components of the recorded magnetisations in a number of thin film disk media were determined from captured isolated replay pulses using inverse filtering techniques. The computed magnetisations exhibit asymmetry in the tails in the approach to saturation. These recovered magnetisation transitions and their derivatives were linearly superposed to investigate the persistence of recorded magnetisation with recording density. Furthermore, the recovered magnetisation distributions were employed to investigate the effect of transition shape on the noise power spectral densities. The results are compared with experimental observations.

Super-resolution of pulsed multipath channels for delay spread characterization

Frequency selective channels can be measured by either a continuous wave frequency sweep, which gives directly the transfer function, or by sounding the channel with pulses, which results in the impulse response. When the sounding pulse becomes a chirp-type waveform, the two approaches become similar. Simple rectangular pulses can be used for partial, but usually sufficient, channel characterization. In this paper, the process is described for resolving impulse response components into bins smaller than the duration of the sounding pulse and smaller than the reciprocal of the channel bandwidth. From such super-resolution, the delay spread of the propagation channel can be established accurately even when it is much less than the sounding pulse duration, allowing longer, higher energy pulses to be used for channel characterization. The process is demonstrated using subtractive deconvolution where a loop gain of unity is shown to be stable; and a modified inverse filter technique, in which the modification caters for the spectral zeros of the sounding pulse.

CONDITION MONITORING AND DIAGNOSTICS IN AUTOMATIC MACHINES: COMPARISON OF VIBRATION ANALYSIS TECHNIQUES

This paper assesses and compares the effectiveness and reliability of different vibration analysis techniques for fault detection and diagnostics in cam mechanisms used in high performance automatic packaging machines. Firstly, the results of the application of widely used techniques are given and their limitations are delineated. Secondly, the application of the wavelet transform is proposed, since the vibrations of this kind of machine are highly non-stationary. Finally, an inverse filtering technique is used to recover the actual output link motion from the casing vibration, in order to enable condition monitoring and diagnostics to be carried out through the analysis of the mechanism motion.

Velopharyngeal insufficiency analysis based on the lossless tube model

The purpose of this study is to design a method to reconstruct the shapes of articulators including the vocal tract and nasal tract during phonation and to quantitatively identify velopharyngeal dysfunction when it occurs. Voice in patients with velopharyngeal insufficiency have the psychoacoustic impression of hypernasality that results from an over abundance of sounds resonating in the nasal tract. Patients with unrepaired soft palate clefts typically demonstrate this abnormal speech. The articulation model of velopharyngeal dysfunction is predicated upon the lossless tube model. During recording of the sustained /a/ sound, subjects were asked to block their nostrils with two fingers. The resulting closed nasal tract introduces zeros and the opened vocal tract contributes poles. Therefore, an ARMA model is formed. PARCOR analysis and the inverse filter technique are used to reconstruct both the shapes of the vocal tract and the nasal tract. The sufficiency of the velopharynx can be determined by measuring the cross-sectional area of velopharynx.