Ultrasound super-resolution imaging with a hierarchical Kalman tracker.

Microbubble (MB) tracking plays an important role in ultrasound super-resolution imaging (SRI) by enabling velocity estimation and improving image quality. This work presents a new hierarchical Kalman (HK) tracker to achieve better performance at scenarios with high concentrations of MBs and high localization uncertainty. The method attempts to follow MBs with different velocity ranges using different Kalman filters. An extended simulation framework for evaluating trackers is also presented and used for comparison of the proposed HK tracker with the nearest-neighbor (NN) and Kalman (K) trackers. The HK tracks were most similar to the ground truth with the highest Jaccard similarity coefficient in 79% of the scenarios and the lowest root-mean-square error in 72% of the scenarios. The HK tracker reconstructed vessels with a more accurate diameter. In a scenario with an uncertainty of 51.2µm in MB localization, a vessel diameter of 250µm was estimated as 257µm by HK tracker, compared with 329µm and 389µm for the K and NN trackers. In the same scenario, the HK tracker estimated MB velocities with a relative bias down to 1.7% and a relative standard deviation down to 8.3%. Finally, the different tracking techniques were applied to in vivo data from rat kidneys, and trends similar to the simulations were observed. Conclusively, the results showed an improvement in tracking performance, when the HK tracker was employed in comparison with the NN and K trackers.

Super-Resolution Imaging with Ultrasound for Visualization of the Renal Microvasculature in Rats Before and After Renal Ischemia: A Pilot Study.

In vivo monitoring of the microvasculature is relevant since diseases such as diabetes, ischemia, or cancer cause microvascular impairment. Super-resolution ultrasound imaging allows in vivo examination of the microvasculature by detecting and tracking sparsely distributed intravascular microbubbles over a minute-long period. The ability to create detailed images of the renal vasculature of Sprague-Dawley rats using a modified clinical ultrasound platform was investigated in this study. Additionally, we hypothesized that early ischemic damage to the renal microcirculation could be visualized. After a baseline scan of the exposed kidney, 10 rats underwent clamping of the renal vein (n = 5) or artery (n = 5) for 45 min. The kidneys were rescanned at the onset of clamp release and after 60 min of reperfusion. Using a processing pipeline for tissue motion compensation and microbubble tracking, super-resolution images with a very high level of detail were constructed. Image filtration allowed further characterization of the vasculature by isolating specific vessels such as the ascending vasa recta with a 15-20 µm diameter. Using the super-resolution images alone, it was only possible for six assessors to consistently distinguish the healthy renal microvasculature from the microvasculature at the onset of vein clamp release. Future studies will aim at attaining quantitative estimations of alterations in the renal microvascular blood flow using super-resolution ultrasound imaging.

The role of spectral detail in the binaural transfer function on perceived externalization in a reverberant environment.

Individual binaural room impulse responses (BRIRs) were recorded at a distance of 1.5 m for azimuth angles of 0° and 50° in a reverberant room. Spectral details were reduced in either the direct or the reverberant part of the BRIRs by averaging the magnitude responses with band-pass filters. For various filter bandwidths, the modified BRIRs were convolved with broadband noise and listeners judged the perceived position of the noise when virtualized over headphones. Only reductions in spectral details of the direct part obtained with filter bandwidths broader than one equivalent rectangular bandwidth affected externalization. Reductions in spectral details of the reverberant part had only little influence on externalization. In both conditions, externalization was not as pronounced at 0° as at 50°. To characterize the auditory processes that may be involved in the perception of externalization, a quantitative model is proposed. The model includes an echo-suppression mechanism, a filterbank describing the frequency selectivity in the cochlea and a binaural stage that measures the deviations of the interaural level differences between the considered input and the unmodified input. These deviations, integrated across frequency, are then mapped to a value that corresponds to the perceived externalization.

The effect of interaural-level-difference fluctuations on the externalization of sound.

Real-world sound sources are usually perceived as externalized and thus properly localized in both direction and distance. This is largely due to (1) the acoustic filtering by the head, torso, and pinna, resulting in modifications of the signal spectrum and thereby a frequency-dependent shaping of interaural cues and (2) interaural cues provided by the reverberation inside an enclosed space. This study first investigated the effect of room reverberation on the spectro-temporal behavior of interaural level differences (ILDs) by analyzing dummy-head recordings of speech played at different distances in a standard listening room. Next, the effect of ILD fluctuations on the degree of externalization was investigated in a psychoacoustic experiment performed in the same listening room. Individual binaural impulse responses were used to simulate a distant sound source delivered via headphones. The ILDs were altered using a gammatone filterbank for analysis and resynthesis, where the envelopes of the left and right-ear signals were modified such that the naturally occurring fluctuations of the ILDs were restricted. This manipulation reduced the perceived degree of externalization. This was consistent with the analysis of short-term ILDs at different distances showing that a decreased distance to the sound source also reduced the ILD fluctuations.

Extracting the invariant model from the feedback paths of digital hearing aids.

Feedback whistling is a severe problem with hearing aids. A typical acoustical feedback path represents a wave propagation path from the receiver to the microphone and includes many complicated effects among which some are invariant or nearly invariant for all users and in all acoustical environments given a specific type of hearing aids. Based on this observation, a feedback path model that consists of an invariant model and a variant model is proposed. A common-acoustical-pole and zero model-based approach and an iterative least-square search-based approach are used to extract the invariant model from a set of impulse responses of the feedback paths. A hybrid approach combining the two methods is also proposed. The general properties of the three methods are studied using artificial datasets, and the methods are cross-validated using the measured feedback paths. The results show that the proposed hybrid method gives the best overall performance, and the extracted invariant model is effective in modeling the feedback path.

Using a reflection model for modeling the dynamic feedback path of digital hearing aids.

Feedback whistling is one of the severe problems with hearing aids, especially in dynamic situations when the users hug, pick up a telephone, etc. This paper investigates the properties of the dynamic feedback paths of digital hearing aids and proposes a model based on a reflection assumption. The model is compared with two existing models: a direct model and an initialization model, using the measured dynamic feedback paths. The comparison shows that the proposed approach is able to model the dynamic feedback paths more efficiently and accurately in terms of mean-square error and maximum stable gain. The method is also extended to dual-microphone hearing aids to assess the possibility of relating the two dynamic feedback paths through the reflection model. However, it is found that in a complicated acoustic environment, the relation between the two feedback paths can be very intricate and difficult to exploit to yield better modeling of the dynamic feedback paths.

Adaptive spectral doppler estimation.

In this paper, 2 adaptive spectral estimation techniques are analyzed for spectral Doppler ultrasound. The purpose is to minimize the observation window needed to estimate the spectrogram to provide a better temporal resolution and gain more flexibility when designing the data acquisition sequence. The methods can also provide better quality of the estimated power spectral density (PSD) of the blood signal. Adaptive spectral estimation techniques are known to provide good spectral resolution and contrast even when the observation window is very short. The 2 adaptive techniques are tested and compared with the averaged periodogram (Welch's method). The blood power spectral capon (BPC) method is based on a standard minimum variance technique adapted to account for both averaging over slow-time and depth. The blood amplitude and phase estimation technique (BAPES) is based on finding a set of matched filters (one for each velocity component of interest) and filtering the blood process over slow-time and averaging over depth to find the PSD. The methods are tested using various experiments and simulations. First, controlled flow-rig experiments with steady laminar flow are carried out. Simulations in Field II for pulsating flow resembling the femoral artery are also analyzed. The simulations are followed by in vivo measurement on the common carotid artery. In all simulations and experiments it was concluded that the adaptive methods display superior performance for short observation windows compared with the averaged periodogram. Computational costs and implementation details are also discussed.

Broadband minimum variance beamforming for ultrasound imaging.

A minimum variance (MV) approach for near-field beamforming of broadband data is proposed. The approach is implemented in the frequency domain, and it provides a set of adapted, complex apodization weights for each frequency subband. The performance of the proposed MV beamformer is tested on simulated data obtained using Field II. The method is validated using synthetic aperture data and data obtained from a plane wave emission. Data for 13 point targets and a circular cyst with a radius of 5 mm are simulated. The performance of the MV beamformer is compared with delay-and-sum (DS) using boxcar weights and Hanning weights and is quantified by the full width at half maximum (FWHM) and the peak-side-lobe level (PSL). Single emission {DS boxcar, DS Hanning, MV} provide a PSL of {-16, -36, -49} dB and a FWHM of {0.79, 1.33, 0.08} mm. Using all 128 emissions, {DS boxcar, DS Hanning, MV} provides a PSL of {-32, -49, -65} dB, and a FWHM of {0.63, 0.97, 0.08} mm. The contrast of the beamformed single emission responses of the circular cyst was calculated as {-18, -37, -40} dB. The simulations have shown that the frequency subband MV beamformer provides a significant increase in lateral resolution compared with DS, even when using considerably fewer emissions. An increase in resolution is seen when using only one single emission. Furthermore, the effect of steering vector errors is investigated. The steering vector errors are investigated by applying an error of the sound speed estimate to the ultrasound data. As the error increases, it is seen that the MV beamformer is not as robust compared with the DS beamformer with boxcar and Hanning weights. Nevertheless, it is noted that the DS does not outperform the MV beamformer. For errors of 2% and 4% of the correct value, the FWHM are {0.81, 1.25, 0.34} mm and {0.89, 1.44, 0.46} mm, respectively.

Coded ultrasound for blood flow estimation using subband processing.

This paper investigates the use of coded excitation for blood flow estimation in medical ultrasound. Traditional autocorrelation estimators use narrow-band excitation signals to provide sufficient signal-to-noise-ratio (SNR) and velocity estimation performance. In this paper, broadband coded signals are used to increase SNR, followed by subband processing. The received broadband signal is filtered using a set of narrow-band filters. Estimating the velocity in each of the bands and averaging the results yields better performance compared with what would be possible when transmitting a narrow-band pulse directly. Also, the spatial resolution of the narrow-band pulse would be too poor for brightness-mode (B-mode) imaging, and additional transmissions would be required to update the B-mode image. For the described approach in the paper, there is no need for additional transmissions, because the excitation signal is broadband and has good spatial resolution after pulse compression. This means that time can be saved by using the same data for B-mode imaging and blood flow estimation. Two different coding schemes are used in this paper, Barker codes and Golay codes. The performance of the codes for velocity estimation is compared with a conventional approach transmitting a narrow-band pulse. The study was carried out using an experimental ultrasound scanner and a commercial linear array 7 MHz transducer. A circulating flow rig was scanned with a beam-to-flow angle of 60 degrees. The flow in the rig was laminar and had a parabolic flow-profile with a peak velocity of 0.09 m/s. The mean relative standard deviation of the velocity estimate using the reference method with an 8-cycle excitation pulse at 7 MHz was 0.544% compared with the peak velocity in the rig. Two Barker codes were tested with a length of 5 and 13 bits, respectively. The corresponding mean relative standard deviations were 0.367% and 0.310%, respectively. For the Golay coded experiment, two 8-bit codes were used, and the mean relative standard deviation was 0.335%.

High frame-rate blood vector velocity imaging using plane waves: simulations and preliminary experiments.

Conventional ultrasound methods for acquiring color images of blood velocity are limited by a relatively low frame-rate and are restricted to give velocity estimates along the ultrasound beam direction only. To circumvent these limitations, the method presented in this paper uses 3 techniques: 1) The ultrasound is not focused during the transmissions of the ultrasound signals; 2) A 13-bit Barker code is transmitted simultaneously from each transducer element; and 3) The 2-D vector velocity of the blood is estimated using 2-D cross-correlation. A parameter study was performed using the Field II program, and performance of the method was investigated when a virtual blood vessel was scanned by a linear array transducer. An improved parameter set for the method was identified from the parameter study, and a flow rig measurement was performed using the same improved setup as in the simulations. Finally, the common carotid artery of a healthy male was scanned with a scan sequence that satisfies the limits set by the Food and Drug Administration. Vector velocity images were obtained with a frame-rate of 100 Hz where 40 speckle images are used for each vector velocity image. It was found that the blood flow approximately followed the vessel wall, and that maximum velocity was approximately 1 m/s, which is a normal value for a healthy person. To further evaluate the method, the test person was scanned with magnetic resonance (MR) angiography. The volume flow derived from the MR scanning was compared with that from the ultrasound scanning. A deviation of 9% between the 2 volume flow estimates was found.

Multi-frequency encoding for fast color flow or quadroplex imaging.

Ultrasonic color flow maps are made by estimating the velocities line by line over the region of interest. For each velocity estimate, multiple repetitions are needed. This sets a limit on the frame rate, which becomes increasingly severe when imaging deeper lying structures or when simultaneously acquiring spectrogram data for triplex imaging. This paper proposes a method for decreasing the data acquisition time by simultaneously sampling multiple lines for color flow maps, using narrow band signals with approximately disjoint spectral support. The signals are separated in the receiver by filters matched to the emitted waveforms, producing a number of data sets with different center frequencies. The autocorrelation estimator is then applied to each of the data sets. The method is presented, various side effects are considered, and the method is tested on data from a recirculating flow phantom. A mean standard deviation across the flow profile of 3.1, 2.5, and 2.1% of the peak velocity was found for bands at 5 MHz, 7 MHz, and 9 MHz, respectively. Alternatively, the method can be used for simultaneously sampling data for a color flow map and for multiple spectrograms using different spectral bands. Using three spectral bands, data for a color flow map and two independent spectrograms can be acquired at the time normally spent on acquiring data for a color flow map only. This yields an expansion of triplex imaging called multifrequency quadroplex imaging, which enables study of the flow over an arterial stenosis by simultaneously acquiring spectrograms on both sides of the stenosis, while maintaining the color flow map. The method was tested in vivo on data from the common carotid artery of a healthy male volunteer, both for fast color flow mapping and for multifrequency quadroplex imaging.

Spatial encoding using a code division technique for fast ultrasound imaging.

This paper describes a method for spatial encoding in synthetic transmit aperture ultrasound imaging. This allows several ultrasonic sources to be active simultaneously. The method is based on transmitting pseudorandom sequences to spatially encode the transmitters. The data can be decoded after only one transmission using the knowledge of the transmitted code sequences as opposed to other spatial encoding techniques, such as Hadamard or Golay encoding. This makes the method less sensitive to motion, and data can be acquired using fewer transmissions. The aim of this paper is to analyze the underlying theory and to test the feasibility in a physical system. The method has been evaluated in simulations using Field II in which the point-spread functions were simulated for different depths for a 7 MHz linear array transducer. A signal-to-noise ratio (SNR) simulation also was included in the study in which an improvement in SNR of approximately 1.5 dB was attained compared to the standard synthetic transmit aperture (STA) firing scheme. Considering the amount of energy transmitted, this value is low. A plausible explanation is given that is verified in simulation. The method also was tested in an experimental ultrasound scanner and compared to a synthetic transmit aperture ultrasound imaging scheme using a sinusoidal excitation. The performance of the proposed method was comparable to the reference with respect to axial and lateral resolution, but it displayed poorer contrast with sidelobe levels at approximately - 40 dB compared to the mainlobe.

Designing waveforms for temporal encoding using a frequency sampling method.

In this paper a method for designing waveforms for temporal encoding in medical ultrasound imaging is described. The method is based on least squares optimization and is used to design nonlinear frequency modulated signals for synthetic transmit aperture imaging. By using the proposed design method, the amplitude spectrum of the transmitted waveform can be optimized, such that most of the energy is transmitted where the transducer has large amplification. To test the design method, a waveform was designed for a BK8804 linear array transducer. The resulting nonlinear frequency modulated waveform was compared to a linear frequency modulated signal with amplitude tapering, previously used in clinical studies for synthetic transmit aperture imaging. The latter had a relatively flat spectrum which implied that the waveform tried to excite all frequencies including ones with low amplification. The proposed waveform, on the other hand, was designed so that only frequencies where the transducer had a large amplification were excited. Hereby, unnecessary heating of the transducer could be avoided and the signal-to-noise ratio could be increased. The experimental ultrasound scanner RASMUS was used to evaluate the method experimentally. Due to the careful waveform design optimized for the transducer at hand, a theoretic gain in signal-to-noise ratio of 4.9 dB compared to the reference excitation was found, even though the energy of the nonlinear frequency modulated signal was 71% of the energy of the reference signal. This was supported by a signal-to-noise ratio measurement and comparison in penetration depth, where an increase of 1 cm was found in favor for the proposed waveform. Axial and lateral resolutions at full-width half-maximum were compared in a water phantom at depths of 42, 62, 82, and 102 mm. The axial resolutions of the nonlinear frequency modulated signal were 0.62, 0.69, 0.60, and 0.60 mm, respectively. The corresponding axial resolutions for the reference waveform were 0.58, 0.65, 0.62, and 0.60 mm, respectively. The compression properties of the matched filter (mismatched filter for the linear frequency modulated signal) were tested for both waveforms in simulation with respect to the Doppler frequency shift occurring when probing moving objects. It was concluded that the Doppler effect of moving targets does not significantly degrade the filtered output. Finally, in vivo measurements are shown for both methods, wherein the common carotid artery on a 27-year-old healthy male was scanned.

Directional velocity estimation using a spatio-temporal encoding technique based on frequency division for synthetic transmit aperture ultrasound.

This paper investigates the possibility of flow estimation using spatio-temporal encoding of the transmissions in synthetic transmit aperture imaging (STA). The spatial encoding is based on a frequency division approach. In STA, a major disadvantage is that only a single transmitter (denoting single transducer element or a virtual source) is used in every transmission. The transmitted acoustic energy will be low compared to a conventional focused transmission in which a large part of the aperture is used. By using several transmitters simultaneously, the total transmitted energy can be increased. However, to focus the data properly, the signals originating from the different transmitters must be separated. To do so, the pass band of the transducer is divided into a number of subbands with disjoint spectral support. At every transmission, each transmitter is assigned one of the subbands. In receive, the signals are separated using a simple filtering operation. To attain high axial resolution, broadband spectra must be synthesized for each of the transmitters. By multiplexing the different waveforms on different transmitters over a number of transmissions, this can be accomplished. To further increase the transmitted energy, the waveforms are designed as linear frequency modulated signals. Therefore, the full excitation amplitude can be used during most of the transmission. The method has been evaluated for blood velocity estimation for several different velocities and incident angles. The program Field II was used. A 128-element transducer with a center frequency of 7 MHz was simulated. The 64 transmitting elements were used as the transmitting aperture and 128 elements were used as the receiving aperture. Four virtual sources were created in every transmission. By beamforming lines in the flow direction, directional data were extracted and correlated. Hereby, the velocity of the blood was estimated. The pulse repetition frequency was 16 kHz. Three different setups were investigated with flow angles of 45, 60, and 75 degrees with respect to the acoustic axis. Four different velocities were simulated for each angle at 0.10, 0.25, 0.50, and 1.00 m/s. The mean relative bias with respect to the peak flow for the three angles was less than 2%, 2%, and 4%, respectively.

Frequency division transmission imaging and synthetic aperture reconstruction.

In synthetic transmit aperture imaging only a few transducer elements are used in every transmission, which limits the signal-to-noise ratio (SNR). The penetration depth can be increased by using all transmitters in every transmission. In this paper, a method for exciting all transmitters in every transmission and separating them at the receiver is proposed. The coding is done by designing narrow-band linearly frequency modulated signals, which are approximately disjointed in the frequency domain and assigning one waveform to each transmitter. By designing a filterbank consisting of the matched filters corresponding to the excitation waveforms, the different transmitters can be decoded at the receiver. The matched filter of a specific waveform will allow information only from this waveform to pass through, thereby separating it from the other waveforms. This means that all transmitters can be used in every transmission, and the information from the different transmitters can be separated instantaneously. Compared to traditional synthetic transmit aperture (STA) imaging, in which the different transmitters are excited sequentially, more energy is transmitted in every transmission, and a better signal-to-noise-ratio is attained. The method has been tested in simulation, in which the resolution and contrast was compared to a standard synthetic transmit aperture system with a single sinusoid excitation. The resolution and contrast was comparable for the two systems. The method also has been tested using the experimental ultrasound scanner RASMUS. The resolution was evaluated using a string phantom. The method was compared to a conventional STA using both sinusoidal excitation and linear frequency modulated (FM) signals as excitation. The system using the FM signals and the frequency division approach yielded the same performance concerning both axial (of approximately equal to 3 wavelengths) and lateral resolution (of approximately equal to 4.5 wavelengths). A SNR measurement showed an increase in SNR of 6.5 dB compared to the system using the conventional STA method and FM signal excitation.