High Frame Rate Imaging Method Research Articles

A phase-shifting method for computation reduction for high-frame-rate imaging

High-frame rate (HFR) imaging using steered plane wave (SPW) or limited-diffraction beam has a high-temporal resolution and thus has found many applications. To further study the HFR imaging methods, computer simulations were performed. However, the simulations require a large number of computations, especially for 3D imaging with 2D array transducers for a large imaging volume. In this paper, a phase-shifting method was developed to reduce the number of computations. In the method, the grid points of the transmit and receive beams were calculated at 1-mm interval in the depth direction that is perpendicular to the transducer surface. The interval is much larger than the 1/4 of the 0.58-mm wavelength required for an accurate interpolation for millions of random scatterers in pulse-echo response without aliasing. Since the HFR imaging uses either SPW or LDB, the wave vectors of these beams are fixed at each frequency. Due to the fact that the amplitude of ultrasound beams changes very little over a couple of wavelengths, the interpolation in the depth direction was replaced with a phase shift. Results show that images reconstructed with the phase-shifting method removed the artifacts caused by aliasing when conventional tri-linear interpolations were used for 3D imaging.

3D ultrasound imaging in frequency domain based on concepts of array beam and synthetic aperture

The high frame rate (HFR) imaging method has the ability to achieve a high frame rate. In this method, only one transmission is required to construct a frame of image. In our previous work, using a moved one-dimensional (1D) array transducer, a three-dimensional (3D) ultrasound imaging method in frequency domain was developed. This imaging method was designed based on the concepts of array beam and synthetic aperture, which can simplify the two-dimensional (2D) array transducer. In this paper, based on array beam and synthetic aperture, the HFR imaging method is demonstrated from a novel view. From this view, the relationship between the HFR imaging method and synthetic aperture is established with the weighting function of array beam. Besides, the HFR imaging method, the imaging method with a moved 1D array transducer, and the synthetic aperture imaging method with a moved single element transducer are unified in the same analytical method with different weighting functions. The same frequency domain signal processing flow can be applied to these imaging methods. Comparisons to these imaging methods are implemented with simulations. Simulation results show that, in the imaging depth of 45 mm, the resolutions calculated as the total width of the −6 dB main lobe in x-direction are 1.099 mm, 1.056 mm and 0.596 mm for the methods with 1D transducer, 2D transducer and the single element transducer, respectively. The resolution in y-direction is 1.054 mm for the methods with 2D transducer, and 0.565 mm, 0.593 mm for the 1D and single element transducers, respectively. The resolutions in z-direction are 0.493 mm, 0.451 mm and 0.452 mm for the 2D, 1D and single element transducers, respectively. The resolution in the moved-direction is improved with a moved transducer, but the contrast of the image is decreased.

High temporal and spatial resolution ultrasonic imaging and application to cardiovascular imaging

Recently, high frame rate ultrasound has become used for various imaging scenarios. It creates multiple scan lines at the same time using unfocused transmit beam to reduce the number of transmissions per image frame. Consequently, high frame rate ultrasound can realize a frame rate of over 1 kHz, but the lateral spatial resolution is degraded by using unfocused transmit beams, such as plane and diverging waves. To overcome such a problem, we have developed novel beamforming methods based on adaptive procedures, such as phase coherence imaging and minimum variance beamforming. As a result, high temporal and spatial resolution could be realized simultaneously. The developed high frame rate imaging methods were applied to imaging of cardiovascular dynamics. Blood flow in the carotid artery and cardiac left ventricle were imaged at frame rates of 2500 Hz and 6250 Hz with plane wave and diverging wave transmissions, respectively, and vortices in the carotid bifurcation and left ventricle could be observed. The...

Real-time distortion correction of spiral and echo planar images using the gradient system impulse response function.

MRI-guided interventions demand high frame rate imaging, making fast imaging techniques such as spiral imaging and echo planar imaging (EPI) appealing. In this study, we implemented a real-time distortion correction framework to enable the use of these fast acquisitions for interventional MRI. Distortions caused by gradient waveform inaccuracies were corrected using the gradient impulse response function (GIRF), which was measured by standard equipment and saved as a calibration file on the host computer. This file was used at runtime to calculate the predicted k-space trajectories for image reconstruction. Additionally, the off-resonance reconstruction frequency was modified in real time to interactively deblur spiral images. Real-time distortion correction for arbitrary image orientations was achieved in phantoms and healthy human volunteers. The GIRF-predicted k-space trajectories matched measured k-space trajectories closely for spiral imaging. Spiral and EPI image distortion was visibly improved using the GIRF-predicted trajectories. The GIRF calibration file showed no systematic drift in 4 months and was demonstrated to correct distortions after 30 min of continuous scanning despite gradient heating. Interactive off-resonance reconstruction was used to sharpen anatomical boundaries during continuous imaging. This real-time distortion correction framework will enable the use of these high frame rate imaging methods for MRI-guided interventions. Magn Reson Med 75:2278-2285, 2016. © 2015 Wiley Periodicals, Inc.

Quantitative assessment of effects of phase aberration and noise on high-frame-rate imaging

The goal of this paper is to quantitatively study effects of phase aberration and noise on high-frame-rate (HFR) imaging using a set of traditional and new parameters. These parameters include the traditional −6-dB lateral resolution, and new parameters called the energy ratio (ER) and the sidelobe ratio (SR). ER is the ratio between the total energy of sidelobe and the total energy of mainlobe of a point spread function (PSF) of an imaging system. SR is the ratio between the peak value of the sidelobe and the peak value of the mainlobe of the PSF. In the paper, both simulation and experiment are conducted for a quantitative assessment and comparison of the effects of phase aberration and noise on the HFR and the conventional delay-and-sum (D&S) imaging methods with the set of parameters. In the HFR imaging method, steered plane waves (SPWs) and limited-diffraction beams (LDBs) are used in transmission, and received signals are processed with the Fast Fourier Transform to reconstruct images. In the D&S imaging method, beams focused at a fixed depth are used in transmission and dynamically focused beams are used in reception for image reconstruction.The simulation results show that the average differences between the −6-dB lateral beam widths of the HFR imaging and the D&S imaging methods are −0.1337mm for SPW and −0.1481mm for LDB, which means that the HFR imaging method has a higher lateral image resolution than the D&S imaging method since the values are negative. In experiments, the average differences are also negative, i.e., −0.2804mm for SPW and −0.3365mm for LDB. The results for the changes of ER and SR between the HFR and the D&S imaging methods have negative values, too. After introducing phase aberration and noise, both simulations and experiments show that the HFR imaging method has also less change in the −6-dB lateral resolution, ER, and SR as compared to the conventional D&S imaging method. This means that the HFR imaging method is less sensitive to the phase aberration and noise.Based on the study of the new parameters on the HFR and the D&S imaging methods, it is expected that the new parameters can also be applied to assess quality of other imaging methods.

Motion Artifacts of Extended High Frame Rate Imaging

Based on the high frame rate (HFR) imaging method developed in our lab, an extended high frame rate imaging method with various transmission schemes was developed recently. In this method, multiple, limited-diffraction array beams or steered plane wave transmissions are used to increase image resolution and field of view as well as to reduce sidelobes. Furthermore, the multiple, limited-diffraction array beam transmissions can be approximated with square-wave aperture weightings, allowing one or two transmitters to be used with a multielement array transducer to simplify imaging systems. By varying the number of transmissions, the extended HFR imaging method allows a continuous trade-off between image quality and frame rate. Because multiple transmissions are needed to obtain one frame of image for the method, motion could cause phase misalignment and thus produce artifacts, reducing image contrast and resolution and leading to an inaccurate clinical interpretation of images. Therefore, it is important to study how motion affects the method and provide a useful guidance of using the method properly in various applications. In this paper, computer simulations, in vitro and in vivo experiments were performed to study the effects of motion on the method in different conditions. Results show that a number of factors may affect the motion effects. However, it was found that the extended HFR imaging method is not sensitive to the motions commonly encountered in the clinical applications, as is demonstrated by an in vivo heart experiment, unless the number of transmissions is large and objects are moving at a high velocity near the surface of a transducer.

Effects of Phase Aberration and Noise on Extended High Frame Rate Imaging

Based on the high frame rate (HFR) imaging theories, an extended HFR imaging method has been developed recently in our lab where multiple limited-diffraction array beams or steered plane waves are used in transmissions to reconstruct a high quality image of an equivalent dynamic focusing in both transmissions and receptions. The method has the potential to simplify imaging systems because the fast Fourier transform and square-wave aperture weightings can be used. The method is also flexible in using different numbers of transmissions for a continuous trade-off between image quality and frame rate. In this paper, we study the effects of phase aberration and noise on the extended HFR imaging method with in vitro experiments and compare the results with those obtained with a conventional delay-and-sum (D&S) method of a fixed-transmission focus and a dynamically-focused reception. In the experiments, an ATS539 tissue-mimicking phantom and an Acuson V2 phase array transducer (128 elements, 2.5 MHz, and 0.15-mm pitch) were used. The transducer was driven by a homemade general-purpose HFR imaging system that was capable of producing both the limited-diffraction array beams and steeredplane waves and echo data were acquired with the same system and then transferred to a personal computer via a universal serial bus (USB) 2.0 link for image reconstructions. The phase aberration was introduced by adding random phase shifts to both transmission and reception beams. The random noise was added to the received radiofrequency echo data. Results show that the phase aberration and noise degrade both the extended HFR and the conventional delay-and-sum (D&S) imaging method. However, images reconstructed with the extended HFR imaging method have an overall higher quality than those with the D&S method given the phase aberration and noise models studied.

Extract echo signal in HFR imaging system

Ultrasonic high frame rate (HFR) method was developed by Jianyu Lu in 1997. This method constructs images in frequency domain; the spectrum of space domain objects is calculated based on the spectrum of echo signals, and the image is obtained by using IFFT (inverse fast Fourier transform). Because of this imaging process, the echo signal from each individual scatterer cannot be obtained directly. In this paper, we present a method, which can pick up the echo signal of every scatterer in HFR imaging method. Computer simulation has been performed to verify our method. With the new method we have developed, many other imaging methods, such as Doppler blood flow and elastic imaging, can be easily combined with HFR method to give more information of the imaged objects.

Design and construction of a high frame rate imaging system

A new high frame rate imaging method has been developed recently [Jian-yu Lu, ‘‘2D and 3D high frame rate imaging with limited diffraction beams,’’ IEEE Trans. Ultrason. Ferroelectr. Freq. Control 44, 839–856 (1997)]. This method may have a clinical application for imaging of fast moving objects such as human hearts, velocity vector imaging, and low-speckle imaging. To implement the method, an imaging system has been designed. The system consists of one main printed circuit board (PCB) and 16 channel boards (each channel board contains 8 channels), in addition to a set-top box for connections to a personal computer (PC), a front panel board for user control and message display, and a power control and distribution board. The main board contains a field programmable gate array (FPGA) and controls all channels (each channel has also an FPGA). We will report the analog and digital circuit design and simulations, multiplayer PCB designs with commercial software (Protel 99), PCB signal integrity testing and system RFI/EMI shielding, and the assembly and construction of the entire system. [Work supported in part by Grant 5RO1 HL60301 from NIH.]

Increasing field of view of high frame rate ultrasonic imaging

A high frame rate imaging method has been developed recently based on the theory of limited diffraction beams for medical diagnosis. This method constructs images in a direction normal to the transducer surface. In this report, the high frame rate method is extended so that it will not only construct images in the normal direction, but also in any oblique directions. This increases the image construction area because images from several oblique angles can be combined. A new equation has been derived for this extension. A computer simulation was performed. Results show that the field of view of images constructed with the extended method is increased from about 15.74 deg to about 96.57 deg with a linear array of a width (foot print) of about 18.288 mm after combining images from only seven angles (the depth of the images is assumed to be 66.12 mm). This work is significant because a large field of view with a small transducer footprint is desirable for high frame rate heart imaging. [Work supported in part by Grant No. HL60301 from NIH.]

High frame rate imaging with a small number of array elements

Recently, a high frame rate imaging method has been developed to construct either 2-D or 3-D images (about 3750 frames or volumes/s at a depth of about 200 mm in biological soft tissues because only one transmission is needed). The signal-to-noise ratio (SNR) is high using this method because all array elements are used in transmission and the transmit beams do not diverge. In addition, imaging hardware with the new method can be greatly simplified. Theoretically, the element spacing (distance between the centers of two neighboring elements) of an array should be lambda/2, where lambda is the wavelength, to avoid grating lobes in imaging. This requires an array of a large number of elements, especially, for 3-D imaging in which a 2-D array is needed. In this paper, we study quantitatively the relationship between the quality of images constructed with the new method and the element spacing of array transducers. In the study, two linear arrays were used. One has an aperture of 18.288 mm, elevation dimension of 12.192 mm, a center frequency of 2.25 MHz, and 48 elements (element spacing is 0.381 mm or 0.591 lambda). The other has a dimension of 38.4 mmx10 mm, a center frequency of 2.5 MHz, and 64 elements (0.6 mm or 1.034 lambda element spacing). Effective larger element spacings were obtained by combining signals from adjacent elements. Experiments were performed with both the new and the conventional delay-and-sum methods. Results show that resolution of constructed images is not affected by the reduction of a number of elements, but the contrast of images is decreased dramatically when the element spacing is larger than about 2.365 lambda for objects that are not too close to the transducers. This suggests that an array of about 2.365 lambda spacing can be used with the new method. This may reduce the total number of elements of a fully sampled 128x128 array (0.5 lambda spacing) from 16384 to about 732 considering that the two perpendicular directions of a 2-D array are independent (ignoring the larger element spacing in diagonal directions of 2-D arrays).