Low-frequency restoration

Post-Acquisition Hyperpolarized 29Silicon Magnetic Resonance Image Processing for Visualization of Colorectal Lesions Using a User-Friendly Graphical Interface

Medical imaging devices often use automated processing that creates and displays a self-normalized image. When improperly executed, normalization can misrepresent information or result in an inaccurate analysis. In the case of diagnostic imaging, a false positive in the absence of disease, or a negative finding when disease is present, can produce a detrimental experience for the patient and diminish their health prospects and prognosis. In many clinical settings, a medical technical specialist is trained to operate an imaging device without sufficient background information or understanding of the fundamental theory and processes involved in image creation and signal processing. Here, we describe a user-friendly image processing algorithm that mitigates user bias and allows for true signal to be distinguished from background. For proof-of-principle, we used antibody-targeted molecular imaging of colorectal cancer (CRC) in a mouse model, expressing human MUC1 at tumor sites. Lesion detection was performed using targeted magnetic resonance imaging (MRI) of hyperpolarized silicon particles. Resulting images containing high background and artifacts were then subjected to individualized image post-processing and comparative analysis. Post-acquisition image processing allowed for co-registration of the targeted silicon signal with the anatomical proton magnetic resonance (MR) image. This new methodology allows users to calibrate a set of images, acquired with MRI, and reliably locate CRC tumors in the lower gastrointestinal tract of living mice. The method is expected to be generally useful for distinguishing true signal from background for other cancer types, improving the reliability of diagnostic MRI.

Short-T2 MRI: Principles and recent advances

Among current modalities of biomedical and diagnostic imaging, MRI stands out by virtue of its versatile contrast obtained without ionizing radiation. However, in various cases, e.g., water protons in tissues such as bone, tendon, and lung, MRI performance is limited by the rapid decay of resonance signals associated with short transverse relaxation times T₂ or T₂*. Efforts to address this shortcoming have led to a variety of specialized short-T₂ techniques. Recent progress in this field expands the choice of methods and prompts fresh considerations with regard to instrumentation, data acquisition, and signal processing. In this review, the current status of short-T₂ MRI is surveyed. In an attempt to structure the growing range of techniques, the presentation highlights overarching concepts and basic methodological options. The most frequently used approaches are described in detail, including acquisition strategies, image reconstruction, hardware requirements, means of introducing contrast, sources of artifacts, limitations, and applications.

Multi-Rate Acquisition for Dead Time Reduction in Magnetic Resonance Receivers: Application to Imaging With Zero Echo Time

For magnetic resonance imaging of tissues with very short transverse relaxation times, radio-frequency excitation must be immediately followed by data acquisition with fast spatial encoding. In zero-echo-time (ZTE) imaging, excitation is performed while the readout gradient is already on, causing data loss due to an initial dead time. One major dead time contribution is the settling time of the filters involved in signal down-conversion. In this paper, a multi-rate acquisition scheme is proposed to minimize dead time due to filtering. Short filters and high output bandwidth are used initially to minimize settling time. With increasing time since the signal onset, longer filters with better frequency selectivity enable stronger signal decimation. In this way, significant dead time reduction is accomplished at only a slight increase in the overall amount of output data. Multi-rate acquisition was implemented with a two-stage filter cascade in a digital receiver based on a field-programmable gate array. In ZTE imaging in a phantom and in vivo, dead time reduction by multi-rate acquisition is shown to improve image quality and expand the feasible bandwidth while increasing the amount of data collected by only a few percent.

Exploring the bandwidth limits of ZTE imaging: Spatial response, out-of-band signals, and noise propagation

PURPOSE

Zero echo time (ZTE) imaging with single-pulse excitation is a fast, robust, and silent three-dimensional (3D) method for MRI of short T2 tissues. In this technique, algebraic reconstruction serves to fill gaps in the center of k-space due to finite acquisition dead time. The purpose of this study was to investigate the effect of this operation on depiction characteristics, noise behavior, and achievable bandwidth.

METHODS

The spatial response function (SRF) and noise covariance resulting from ZTE reconstruction were studied using formal analysis, simulations, and phantom experiments.

RESULTS

Three prominent limiting phenomena were identified: SRF behavior within the field of view, heightened sensitivity to out-of-band signal sources, and noise amplification. The related errors all appear as image distortions of low spatial frequency and are strongly attenuated upon the transition from one-dimensional projections to 3D image data. Relying on these observations, ZTE imaging was accomplished with a previously unreached gap size, permitting the depiction of a solid sample with T2 ≈ 25 µs at a bandwidth of 500 kHz.

CONCLUSION

The tightest bandwidth limits in ZTE arise from background signal and radiofrequency (RF) switching transients. Significant advances in ZTE performance will be afforded by faster transmit-receive (T/R) switching with negligible transients and RF coils free of background signal.

Hybrid-SPRITE MRI

In a FID based frequency encoding MRI experiment the central part of k-space is not generally accessible due to the probe dead time. This portion of k-space is however crucial for image reconstruction. SPRITE (Single Point Ramped Imaging with T1 Enhancement), SPI with a linearly ramped phase encode gradient, has been employed to image short relaxation time systems for many years with great success. It is a robust imaging method in significant measure because it provides acquisition of high quality k-space origin data. We propose a new sampling scheme, termed hybrid-SPRITE, combining phase and frequency encoding to ensure high quality images with reduced acquisition times, reduced gradient duty cycle and increased sensitivity. In hybrid-SPRITE, numerous time domain points are collected to assist image reconstruction. An Inverse Non-uniform Discrete Fourier Transform (INDFT) is employed in 1D applications. A pseudo-polar grid is exploited in 2D hybrid-SPRITE for rapid and accurate image reconstruction.

ZTE imaging in humans

PURPOSE

Zero echo time (ZTE) imaging is a robust and silent 3D radial technique suitable for direct MRI of tissues with very rapid transverse relaxation. Given its successful application on micro- and animal MRI systems, the purpose of this work is to enable and demonstrate ZTE imaging in humans using a whole-body magnet.

METHODS

A commercial 7 T MRI scanner was complemented by rapid high-power transmit-receive switches, a custom-built spectrometer, and a proton-free detector coil. With this setup, transmit-receive switching is achieved within 1 µs, radiofrequency (RF) excitation is performed in 3 µs, and digital bandpass filtering takes 5.3 µs, resulting in an effective dead time of only 5 µs.

RESULTS

ZTE imaging was performed at 250 and 500 kHz bandwidth with central k-space gaps of 1.2 and 2.5 Nyquist intervals and repetition times of 739 and 471 µs. The technique was applied for silent 3D imaging of the head and joints of human volunteers at an isotropic resolution down to 0.83 mm. A sound pressure level of 41 dB(A) was measured, which is a reduction of more than 40 dB(A) compared to gradient-switched MRI.

CONCLUSION

ZTE imaging in humans was demonstrated for the first time, enabled by dedicated, high-performing RF hardware.

Direct depiction of bone microstructure using MRI with zero echo time

This paper reports a proof of principle of direct depiction of trabecular bone microstructure in vitro by means of magnetic resonance imaging (MRI). Such depiction is achieved by (1)H imaging of water embedded in the bone matrix. The fast transverse relaxation of this compartment with T2(⁎) on the order of a few hundreds of microseconds is addressed by a three-dimensional MRI technique with zero echo time (ZTE). ZTE imaging at an isotropic spatial resolution of 56 μm is demonstrated in a trabecular bone specimen extracted from a bovine bone. In the MR images, the trabecular bone structure is clearly depicted and a high level of robustness against off-resonance artefacts is observed. The structural accuracy of the MR data is investigated by comparison with x-ray micro-computed tomography.

High-resolution ZTE imaging of human teeth

MRI with zero echo time (ZTE) is achieved by 3D radial centre-out encoding and hard-pulse RF excitation while the projection gradient is already on. Targeting short-T(2) samples, the efficient, robust and silent ZTE approach was implemented for high-bandwidth high-resolution imaging requiring particularly rapid transmit-receive switching and algebraic image reconstruction. The ZTE technique was applied to image extracted human teeth at 11.7T field strength, yielding detailed depictions with very good delineation of the mineralised dentine and enamel layers. ZTE results are compared with UTE (ultra-short echo time) MRI and micro-computed tomography (μCT), revealing significant differences in SNR and CNR yields. Compared to μCT, ZTE MRI appears to be less susceptible to artefacts caused by dental fillings and to offer superior sensitivity for the detection of early demineralisation and caries lesions.

A correction algorithm for undersampled images using dynamic segmentation and entropy based focus criterion

A post-processing technique is presented for correcting images undersampled in k-space. The method works by taking advantage of the image's background zeros (dynamically segmented through the application of a threshold) to extrapolate the missing k-space samples. The algorithm can produce good quality images from a small set of k-space frequencies with only a few iterations of simple matrix operations, using the image entropy as the focus criterion. It does not require any special patient preparation, extra pulse sequences, complex gradient programming or specialized hardware. This makes it a good candidate for any application that requires short scan times or where only few frequencies can be sampled.

Improved resolution and signal-to-noise ratio in MRI via enhanced signal digitization

The high frequency k-space data in magnetic resonance imaging is often poorly reproduced due to the finite dynamic range of an analog-to-digital converter. The magnitude of this digitization error can equal and even exceed the magnitude of the thermal noise. Under such conditions, attempts to increase image signal-to-noise ratio via signal averaging meet with diminishing success. Because the relative size of the digitization error increases at higher spatial frequencies, a reduction in image resolution is incurred as well. By adjusting the level of the analog signal sampled by the analog-to-digital converter during the course of an imaging experiment, the magnitude of the digitization artifact can be greatly reduced. The results of simulations and imaging experiments are presented which demonstrate that this strategy improves both the signal-to-noise ratio and resolution of magnetic resonance images.

MRS imaging using anatomically based k-space sampling and extrapolation

A comprehensive strategy for the acquisition, reconstruction, and postprocessing of MR spectroscopic images is presented. The reconstruction algorithm is the most critical component of this strategy. It is assumes that the desired image is spatially bounded, meaning that the desired image contains an object that is surrounded by a background of zeros. The reconstruction algorithm relies on prior knowledge of the background zeros for k-space extrapolation. This algorithm is a good candidate for proton MR spectroscopic image reconstruction because these images are often spatially bounded and prior knowledge of the zeros is easily obtained from a rapidly acquired high resolution conventional MRI. Although the reconstruction algorithm can be used with the standard 3DFT k-space distribution, a distribution that relies on anatomical features that are likely to occur in the spectroscopic image can produce better results. Prior knowledge of these anatomical features is also obtained from a conventional MRI. Finally, the postprocessing component of this strategy is valuable for reducing subcutaneous lipid contamination. Overall, the comprehensive approach presented here produces images that are better resolved than standard approaches without increasing acquisition time or reducing SNR. Examples using NAA data are provided.

Spectral extrapolation of spatially bounded images [MRI application]

A spectral extrapolation algorithm for spatially bounded images is presented. An image is said to be spatially bounded when it is confined to a closed region and is surrounded by a background of zeros. With prior knowledge of the spatial domain zeros, the extrapolation algorithm extends the image's spectrum beyond a known interval of low-frequency components. The result, which is referred to as the finite support solution, has space variant resolution; features near the edge of the support region are better resolved than those in the center. The resolution of the finite support solution is discussed as a function of the number of known spatial zeros and known spectral components. A regularized version of the finite support solution is included for handling the case where the known spectral components are noisy. For both the noiseless and noisy cases, the resolution of the finite support solution is measured in terms of its impulse response characteristics, and compared to the resolution of the zerofilled and Nyquist solutions. The finite support solution is superior to the zerofilled solution for both the noisy and noiseless data cases. When compared to the Nyquist solution, the finite support solution may be preferred in the noisy data case. Examples using medical image data are provided.

Phase-correction method for reduction of B0 instability artifacts

View-to-view phase shifts due to B0 instability are corrected by applying a phase correction to the low-spatial-frequency views. The algorithm described herein assumes that the signal intensity in the space outside of an object represents only noise. A set of phases for the low-spatial-frequency views are calculated by minimizing, in the image, the intensity of the induced artifacts in the space outside of the object.

Quantization improvement in MRI using dual quantizers

The quantization of magnetic resonance imaging (MRI) data can cause information loss due to quantizer/data mismatch. The authors address a method for improved quantization as well as techniques for measuring the improvement in such methods. A dual quantizer scheme is described and simulated which is fast and more accurately quantizes MRI data than conventional methods. The approach is to use two quantizers, one for the high-level data and one for the low-level data. This adaptive, dual quantization scheme is simple and provides significant improvements in image quality, especially for three-dimensional (3-D) acquisition. Results are given which show how well the low frequencies are represented and indicate the increased fidelity of the high-frequency components. These results show a significant improvement in signal-to-noise ratio as well as in detection tasks for both noiseless data and data which include varying amounts of system noise.

The relation of low frequency restoration methods to the Gerchberg-Papoulis algorithm

In magnetic resonance imaging, low frequency components can be allowed to saturate the analog to digital converter to reduce the quantization noise. These components can be estimated using least squares error estimation based low frequency restoration methods or the iterative Gerchberg-Papoulis algorithm. In this paper, we show the relationship between the closed form estimation methods and the iterative algorithm, propose a method for improving the speed of iteration, and discuss the advantages and disadvantages of two types of methods.

An efficient algorithm for MR image reconstruction without low spatial frequencies

It is demonstrated that if the image to be reconstructed is known to have some zero-valued pixels, the dynamic ranges can be better used by disregarding the largest signals and using signal restoration methods. Low-frequency and high-frequency signals are related, using the knowledge that some pixels are zero, by a set of linear equations in which the number of equations is equal to the number of zero pixels, and the number of unknowns is equal to the number of low-frequency signal samples rejected. An improved Fourier transform (FT), magnetic resonance (MR) imaging method based on a least-square-error (LSE) technique, and an efficient algorithm for signal restoration when the low-frequency components are discarded are presented. In this method, the regions of support in both the image domain and the frequency domain can have arbitrary shapes, and all zero pixels in the image domain can be taken into account. The algorithm has been tested on simulated and experimental data with acceptable results.