Clifton R. Haider

3D high temporal and spatial resolution contrast-enhanced MR angiography of the whole brain

Sensitivity encoding (SENSE) and partial Fourier techniques have been shown to reduce the acquisition time and provide high diagnostic quality images. However, for time‐resolved acquisitions there is a need for both high temporal and spatial resolution. View sharing can be used to provide an increased frame rate but at the cost of acquiring spatial frequencies over a duration longer than a frame time. In this work we hypothesize that a CArtesian Projection Reconstruction‐like (CAPR) technique in combination with 2D SENSE, partial Fourier, and view sharing can provide 1–2 mm isotropic resolution with sufficient temporal resolution to distinguish intracranial arterial and venous phases of contrast passage in whole‐brain angiography. In doing so, the parameter of “temporal footprint” is introduced as a descriptor for characterizing and comparing time‐resolved view‐shared pulse sequences. It is further hypothesized that short temporal footprint sequences have higher temporal fidelity than similar sequences with longer temporal footprints. The tradeoff of temporal footprint and temporal acceleration is presented and characterized in numerical simulations. Results from 11 whole‐brain contrast‐enhanced MR angiography studies with the new method with SENSE acceleration factors R = 4 and 5.3 are shown to provide images of comparable or higher diagnostic quality than the unaccelerated reference. Magn Reson Med 60:749–760, 2008.

Peripheral vasculature: high-temporal- and high-spatial-resolution three-dimensional contrast-enhanced MR angiography.

PURPOSE To prospectively evaluate the feasibility of performing high-spatial-resolution (1-mm isotropic) time-resolved three-dimensional (3D) contrast material-enhanced magnetic resonance (MR) angiography of the peripheral vasculature with Cartesian acquisition with projection-reconstruction-like sampling (CAPR) and eightfold accelerated two-dimensional (2D) sensitivity encoding (SENSE). MATERIALS AND METHODS All studies were approved by the institutional review board and were HIPAA compliant; written informed consent was obtained from all participants. There were 13 volunteers (mean age, 41.9; range, 27-53 years). The CAPR sequence was adapted to provide 1-mm isotropic spatial resolution and a 5-second frame time. Use of different receiver coil element sizes for those placed on the anterior-to-posterior versus left-to-right sides of the field of view reduced signal-to-noise ratio loss due to acceleration. Results from eight volunteers were rated independently by two radiologists according to prominence of artifact, arterial to venous separation, vessel sharpness, continuity of arterial signal intensity in major arteries (anterior and posterior tibial, peroneal), demarcation of origin of major arteries, and overall diagnostic image quality. MR angiographic results in two patients with peripheral vascular disease were compared with their results at computed tomographic angiography. RESULTS The sequence exhibited no image artifact adversely affecting diagnostic image quality. Temporal resolution was evaluated to be sufficient in all cases, even with known rapid arterial to venous transit. The vessels were graded to have excellent sharpness, continuity, and demarcation of the origins of the major arteries. Distal muscular branches and the communicating and perforating arteries were routinely seen. Excellent diagnostic quality rating was given for 15 (94%) of 16 evaluations. CONCLUSION The feasibility of performing high-diagnostic-quality time-resolved 3D contrast-enhanced MR angiography of the peripheral vasculature by using CAPR and eightfold accelerated 2D SENSE has been demonstrated.

Sparse-CAPR: Highly accelerated 4D CE-MRA with parallel imaging and nonconvex compressive sensing

Cartesian Acquisition with Projection‐Reconstruction‐like sampling is a SENSE‐type parallel 3DFT acquisition paradigm for 4D contrast‐enhanced magnetic resonance angiography that has been demonstrated capable of providing high spatial and temporal resolution, diagnostic‐quality images at very high acceleration rates. However, Cartesian Acquisition with Projection–Reconstruction‐like sampling images are typically reconstructed online using Tikhonov regularization and partial Fourier methods, which are prone to exhibit noise amplification and undersampling artifacts when operating at very high acceleration rates. In this work, a sparsity‐driven offline reconstruction framework for Cartesian Acquisition with Projection‐Reconstruction‐like sampling is developed and demonstrated to consistently provide improvements over the currently‐employed reconstruction strategy against these ill‐effects. Moreover, the proposed reconstruction strategy requires no changes to the existing Cartesian Acquisition with Projection–Reconstruction‐like sampling acquisition protocol, and an efficient numerical optimization and hardware system are described that allow for a 256 × 160 × 80 volume contrast‐enhanced magnetic resonance angiography volume to be reconstructed from an eight‐channel data set in less than 2 min. Magn Reson Med, 2011.

Controlled experimental study depicting moving objects in view-shared time-resolved 3D MRA

Various methods have been used for time‐resolved contrast‐enhanced magnetic resonance angiography (CE‐MRA), many involving view sharing. However, the extent to which the resultant image time series represents the actual dynamic behavior of the contrast bolus is not always clear. Although numerical simulations can be used to estimate performance, an experimental study can allow more realistic characterization. The purpose of this work was to use a computer‐controlled motion phantom for study of the temporal fidelity of three‐dimensional (3D) time‐resolved sequences in depicting a contrast bolus. It is hypothesized that the view order of the acquisition and the selection of views in the reconstruction can affect the positional accuracy and sharpness of the leading edge of the bolus and artifactual signal preceding the edge. Phantom studies were performed using dilute gadolinium‐filled vials that were moved along tabletop tracks by a computer‐controlled motor. Several view orders were tested using view‐sharing and Cartesian sampling. Compactness of measuring the k‐space center, consistency of view ordering within each reconstruction frame, and sampling the k‐space center near the end of the temporal footprint were shown to be important in accurate portrayal of the leading edge of the bolus. A number of findings were confirmed in an in vivo CE‐MRA study. Magn Reson Med, 2009.

Practical nonconvex Compressive Sensing reconstruction of highly-accelerated 3D parallel MR angiograms

In this work, a nonconvex Compressive Sensing model targeted at true 3D reconstructions of highly undersampled MR angiograms acquired with parallel imaging is proposed. When combined with the Max-CAPR acquisition sequence, it is demonstrated that high quality, non-view-shared, 3D images of the contrast-filled neurovasculature can be acquired (at acceleration factors exceeding the number of coils) in just over 2 seconds and reconstructed in as few as 14 minutes on a high-performance workstation.

Undersampled elliptical centric view-order for improved spatial resolution in contrast-enhanced MR angiography.

Although contrast‐enhanced MR angiography (CE‐MRA) has been successfully developed into a routine clinical imaging technique, there is still need for improved spatial resolution in a given acquisition time. Undersampled projection reconstruction (PR) techniques maintain spatial resolution with reduced scan times, and the elliptical centric (EC) view order provides high quality arterial phase images without venous contamination. In this work, we present a hybrid elliptical centric–projection reconstruction (EC‐PR) technique to provide spatial resolution improvement over standard EC in a given time. The k‐space sampling was performed by undersampling the periphery of the kY‐kZ phase encoding plane of an EC view order in a PR like manner. The sampled views were maintained on a rectilinear grid, and thus reconstructed by standard 3DFT. The non‐sampled views were compensated either by zero‐filling or performing a 2D homodyne reconstruction. Compared to a fully sampled k‐space, the EC‐PR sequence acquired in the same scan time provides a resolution improvement of about two, as shown by point spread function analysis and phantom experiments. The hypothesis that EC‐PR provides improved resolution while retaining diagnostically adequate SNR was tested in 11 CE‐MRA studies of the popliteal and carotid arteries and shown to be true (P < 0.03). Magn Reson Med 55:000–000, 2006.

Intrinsic Signal Amplification in the Application of 2D SENSE Parallel Imaging to 3D Contrast-Enhanced Elliptical Centric MRA and MRV

The relative signal‐to‐noise ratio (SNR) provided by 2D sensitivity encoding (SENSE) when applied to 3D contrast‐enhanced MR angiography (CE‐MRA) is studied. If an elliptical centric phase‐encoding order is used to map the waning magnetization of the contrast bolus to k‐space, the application of SENSE will reduce the degree of k‐space signal modulation, providing a signal amplification A over corresponding nonaccelerated acquisitions. This offsets the SNR loss in R‐accelerated SENSE due to and the geometry (g) factor. The theoretical bound on A is R and is reduced from this depending on the properties of the bolus profile and the duration over which it is imaged. In this work a signal amplification of 1.14–1.23 times that of nonvascular background tissue is demonstrated in a study of 20 volunteers using R = 4 2D SENSE whole‐brain MR venography (MRV). The effects of a nonuniform g‐factor and inhomogeneity of background tissue are accounted for. The observed amplification compares favorably with the value of 1.31 predicted numerically from a measured bolus curve. Magn Reson Med 58:855–864, 2007.

High-performance 3D compressive sensing MRI reconstruction using many-core architectures

Compressive sensing (CS) describes how sparse signals can be accurately reconstructed from many fewer samples than required by the Nyquist criterion. Since MRI scan duration is proportional to the number of acquired samples, CS has been gaining significant attention in MRI. However, the computationally intensive nature of CS reconstructions has precluded their use in routine clinical practice. In this work, we investigate how different throughput-oriented architectures can benefit one CS algorithm and what levels of acceleration are feasible on different modern platforms. We demonstrate that a CUDA-based code running on an NVIDIA Tesla C2050 GPU can reconstruct a 256 × 160 × 80 volume from an 8-channel acquisition in 19 seconds, which is in itself a significant improvement over the state of the art. We then show that Intels Knights Ferry can perform the same 3D MRI reconstruction in only 12 seconds, bringing CS methods even closer to clinical viability.

High temporal and spatial resolution 3D time-resolved contrast-enhanced magnetic resonance angiography of the hands and feet

Methods are described for generating 3D time‐resolved contrast‐enhanced magnetic resonance (MR) angiograms of the hands and feet. Given targeted spatial resolution and frame times, it is shown that acceleration of about one order of magnitude or more is necessary. This is obtained by a combination of 2D sensitivity encoding (SENSE) and homodyne (HD) acceleration methods. Image update times from 3.4–6.8 seconds are provided in conjunction with view sharing. Modular receiver coil arrays are described which can be designed to the targeted vascular region. Images representative of the technique are generated in the vasculature of the hands and feet in volunteers and in patient studies. J. Magn. Reson. Imaging 2011;.

Time-of-Arrival Mapping at Three-dimensional Time-resolved Contrast-enhanced MR Angiography

This study was HIPAA compliant and institutional review board approved, and informed consent was obtained from all volunteers. The authors describe a method for generating a time-of-arrival (TOA) map of intravenously administered contrast material, as observed in a time series of three-dimensional (3D) contrast material-enhanced magnetic resonance (MR) angiograms. The method may enable visualization and interpretation, on one 3D image, of the temporal enhancement patterns that occur in the vasculature. Colorization of TOA values may further aid interpretation. The quality of the results depends not only on the adequacy of the frame rate, spatial resolution, and signal-to-noise ratio of the MR image acquisition method but also on the accuracy and clarity with which the leading edge of the contrast material bolus is depicted. The criteria for optimizing these parameters are described. The TOA mapping technique is demonstrated by using vascular studies of the hands, brain, and lower leg regions.