SoHyun Han

Temporal/spatial resolution improvement of in vivo DCE-MRI with compressed sensing-optimized FLASH

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) provides critical information regarding tumor perfusion and permeability by injecting a T(1) contrast agent, such as Gd-DTPA, and making a time-resolved measurement of signal increase. Both temporal and spatial resolutions are required to be high to achieve an accurate and reproducible estimation of tumor perfusion. However, the dynamic nature of the DCE experiment limits simultaneous improvement of temporal and spatial resolution by conventional methods. Compressed sensing (CS) has become an important tool for the acceleration of imaging times in MRI, which is achieved by enabling the reconstruction of subsampled data. Similarly, CS algorithms can be utilized to improve the temporal/spatial resolution of DCE-MRI, and several works describing retrospective simulations have demonstrated the feasibility of such improvements. In this study, the fast low angle shot sequence was modified to implement a Cartesian, CS-optimized, sub-Nyquist phase encoding acquisition/reconstruction with multiple two-dimensional slice selections and was tested on water phantoms and animal tumor models. The mean voxel-level concordance correlation coefficient for Ak(ep) values obtained from ×4 and ×8 accelerated and the fully sampled data was 0.87±0.11 and 0.83±0.11, respectively (n=6), with optimized CS parameters. In this case, the reduction of phase encoding steps made possible by CS reconstruction improved effectively the temporal/spatial resolution of DCE-MRI data using an in vivo animal tumor model (n=6) and may be useful for the investigation of accelerated acquisitions in preclinical and clinical DCE-MRI trials.

Robust MR assessment of cerebral blood volume and mean vessel size using SPION-enhanced ultrashort echo acquisition.

Intravascular superparamagnetic iron oxide nanoparticles (SPION)-enhanced MR transverse relaxation rates (∆R2(⁎) and ∆R2) are widely used to investigate in vivo vascular parameters, such as the cerebral blood volume (CBV), microvascular volume (MVV), and mean vessel size index (mVSI, ∆R2(⁎)/∆R2). Although highly efficient, regional comparison of vascular parameters acquired using gradient-echo based ∆R2(⁎) is hampered by its high sensitivity to magnetic field perturbations arising from air-tissue interfaces and large vessels. To minimize such demerits, we took advantage of the dual contrast property of SPION and both theoretically and experimentally verified the direct benefit of replacing gradient-echo based ∆R2(⁎) measurement with ultra-short echo time (UTE)-based ∆R1 contrast to generate the robust CBV and mVSI maps. The UTE acquisition minimized the local measurement errors from susceptibility perturbations and enabled dose-independent CBV measurement using the vessel/tissue ∆R1 ratio, while independent spin-echo acquisition enabled simultaneous ∆R2 measurement and mVSI calculation of the cortex, cerebellum, and olfactory bulb, which are animal brain regions typified by significant susceptibility-associated measurement errors.

Multi-GPU Reconstruction of Dynamic Compressed Sensing MRI

Magnetic resonance imaging (MRI) is a widely used in-vivo imaging technique that is essential to the diagnosis of disease, but its longer acquisition time hinders its wide adaptation in time-critical applications, such as emergency diagnosis. Recent advances in compressed sensing (CS) research have provided promising theoretical insights to accelerate the MRI acquisition process, but CS reconstruction also poses computational challenges that make MRI less practical. In this paper, we introduce a fast, scalable parallel CS-MRI reconstruction method that runs on graphics processing unit (GPU) cluster systems for dynamic contrast-enhanced (DCE) MRI. We propose a modified Split-Bregman iteration using a variable splitting method for CS-based DCE-MRI. We also propose a parallel GPU Split-Bregman solver that scales well across multiple GPUs to handle large data size. We demonstrate the validity of the proposed method on several synthetic and real DCE-MRI datasets and compare with existing methods.

Magnetic field anisotropy based MR tractography.

Non-invasive measurements of structural orientation provide unique information regarding the connectivity and functionality of fiber materials. In the present study, we use a capillary model to demonstrate that the direction of fiber structure can be obtained from susceptibility-induced magnetic field anisotropy. The interference pattern between internal and external magnetic field gradients carries the signature of the underlying anisotropic structure and can be measured by MRI-based water diffusion measurements. Through both numerical simulation and experiments, we found that this technique can determine the capillary orientation within 3°. Therefore, susceptibility-induced magnetic field anisotropy may be useful for an alternative tractography method when diffusion anisotropy is small at higher magnetic field strength without the need to rotate the subject inside the scanner.

Optimization of sparse phase encodings for variable repetition-delay turbo-spin echo (TSE) T1 measurements for preclinical applications

A variable repetition-delay (TR) spin echo sequence with repeated refocusing pulses, i.e., a variable TR turbo-spin echo (TSE), provides an attractive means of acquiring an accurate T1 map information that is free from gradient echo based artifacts such as magnetic field inhomogeneities particularly for ultra-high field (at 7T and above) preclinical applications. However, the applicability of multi-slice TSE sequences is often limited by signal distortion from T2 relaxation due to echo-train acquisitions for short T2 tissues, inter-slice cross talks and magnetization transfer (MT) from repetitive slice-selective 180° pulse, and extended scan times with multiple TR excitations. These TSE shortcomings are difficult to remedy for preclinical applications, where small sizes of target organs usually limit the slice-gap control with restricted parallel imaging capabilities. In this study, compressed-sensing-assisted turbo-spin echo (CS-TSE) acquisitions for variable TR T1 measurements at 7T preclinical scanner were implemented to reduce the echo-trains by sparse phase encodings. Following the sparse signal simulation and sampling scheme optimization, the measured T1 values from CS-TSE and TSE were compared for phantoms, ex vivo, and in vivo subjects. The phantom T1 values from CS-TSE and TSE were identical to those from the inversion recovery spin echo. For both ex vivo and in vivo multi-slice T1 mapping, the shortened echo-trains of CS-TSE relieved the T2 relaxation, reduced the inter-slice interferences of multi-slice acquisition, and made room for additional slice encodings while maintaining a shorter scan time than the conventional TSE at the expense of local image smoothness from CS regularizations.

Simulational validation of color magnetic particle imaging (cMPI)

Exploiting the field response of magnetic tracers, magnetic particle imaging (MPI) allows direct, local quantification of the tracer concentration in bulk structures. Here, we investigated the use of characteristic field response functions to spatially resolve the absolute concentration of multiple nanoparticle species by simulation. In particular, using various drive and selection field strengths, we devised color MPI (i.e. cMPI) to quantify and disentangle MPI signals from the mixed Langevin particles of variable concentration and magnetic susceptibility. Specifically, the drive field strength was optimized to distinguish individual field responses from differently sized iron-oxide nanoparticles without compromising the image quality. The proposed cMPI technique is implementable on an existing MPI setup and can be used to quantify biophysical parameters including size-dependent bio-distribution and altered magnetic property of particles. The current study result, simultaneous visualization of the multiple magnetic tracers, theoretically validates the potential feasibility of cMPI as a versatile biosensor and contrast imaging method.

UTE-ΔR2 -ΔR2 * combined MR whole-brain angiogram using dual-contrast superparamagnetic iron oxide nanoparticles.

The ability to visualize whole‐brain vasculature is important for quantitative in vivo investigation of vascular malfunctions in cerebral small vessel diseases, including cancer, stroke and neurodegeneration. Transverse relaxation‐based ΔR2 and ΔR2* MR angiography (MRA) provides improved vessel–tissue contrast in animal deep brain with the aid of intravascular contrast agents; however, it is susceptible to orientation dependence, air–tissue interface artifacts and vessel size overestimation. Dual‐mode MRA acquisition with superparamagnetic iron oxide nanoparticles (SPION) provides a unique opportunity to systematically compare and synergistically combine both longitudinal (R1) and transverse (ΔR2 and ΔR2*) relaxation‐based MRA. Through Monte Carlo (MC) simulation and MRA experiments in normal and tumor‐bearing animals with intravascular SPION, we show that ultrashort TE (UTE) MRA acquires well‐defined vascularization on the brain surface, minimizing air–tissue artifacts, and combined ΔR2 and ΔR2* MRA simultaneously improves the sensitivity to intracortical penetrating vessels and reduces vessel size overestimation. Consequently, UTE–ΔR2–ΔR2* combined MRA complements the shortcomings of individual angiograms and provides a strategy to synergistically merge longitudinal and transverse relaxation effects to generate more robust in vivo whole‐brain micro‐MRA. Copyright

Gradient-echo and spin-echo blood oxygenation level-dependent functional MRI at ultrahigh fields of 9.4 and 15.2 Tesla: BOLD fMRI at ultrahigh fields

Sensitivity and specificity of blood oxygenation level–dependent (BOLD) functional MRI (fMRI) is sensitive to magnetic field strength and acquisition methods. We have investigated gradient‐echo (GE)‐ and spin‐echo (SE)‐BOLD fMRI at ultrahigh fields of 9.4 and 15.2 Tesla.

Optimal sampling with prior information of the image geometry in microfluidic MRI

Recent advances in MRI acquisition for microscopic flows enable unprecedented sensitivity and speed in a portable NMR/MRI microfluidic analysis platform. However, the application of MRI to microfluidics usually suffers from prolonged acquisition times owing to the combination of the required high resolution and wide field of view necessary to resolve details within microfluidic channels. When prior knowledge of the image geometry is available as a binarized image, such as for microfluidic MRI, it is possible to reduce sampling requirements by incorporating this information into the reconstruction algorithm. The current approach to the design of the partial weighted random sampling schemes is to bias toward the high signal energy portions of the binarized image geometry after Fourier transformation (i.e. in its k-space representation). Although this sampling prescription is frequently effective, it can be far from optimal in certain limiting cases, such as for a 1D channel, or more generally yield inefficient sampling schemes at low degrees of sub-sampling. This work explores the tradeoff between signal acquisition and incoherent sampling on image reconstruction quality given prior knowledge of the image geometry for weighted random sampling schemes, finding that optimal distribution is not robustly determined by maximizing the acquired signal but from interpreting its marginal change with respect to the sub-sampling rate. We develop a corresponding sampling design methodology that deterministically yields a near optimal sampling distribution for image reconstructions incorporating knowledge of the image geometry. The technique robustly identifies optimal weighted random sampling schemes and provides improved reconstruction fidelity for multiple 1D and 2D images, when compared to prior techniques for sampling optimization given knowledge of the image geometry.

Gaussian Mixture Model-based Classification of DCE-MRI data For Identifying Diverse Tumor Microenvironments: Preliminary Results

Tumor hypoxia develops heterogeneously, affects radiation sensitivity and the development of metastases. Prognostic information derived from the in vivo characterization of the spatial distribution of hypoxic areas in solid tumors can be of value for radiation therapy planning and for monitoring the early treatment response. Tumor hypoxia is caused by an imbalance between the supply and consumption of oxygen. The tumor oxygen supply is inherently linked to its vasculature and perfusion which can be evaluated by dynamic contrast enhanced (DCE‐) MRI using the contrast agent Gd‐DTPA. Thus, we hypothesize that DCE‐MRI data may provide surrogate information regarding tumor hypoxia. In this study, DCE‐MRI data from a rat prostate tumor model were analysed with a Gaussian mixture model (GMM)‐based classification to identify perfused, hypoxic and necrotic areas for a total of ten tumor slices from six rats, of which one slice was used as training data for GMM classifications. The results of pattern recognition analyzes were validated by comparison to corresponding Akep maps defining the perfused area (0.84 ± 0.09 overlap), hematoxylin and eosin (H&E)‐stained tissue sections defining necrosis (0.64 ± 0.15 overlap) and pimonidazole‐stained sections defining hypoxia (0.72 ± 0.17 overlap), respectively. Our preliminary data indicate the feasibility of a GMM‐based classification to identify tumor hypoxia, necrosis and perfusion/permeability from non‐invasively acquired, in vivo DCE‐MRI data alone, possibly obviating the need for invasive procedures, such as biopsies, or exposure to radioactivity, such as positron emission tomography (PET) exams. Copyright