White matter maturation reshapes structural connectivity in the late developing human brain.

From toddler to late teenager, the macroscopic pattern of axonal projections in the human brain remains largely unchanged while undergoing dramatic functional modifications that lead to network refinement. These functional modifications are mediated by increasing myelination and changes in axonal diameter and synaptic density, as well as changes in neurochemical mediators. Here we explore the contribution of white matter maturation to the development of connectivity between ages 2 and 18 y using high b-value diffusion MRI tractography and connectivity analysis. We measured changes in connection efficacy as the inverse of the average diffusivity along a fiber tract. We observed significant refinement in specific metrics of network topology, including a significant increase in node strength and efficiency along with a decrease in clustering. Major structural modules and hubs were in place by 2 y of age, and they continued to strengthen their profile during subsequent development. Recording resting-state functional MRI from a subset of subjects, we confirmed a positive correlation between structural and functional connectivity, and in addition observed that this relationship strengthened with age. Continuously increasing integration and decreasing segregation of structural connectivity with age suggests that network refinement mediated by white matter maturation promotes increased global efficiency. In addition, the strengthening of the correlation between structural and functional connectivity with age suggests that white matter connectivity in combination with other factors, such as differential modulation of axonal diameter and myelin thickness, that are partially captured by inverse average diffusivity, play an increasingly important role in creating brain-wide coherence and synchrony.

Improving diffusion MRI using simultaneous multi-slice echo planar imaging.

In diffusion MRI, simultaneous multi-slice single-shot EPI acquisitions have the potential to increase the number of diffusion directions obtained per unit time, allowing more diffusion encoding in high angular resolution diffusion imaging (HARDI) acquisitions. Nonetheless, unaliasing simultaneously acquired, closely spaced slices with parallel imaging methods can be difficult, leading to high g-factor penalties (i.e., lower SNR). The CAIPIRINHA technique was developed to reduce the g-factor in simultaneous multi-slice acquisitions by introducing inter-slice image shifts and thus increase the distance between aliased voxels. Because the CAIPIRINHA technique achieved this by controlling the phase of the RF excitations for each line of k-space, it is not directly applicable to single-shot EPI employed in conventional diffusion imaging. We adopt a recent gradient encoding method, which we termed "blipped-CAIPI", to create the image shifts needed to apply CAIPIRINHA to EPI. Here, we use pseudo-multiple replica SNR and bootstrapping metrics to assess the performance of the blipped-CAIPI method in 3× simultaneous multi-slice diffusion studies. Further, we introduce a novel image reconstruction method to reduce detrimental ghosting artifacts in these acquisitions. We show that data acquisition times for Q-ball and diffusion spectrum imaging (DSI) can be reduced 3-fold with a minor loss in SNR and with similar diffusion results compared to conventional acquisitions.

Projective imaging of pulsatile flow with magnetic resonance.

Noninvasive angiography with magnetic resonance is demonstrated. Signal arising in all structures except vessels that carry pulsatile flow is eliminated by means of velocity-dependent phase contrast, electrocardiographic gating, and image subtraction. Background structures become in effect transparent, enabling the three-dimensional vascular tree to be imaged by projection to a two-dimensional image plane. Image acquisition and processing are accomplished with entirely conventional two-dimensional Fourier transform magnetic resonance imaging techniques. When imaged at 0.6 tesla, vessels 1 to 2 millimeters in diameter are routinely detected in a 50-centimeter field of view with data acquisition times less than 15 minutes. Studies of normal and pathologic anatomy are illustrated in human subjects.

Use of gadolinium-DTPA as a myocardial perfusion agent: potential applications and limitations for magnetic resonance imaging.

To establish the effect of the paramagnetic contrast agent gadolinium diethylenetriaminepentaacetic acid ([Gd]DTPA) on myocardial magnetic resonance relaxation parameters T1 and T2, and its relationship to myocardial perfusion, we administered [Gd] DTPA 0.2 mM/kg to two groups of dogs. Group I had severe, resting myocardial ischemia induced by coronary occlusion, followed in 2 min by [Gd]DTPA infusion and heart excision 1 min later. Group II had a variable reduction in blood flow. In Group II the coronary vasodilator dipyridamole was infused to enhance blood flow to the normal myocardium before [Gd]DTPA was given. In Group I [Gd]DTPA caused a significant difference in T1 between the normal and severely ischemic zones; changes in T1 correlated with the severity of myocardial ischemia. Although vasodilatation delivered more Gd-DTPA to the normal myocardium in Group II, the lack of further decrease in T1 suggested that it was cleared more rapidly. Thus, [Gd]DTPA permits the detection and characterization of severe, resting myocardial ischemia by magnetic resonance techniques. Using the experimental techniques described in this study, less severe flow differences caused by vasodilatation and resultant hyperemia are not detected.

Intramural mechanics of the human tongue in association with physiological deformations.

Contraction of the tongue musculature during speech and swallowing is associated with characteristic patterns of tissue deformation. In order to quantify local deformation (strain) in the human tongue, we used a non-invasive NMR tagging technique that represents tissue as discrete deforming elements. Subjects were studied with a fast gradient echo pulse sequence (TR,TE 2.3/0.8 ms, slice thickness 10 mm, and effective spatial resolution 1.3x1.3 mm). Individual elements were defined by selectively supersaturating bands of magnetic spills in resting tongue tissue along the antero-posterior and superior inferior directions of the mid-sagittal plane, resulting in a rectilinear square grid. Axial and shear strains relative to the rest condition were determined for each clement and represented by two-dimensional surface strain maps. During forward protrusion, the anterior tongue underwent positive antero posterior strain (elongation) (maximum 200%) and symmetrical negative medial lateral and superior inferior strain (contraction). During sagittal curl directed to the hard palate, the tongue exhibited positive asymmetrical antero posterior strain (maximum 160%) that increased radially as a function of distance from the center of curvature (r = 0.9216, p<0.0005), and commensurate negative strain in the medial lateral direction. Similarly, the magnitude of anterior posterior strain during left-directed tongue curl was proportional to the distance from the curved inner surface (r = O.8978, p<0.0005). We conclude that the regulation of tongue position for the motions studied was related to regional activation of the intrinsic lingual musculature.

Diffusion spectrum magnetic resonance imaging (DSI) tractography of crossing fibers.

MRI tractography is the mapping of neural fiber pathways based on diffusion MRI of tissue diffusion anisotropy. Tractography based on diffusion tensor imaging (DTI) cannot directly image multiple fiber orientations within a single voxel. To address this limitation, diffusion spectrum MRI (DSI) and related methods were developed to image complex distributions of intravoxel fiber orientation. Here we demonstrate that tractography based on DSI has the capacity to image crossing fibers in neural tissue. DSI was performed in formalin-fixed brains of adult macaque and in the brains of healthy human subjects. Fiber tract solutions were constructed by a streamline procedure, following directions of maximum diffusion at every point, and analyzed in an interactive visualization environment (TrackVis). We report that DSI tractography accurately shows the known anatomic fiber crossings in optic chiasm, centrum semiovale, and brainstem; fiber intersections in gray matter, including cerebellar folia and the caudate nucleus; and radial fiber architecture in cerebral cortex. In contrast, none of these examples of fiber crossing and complex structure was identified by DTI analysis of the same data sets. These findings indicate that DSI tractography is able to image crossing fibers in neural tissue, an essential step toward non-invasive imaging of connectional neuroanatomy.

Magnetic resonance imaging of myocardial kinematics. Technique to detect, localize, and quantify the strain rates of the active human myocardium.

A magnetic resonance imaging (MRI) method is presented to detect, localize, and quantify myocardial kinematics by measuring the material rate-of-strain tensor at each pixel in gated NMR images of the heart. The immediate, local effect of muscular activity is self-deformation, and the strain tensor is the basic mathematical device by which such deformation may be quantified. The present method, called "strain-phase" MRI (SP-MRI), entails four steps: (1) the velocity of the myocardium is encoded by means of a set of motion-sensitive NMR image acquisitions, one image per velocity component; (2) the spatial derivatives of the velocity are computed at each pixel; (3) the velocity-derivative data are combined to compute an approximation of the strain-rate tensor of the myocardium at each pixel; and (4) the strain-rate tensor data are simplified to produce a color-coded functional image which represents strain-rate components which are of particular biomedical interest in the myocardium. We present a quantitative SP-MRI methodology suited to conventional MRI, and in addition present an "echo-planar" methodology, able to produce qualitative functional images of myocardial kinematics at almost real-time speeds. Two-dimensional strain-phase MRI data acquired in normal human subjects are presented. These data demonstrate the practicability of SP-MRI in vivo, that SP-MRI resolves myocardial kinematics at the single-pixel scale, having resolution comparable to that of conventional MRI, and that SP-MRI data may have a signal-to-noise ratio up to 50% as great as that of the conventional MRI data from which they are produced. SP-MRI measurements of the local instantaneous strain rates in the human left ventricular myocardium are quantitatively consistent with known transmural average values of myocardial strain.

Rapid three-dimensional angiography with undersampled MR imaging.

Techniques for subtraction angiography with magnetic resonance imaging have been extended from two to three dimensions, and a novel method that reduces the expected data acquisition time by at least an order of magnitude is presented. Electrocardiogram-gated three-dimensional (3D) images are acquired by Fourier transform technique, and flow contrast is obtained by subtracting pairs of images acquired at different points in the cardiac cycle. The vascular tree is shown in 3D perspective by means of a surface detection and a 3D display program. Isotropic 3D angiography requires determining the disposition of the blood vessels in a matrix of cubical voxels. Using orthodox Fourier transform technique, for an image matrix with 256 voxels to the edge, a data acquisition with 256 X 256 = 65 K phase-encodings would be needed. If gated, this would require approximately 1 day. In this study we abbreviate the data acquisition by doing only 1/64 of the usual set of phase-encoding gradient pulses. Spatial resolution is undiminished, but aliasing or "wraparound" results in each of the two phase-encoded coordinates of the 3D image. This aliasing is rectified in a two stage process. First, 64 copies of the undersampled 3D arteriogram are juxtaposed in a two-dimensional grid pattern. This assembles many copies of the complete vascular tree. Because they occupy only a small fraction of ambient volume, these copies are unlikely to overlap or collide with one another. Second, a single copy of the vascular tree is isolated by a surface detection program that takes advantage of the fact that the vascular tree is topologically connected. Studies of the abdominal aorta are presented.

MRI signal void due to in-plane motion is all-or-none.

The process of MRI signal attenuation due to in-plane intravoxel velocity inhomogeneity is described. Given rigid rotation or linear shear, velocity phase-sensitivity will induce a phase distribution that varies linearly with position, which is exactly equivalent to the effect of a spatial phase encoding gradient pulse. It follows that the effect of such motion on the raw MRI signal is to displace it a fixed distance in kappa-space. Attenuation becomes marked when the center of the spin-echo reaches an edge of kappa-space, which happens when intravoxel phase shifts reach pi radian/voxel. Because spin echoes are typically peaked sharply at center, this attenuation usually is abrupt. Analytic and numerical simulations of linear and nonlinear velocity fields confirm abrupt MRI attenuation where phase dispersion exceeds pi radian/voxel. Examples of this phenomenon include the abrupt loss of blood signal adjacent the vessel wall in laminar flow, abrupt loss of subendocardial signal in early diastole, and sudden disappearance due to rotation of a kidney during a measurement of diffusion.

Dynamic range compression in MRI by means of a nonlinear gradient pulse.

In current magnetic resonance imaging (MRI), valuable information must often be discarded because the NMR signal has greater dynamic range than the analog-to-digital converter (ADC) hardware. Typically, a small set of high-intensity data points near the center of the spin echo is responsible for most of the MRI data dynamic range. We predict that it is possible to reduce the dynamic range of the MRI spin echo by incorporating an identical nonlinear gradient pulse into each repetition of the imaging pulse sequence, prior to data sampling. This pulse converts the phase distribution of the subject, ordinarily a linear function of image coordinates, into a nonlinear function. A nonlinear phase distribution can have a negligible impact on image magnitude and yet a profound impact on spin-echo magnitude. Given a nonlinear phase distribution, there will no longer be a single data point at which all of the protons have an identical phase (the echo center). Instead, the protons become phase coherent on a piecemeal basis, the echo peak is smoothed out, and its maximum amplitude and dynamic range are greatly diminished. Using gradient pulses of quadratic spatial variation, we estimate that maximum echo amplitude and dynamic range can be reduced in most cases by an order of magnitude.

Motional phase artifacts in Fourier transform MRI.

Most magnetic resonance imaging (MRI) techniques are subject to a "motional blurring" arising from the acquisition of data in the presence of a frequency-encoding gradient. The Fourier transform of the signal from a spin moving along a magnetic field gradient obeys an equation analogous to the free space Schrödinger equation. Computer simulations of the Bloch equations illustrate the implications of this motional blurring in MRI.

Fiber crossing in human brain depicted with diffusion tensor MR imaging.

Human white matter fiber crossings were investigated with use of the full eigenstructure of the magnetic resonance diffusion tensor. Intravoxel fiber dispersions were characterized by the plane spanned by the major and medium eigenvectors and depicted with three-dimensional graphics. This method improves the analysis of fiber orientations, beyond the principal fiber directions, to a broader range of complex fiber architectures.

Selective saturation NMR imaging.

The application of selective saturation (or solvent suppression) techniques in nuclear magnetic resonance (NMR) imaging offers the opportunity to significantly expand the range of NMR studies. Data acquired at 1.44 T are presented using a two-dimensional spin-echo sequence preceded by a selective (saturating) radiofrequency pulse. Individual water or lipid proton resonances were eliminated (greater than 90% reduction in signal intensity) resulting in images of H2O or -CH2- distribution with resolution and imaging time equivalent to conventional proton images. Data are also presented demonstrating the feasibility of using selective saturation to image proton metabolites at low concentrations with a three-dimensional chemical shift imaging approach. Lactate was investigated because of its importance in the pathophysiology of ischemic insult. Phantom studies without solvent suppression failed to detect lactate at 80 mM; however, with solvent suppression, lactate at 40 mM was imaged in a reasonable time (approximately 50 min). With the favorable NMR characteristics of the methyl protons of lactate and with improvements in imaging systems, this technique may play an important role in the noninvasive evaluation of tissue ischemia using 1H NMR.

Time of flight quantification of coronary flow with echo-planar MRI.

Detection and quantification of flow of the left anterior descending (LAD) coronary artery in healthy volunteers are demonstrated using echo-planar imaging (EPI). A time-of-flight (TOF) model was used to derive coronary flow velocities from wash-in curves, free of cardiac wall motion contamination. Short-axis cardiac studies were performed using a gated, gradient echo EPI technique to limit the effect of cardiac wall motion on coronary vessel imaging. A series of 10 to 20 single or multislice images were acquired within a single breath-hold. Real-time cine series showed the LAD coronary artery with a detectability of 91% (n = 23) and revealed beat-to-beat variability in vessel position of a magnitude equal to or greater than its diameter. Flow velocity was measured in the proximal portion of the artery at rest and during exercise. The data demonstrated the known phasic pattern of LAD flow: Vsystole < or = 5 cm/s and peak Vdiastole = 14 +/- 3 cm/s (n = 11, V = mean laminar flow velocity). During isometric exercise, a LAD flow velocity increase (52 +/- 24%) was detected in eight of nine subjects. The capacity of the EPI TOF method to detect flow velocity changes should prove clinically useful for future assessment of coronary flow reserve.

Anisotropy of water diffusion in the myocardium of the rat.

Pulsed field gradient nuclear magnetic resonance methods combined with nuclear magnetic resonance imaging were used to determine the water diffusion anisotropy in perfused rat hearts at 37 degrees C. It was found that the observed diffusion coefficient D(app) (apparent diffusion coefficient) depends on the orientation of the applied gradient g. When g is parallel to the epicardial surface, the observed diffusivity is D(app) parallel = 1.8 +/- 0.4 x 10(-9) m2.s-1, whereas when g is perpendicular to it, diffusivity is D(app) perpendicular = 2.5 +/- 0.5 x 10(-9) m2.s-1. To better characterize this directional dependence, images of the second-order diffusion tensor D of the myocardium were obtained. These data demonstrate several essential features of cardiac myoarchitecture, including the helicity of fiber orientation with respect to the ventricular axis and the variation of fiber pitch angle with transmural depth. Diffusion anisotropy may be quantified in a coordinate-independent manner by the eigenvalues of the diffusion tensor. In the myocardial midwall, these eigenvalues were E1 = 3.29 +/- 0.57, E2 = 2.01 +/- 0.42, and E3 = 0.77 +/- 0.58 x 10(-9) m2.s-1 (mean +/- SD). These data suggest that myocardial water diffusion is essentially unrestricted parallel to the myofibers. They further show that failure to measure the complete diffusion tensor may lead to substantial underestimates of diffusion anisotropy in the myocardium.

Cardiac diffusion tensor MRI in vivo without strain correction.

Cardiac diffusion MRI with diffusion encoding that spans a cardiac cycle is complicated by myocardial strains. This paper presents a method to obtain accurate diffusion data without strain correction. Owing to the synchrony of normal cardiac motion, there are time points in the cardiac cycle, "sweet spots," when the cardiac configuration approximates its temporal mean. If the diffusion is encoded then, the net effect of strain on the observed diffusion approximates zero. To test this, MRI diffusion and strain-rate movies are performed on cyclically deformed gel phantoms and in five normal subjects. In phantoms, the sweet spots predicted from the strain time curves agree with the times when the observed diffusion equals the true diffusion. In humans, the strain prediction of the sweet spots and the locations determined by the diffusion trace show a high correlation, r = 0.99. In all subjects, diffusion MRI presents a fiber orientation pattern comparable to that obtained from a stationary specimen. Magn Reson Med 42:393-403, 1999.

Biomechanical basis for lingual muscular deformation during swallowing.

Our goal was to quantify intramural mechanics in the tongue through an assessment of local strain during the physiological phases of swallowing. Subjects were imaged with an ultrafast gradient echo magnetic resonance imaging (MRI) pulse sequence after the application of supersaturated magnetized bands in the x and y directions. Local strain was defined through deformation of discrete triangular elements defined by these bands and was depicted graphically either as color-coded two-dimensional strain maps or as three-dimensional octahedra whose axes correspond to the principal strains for each element. During early accommodation, the anterior tongue showed positive strain (expansive) in the anterior-posterior direction (x), whereas the middle tongue showed negative strain (contractile) in the superior-inferior direction (y). During late accommodation, the anterior tongue displayed increased positive x-direction and y-direction strain, whereas the posterior tongue displayed increased negative y-direction strain. These findings were consistent with contraction of the anterior-located intrinsic muscles and the posterior-located genioglossus and hyoglossus muscles. During propulsion, posterior displacement of the tongue was principally associated with positive strain directed in the x and y directions. These findings were consistent with posterior passive stretch in the midline due to contraction of the laterally inserted styloglossus muscle, as well as contraction of the posterior located transversus muscle. We conclude that MRI of lingual deformation during swallowing resolves the synergistic contractions of the intrinsic and extrinsic muscle groups.

Projective MRI angiography and quantitative flow-volume densitometry.

Projective MR images of vascular anatomy and flow are performed at 0.14 T by using phase contrast to suppress the signal contribution of the stationary background. The source of the contrast is the distinctive phase evolution of moving protons under the influence of the read-out gradient of a conventional two-dimensional Fourier transform (2D FT) spin-echo pulse sequence. By using short echo times, small phase shifts may be obtained. When phase shifts are less than about 45 degrees, the phase contrast assumes a simple and useful form. The flow image intensity at any pixel becomes proportional to the net flux or flow volume of protons which cross the corresponding voxel. This proportionality is demonstrated in images of flow phantoms as is the reproducibility of measured flow volume under a variety of transformations of imaging conditions and of the subject. Projective images gated in vivo produce angiographic views of arteries and veins, in systole and diastole, in the neck of a dog and in the lower extremities of a human subject.