Analysis and use of FMRI response delays.

In this study, we implemented a new method for measuring the temporal delay of functional magnetic resonance imaging (fMRI) responses and then estimated the statistical distribution of response delays evoked by visual stimuli (checkered annuli) within and across voxels in human visual cortex. We assessed delay variability among different cortical sites and between parenchyma and blood vessels. Overall, 81% of all responsive voxels showed activation in phase with the stimulus while the remaining voxels showed antiphase, suppressive responses. Mean delays for activated and suppressed voxels were not significantly different (P < 0.001). Cortical flat maps showed that the pattern of activated and suppressed voxels was dynamically induced and depended on stimulus size. Mean delays for blood vessels were 0.7-2.4 sec longer than for parenchyma (P < 0.01). However, both parenchyma and blood vessels produced responses with long delays. We developed a model to identify and quantify different components contributing to variability in the empirical delay measurements. Within-voxel changes in delay over time were fully accounted for by the effects of empirically measured fMRI noise with virtually no measurable variability associated with the stimulus-induced response itself. Across voxels, as much as 47% of the delay variance was also the result of fMRI noise, with the remaining variance reflecting fixed differences in response delay among brain sites. In all cases, the contribution of fMRI noise to the delay variance depended on the noise power at the stimulus frequency. White noise models significantly underestimated the fMRI noise effects.

Direct reconstruction of non-Cartesian k-space data using a nonuniform fast Fourier transform.

An algorithm of Dutt and Rokhlin (SIAM J Sci Comput 1993;14:1368-1383) for the computation of a fast Fourier transform (FFT) of nonuniformly-spaced data samples has been extended to two dimensions for application to MRI image reconstruction. The 2D nonuniform or generalized FFT (GFFT) was applied to the reconstruction of simulated MRI data collected on radially oriented sinusoidal excursions in k-space (ROSE) and spiral k-space trajectories. The GFFT was compared to conventional Kaiser-Bessel kernel convolution regridding reconstruction in terms of image reconstruction quality and speed of computation. Images reconstructed with the GFFT were similar in quality to the Kaiser-Bessel kernel reconstructions for 256(2) pixel image reconstructions, and were more accurate for smaller 64(2) pixel image reconstructions. Close inspection of the GFFT reveals it to be equivalent to a convolution regridding method with a Gaussian kernel. The Gaussian kernel had been dismissed in earlier literature as nonoptimal compared to the Kaiser-Bessel kernel, but a theorem for the GFFT, bounding the approximation error, and the results of the numerical experiments presented here show that this dismissal was based on a nonoptimal selection of Gaussian function.

Specialized neural systems underlying representations of sequential movements.

The ease by which movements are combined into skilled actions depends on many factors, including the complexity of movement sequences. Complexity can be defined by the surface structure of a sequence, including motoric properties such as the types of effectors, and by the abstract or sequence-specific structure, which is apparent in the relations amongst movements, such as repetitions. It is not known whether different neural systems support the cognitive and the sensorimotor processes underlying different structural properties of sequential actions. We investigated this question using whole-brain functional magnetic resonance imaging (fMRI) in healthy adults as they performed sequences of five key presses involving up to three fingers. The structure of sequences was defined by two factors that independently lengthen the time to plan sequences before movement: the number of different fingers (1-3; surface structure) and the number of finger transitions (0-4; sequence-specific structure). The results showed that systems involved in visual processing (extrastriate cortex) and the preparation of sensory aspects of movement (rostral inferior parietal and ventral premotor cortex (PMv)) correlated with both properties of sequence structure. The number of different fingers positively correlated with activation intensity in the cerebellum and superior parietal cortex (anterior), systems associated with sensorimotor, and kinematic representations of movement, respectively. The number of finger transitions correlated with activation in systems previously associated with sequence-specific processing, including the inferior parietal and the dorsal premotor cortex (PMd), and in interconnecting superior temporal-middle frontal gyrus networks. Different patterns of activation in the left and right inferior parietal cortex were associated with different sequences, consistent with the speculation that sequences are encoded using different mnemonics, depending on the sequence-specific structure. In contrast, PMd activation correlated positively with increases in the number of transitions, consistent with the role of this area in the retrieval or preparation of abstract action plans. These findings suggest that the surface and the sequence-specific structure of sequential movements can be distinguished by distinct distributed systems that support their underlying mental operations.

Event-related fMRI contrast when using constant interstimulus interval: theory and experiment.

Event-related functional magnetic resonance imaging (ER-fMRI) involves the mapping of averaged hemodynamic changes resulting from repeated, brief (<3 sec) brain activation episodes. In this paper, two issues regarding constant-interstimulus interval ER-fMRI were addressed. First, the optimal interstimulus interval (ISI), given a stimulus duration (SD), was determined. Second, the statistical power of ER-fMRI relative to that of a blocked-design paradigm was determined. Experimentally, it was found that with a 2-sec SD, the optimal ISI is 12 to 14 sec. Theoretically, the optimal repetition interval (T(opt) = ISI + SD) is 12 to 14 sec for stimuli of 2 sec or less. For longer stimuli, T(opt) is 8 + 2 x SD. At the optimal ISI for SD = 2 sec, the experimentally determined functional contrast of ER-fMRI was only -35% lower than that of blocked-design fMRI. Simulations that assumed a linear system demonstrated an event-related functional contrast that was -65% lower than that of the blocked design. These differences between simulated and experimental contrast suggest that the ER-fMRI amplitude is greater than that predicted by a linear shift-invariant system.

k-Space partition diagrams: a graphical tool for analysis of MRI pulse sequences.

A new type of graphical tool for explaining and analyzing magnetic resonance imaging pulse sequences is developed and illustrated. This tool combines the partition diagram, which shows the evolution of multiple echoes with the application of multiple RF pulses, and k-space graphs, which show the evolution of the transverse magnetization as gradients are applied. The strength of the new tool lies in its ability to depict clearly the progression of complex imaging pulse sequences. Several complicated excitation sequences are used to illustrate this method.

Real-time 3D image registration for functional MRI.

Subject head movements are one of the main practical difficulties with brain functional MRI. A fast, accurate method for rotating and shifting a three-dimensional (3D) image using a shear factorization of the rotation matrix is described. Combined with gradient descent (repeated linearization) on a least squares objective function, 3D image realignment for small movements can be computed as rapidly as whole brain images can be acquired on current scanners. Magn Reson Med 42:1014-1018, 1999.

Mapping of semantic, phonological, and orthographic verbal working memory in normal adults with functional magnetic resonance imaging.

Twelve neurologically normal participants (4 men and 8 women) performed semantic, phonological, and orthographic working memory tasks and a control task during functional magnetic resonance imaging. Divergent regions of the posterior left hemisphere used for decoding and storage of information emerged in each working memory versus control task comparison. These regions were consistent with previous literature on processing mechanisms for semantic, phonological, and orthographic information. Further, working memory versus control task differences extended into the left frontal lobe, including premotor cortex, and even into subcortical structures. Findings were consistent with R. C. Martin and C. Romani's (1994) contention that different forms of verbal working memory exist and further suggest that a reconceptualization of premotor cortex functions is needed.

Rotation of NMR images using the 2D chirp-z transform.

A quick and accurate way to rotate and shift nuclear magnetic resonance (NMR) images using the two-dimensional chirp-z transform is presented. When the desired image grid is rotated and shifted from the original grid due to patient motion, the chirp-z transform can reconstruct NMR images directly onto the ultimate grid instead of reconstructing onto the original grid and then applying interpolation to get the final real-space image in the conventional way. The rotation angle and shift distances are embedded in the parameters of the chirp-z transform. The chirp-z transform implements discrete sinc interpolation to get values at grid points that are not exactly on the original grid when applying the inverse Fourier transform. Therefore, the chirp-z transform is more accurate than methods such as linear or bicubic interpolation and is more efficient than direct implementation of sinc interpolation because the sinc interpolation is implemented at the same time as reconstruction from k-space.

Language processing is strongly left lateralized in both sexes. Evidence from functional MRI.

Functional MRI (fMRI) was used to examine gender effects on brain activation during a language comprehension task. A large number of subjects (50 women and 50 men) was studied to maximize the statistical power to detect subtle differences between the sexes. To estimate the specificity of findings related to sex differences, parallel analyses were performed on two groups of randomly assigned subjects. Men and women showed very similar, strongly left lateralized activation patterns. Voxel-wise tests for group differences in overall activation patterns demonstrated no significant differences between women and men. In further analyses, group differences were examined by region of interest and by hemisphere. No differences were found between the sexes in lateralization of activity in any region of interest or in intrahemispheric cortical activation patterns. These data argue against substantive differences between men and women in the large-scale neural organization of language processes.

Event-related fMRI of tasks involving brief motion.

The assessment of brain function by blood oxygenation level dependent (BOLD) functional magnetic resonance imaging (fMRI) for tasks involving motion near the field of view is compromised by artifacts arising from the motion. The aim of this study is to demonstrate that these artifacts can be reduced by acquiring the average response from a brief stimulus (a "single-trial," or "event-related," paradigm) as opposed to alternating blocks of repeated task with rest (a "block-trial" paradigm). The basis of this technique is that the NMR signal changes from neuronal activation are delayed relative to the motion due to a slow hemodynamic response. By acquiring the average response from a brief stimulus, motion-induced signal changes occur prior to neuronal activation-induced signal changes, and the two can thus be distinguished. This technique is applied to the tasks of speaking out loud, swallowing, jaw clenching, and tongue movement. Functional activation maps derived from the single-trial paradigm contain significantly less artifact than functional activation maps derived from a more traditional block-trial paradigm.

Conceptual processing during the conscious resting state. A functional MRI study.

Localized, task-induced decreases in cerebral blood flow are a frequent finding in functional brain imaging research but remain poorly understood. One account of these phenomena postulates processes ongoing during conscious, resting states that are interrupted or inhibited by task performance. Psychological evidence suggests that conscious humans are engaged almost continuously in adaptive processes involving semantic knowledge retrieval, representation in awareness, and directed manipulation of represented knowledge for organization, problem-solving, and planning. If interruption of such 'conceptual' processes accounts for task-induced deactivation, tasks that also engage these conceptual processes should not cause deactivation. Furthermore, comparisons between conceptual and nonconceptual tasks should show activation during conceptual tasks of the same brain areas that are 'deactivated' relative to rest. To test this model, functional magnetic resonance imaging data were acquired during a resting state, a perceptual task, and a semantic retrieval task. A network of left-hemisphere polymodal cortical regions showed higher signal values during the resting state than during the perceptual task but equal values during the resting and semantic conditions. This result is consistent with the proposal that perceptual tasks interrupt processes ongoing during rest that involve many of the same brain areas engaged during semantic retrieval. As further evidence for this model, the same network of brain areas was activated in two direct comparisons between semantic and perceptual processing tasks. This same 'conceptual processing' network was also identified in several previous studies that contrasted semantic and perceptual tasks or resting and active states. The model proposed here offers a unified account of these findings and may help to explain several unanticipated results from prior studies of semantic processing.

Steady-state currents in sharp stochastic ratchets.

We develop and analyze a model for particle transport in a stochastic ratchet with a periodic piecewise linear potential, with diffusion coefficient D, where the force is discontinuous in position and fluctuating in time via additive telegraph noise with correlation time tau. We find asymptotic formulas for the steady-state particle current J for large and small D and tau. For example, for small tau, the sharp corners in the potential lead to J=O(tau(2) exp[-(Dtau)(-1/2)])+O(tau(5/2)), in contrast to O(tau(3)) when the potential is smooth. We show that diffusion can increase or decrease J, and derive an approximate equation for the value of D that maximizes J.

Two- and three-dimensional image rotation using the FFT.

Techniques for rotating two- and three-dimensional (2-D and 3-D) images using fast Fourier transforms (FFT's) are presented. The methods are applications of the multidimensional chirp algorithm. In the 2-D case, one chirp transformation is sufficient, requiring four 2-D FFT's, In the 3-D case, two successive chirp transformations are required, needing six 3-D FFT's.

Identification and characterization of cerebral cortical response to esophageal mucosal acid exposure and distention.

BACKGROUND & AIMS: Esophageal acid exposure is a common occurrence in healthy individuals and patients with esophagitis. Clinically, perception of this exposure ranges from no perception to severe heartburn and chest pain. Cerebral cortical response to esophageal mucosal contact to acid has not been systematically studied. The aim of this study was to elucidate cerebral cortical response to esophageal acid exposure in normal individuals by functional magnetic resonance imaging (FMRI). METHODS: We studied 10 normal healthy volunteers. Cortical FMRI response to 10 minutes of intraesophageal perfusion of 0.1N HCl (1 mL/min) was determined, and the results were compared with those of saline infusion and balloon distention. RESULTS: Acid perfusion did not induce heartburn or chest pain but increased FMRI signal intensity by 6.7% +/- 2.0% over the preperfusion values. No increase was detected for saline infusion. FMRI signal intensity to balloon distention was similar to that of acid perfusion. Activation latency, activation to peak, and the deactivation periods for response to acid perfusion were significantly longer than those of balloon distention (P < 0.05). CONCLUSIONS: Contact of esophageal mucosa with acid, before inducing heartburn, evokes a cerebral cortical response detectable by FMRI. Temporal characteristics of this response are significantly different from those induced by esophageal balloon distention.

Simultaneous gradient-echo/spin-echo EPI of graded ischemia in human skeletal muscle.

The goal of this study was to evaluate the usefulness of blood oxygenation level-dependent (BOLD) methodologies to provide temporal and spatial information about skeletal muscle perfusion. A simultaneous gradient echo (GE) and spin-echo (SE) imaging sequence (GE/SE) with alternating TE was used to acquire images of leg skeletal muscle throughout a stepped reactive hyperemia paradigm. The change in both the GE and SE relaxation rates (deltaR2*, deltaR2) measured during ischemia and reactive hyperemia scaled with the duration of cuff inflation (the ischemic period) plateaued for cuff inflations lasting longer than 120 seconds and were greater in soleus muscle than in gastrocnemius. The ratio deltaR2*/deltaR2 was found to be less during the reactive hyperemia period relative to ischemia. Considering that a greater proportion of capillary vessels are perfused during reactive hyperemia than during ischemia, this finding suggests that magnetic susceptibility methodologies, with their dependence on compartment size, may provide a measure of the relative distribution of small and large vessels in skeletal muscle.

Magnetic field changes in the human brain due to swallowing or speaking.

Variations in the magnetic field in the human brain caused by the processes of swallowing or speaking are measured. In both processes, motion of the pharyngeal muscles, especially the tongue and jaw, alter the susceptibility-induced magnetic field distribution at the brain slice being imaged. This leads to image warping, compromising the analysis of a time series of images, such as in functional magnetic resonance imaging (fMRI). These dynamic changes are assessed by acquiring a time series of images using a gradient-echo asymmetric-spin-echo sequence (GREASE), a technique in which two images are acquired for each excitation--one during the gradient echo, and one during the latter part of the spin echo. The NMR phase difference between the two images is a measure of the magnetic field distribution. A series of brain images, acquired with this sequence while the subject either swallows or speaks, indicated negative magnetic field changes of up to 0.087 ppm in the inferior region of the brain for both speaking and swallowing, and in some speech, additional positive changes of up to 0.056 ppm in the frontal region of the brain were indicated.

Functional MR activation correlated with intraoperative cortical mapping.

PURPOSE: To evaluate the spatial specificity of functional MR imaging by comparing it with intraoperative electrocortical mapping. METHODS: Functional MR imaging was performed in 28 patients before awake craniotomy and intraoperative electrocortical mapping. Activation was mapped for finger movement, lip movement, tongue movement, word generation, and counting paradigms. During surgery, finger movement, lip movement, tongue movement, counting, and/or speaking were mapped. The functional images and the photographic recordings of the brain functions mapped during surgery were converted to bit maps and coregistered by a computer program. The distance between the intraoperatively mapped function site and the MR activation site for a comparable function was measured. RESULTS: Forty-six functions were recorded on MR images and intraoperative maps. In 100% of correlations, the intraoperative site and the MR activation site were within 20 mm; in 87% of correlations they were within 10 mm. For each paradigm, 67% or more of the intraoperative stimulation maps correlated within 10 mm of the MR activation site. CONCLUSIONS: For the tasks used in this study, the activation site on functional MR images correlated well with the site at which intraoperative stimulation identified function.

The effect of magnetization transfer on functional MRI signals.

A magnetization transfer (MT)-prepared echo-planar imaging (EPI) pulse sequence was developed to study motor cortex activation, using a finger tapping paradigm. MT weighting resulted in a reduction of both the activated area and, in the majority of activated pixels, the functional MRI signal, regardless of the correlation coefficient threshold used in generating the activation map. The magnetization transfer ratio (MTR) was higher during task activation than during rest. Because the MT effect is strongly tissue-dependent, these results support the hypothesis that incorporation of MT into functional MRI will help to understand the origin of the functional MRI signal.

Graded effects of spatial and featural attention on human area MT and associated motion processing areas.

Functional magnetic resonance imaging was used to quantify the effects of changes in spatial and featural attention on brain activity in the middle temporal visual area and associated motion processing regions (hMT+) of normal human subjects. When subjects performed a discrimination task that directed their spatial attention to a peripherally presented annulus and their featural attention to the speed of points in the annulus, activity in hMT+ was maximal. If subjects were instead asked to discriminate the color of points in the annulus, the magnitude and volume of activation in hMT+ fell to 64 and 35%, respectively, of the previously observed maximum response. In another experiment, subjects were asked to direct their spatial attention away from the annulus toward the fixation point to detect a subtle change in luminance. The response magnitude and volume dropped to 40 and 9% of maximum. These experiments demonstrate that both spatial and featural attention modulate hMT+ and that their effects can work in concert to modulate cortical activity. The high degree of modulation by attention suggests that an understanding of the stimulus-driven properties of visual cortex needs to be complemented with an investigation of the effects of task-related factors on visual processing.

Distributed neural systems underlying the timing of movements.

Timing is essential to the execution of skilled movements, yet our knowledge of the neural systems underlying timekeeping operations is limited. Using whole-brain functional magnetic resonance imaging, subjects were imaged while tapping with their right index finger in synchrony with tones that were separated by constant intervals [Synchronization (S)], followed by tapping without the benefit of an auditory cue [Continuation (C)]. Two control conditions followed in which subjects listened to tones and then made pitch discriminations (D). Both the S and the C conditions produced equivalent activation within the left sensorimotor cortex, the right cerebellum (dorsal dentate nucleus), and the right superior temporal gyrus (STG). Only the C condition produced activation of a medial premotor system, including the caudal supplementary motor area (SMA), the left putamen, and the left ventrolateral thalamus. The C condition also activated a region within the right inferior frontal gyrus (IFG), which is functionally interconnected with auditory cortex. Both control conditions produced bilateral activation of the STG, and the D condition also activated the rostral SMA. These results suggest that the internal generation of precisely timed movements is dependent on three interrelated neural systems, one that is involved in explicit timing (putamen, ventrolateral thalamus, SMA), one that mediates auditory sensory memory (IFG, STG), and another that is involved in sensorimotor processing (dorsal dentate nucleus, sensorimotor cortex).