Brain Mapping seminar abstracts 2006-07

Irene Tracey, Ph.D.
Professor of Pain Research and Director, FMRIB Centre, Oxford University
Wednesday, 08/30/06, noon, MGH bldg 149

Imaging Pain: from Agony to Analgesia
Pain is a conscious experience, an interpretation of the nociceptive input influenced by memories, emotional, pathological and cognitive factors. The behavioural response by a subject to a painful event is modified according to what is appropriate or possible in any particular situation. Pain is, therefore, a subjective experience as illustrated by the definition given by The International Association for the Study of Pain (IASP): An unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage. Until recently it has been difficult to obtain reliable objective information from healthy volunteers and patients regarding their subjective experience of pain. Since the introduction of functional neuroimaging methods, such as functional magnetic resonance imaging (FMRI), positron emission tomography (PET) and electroencephalography (EEG), we and others have been able to show robust and reproducible activation in response to nociceptive (i.e. painful) stimuli within the human brain and spinal cord. This activation can be related to what the subject describes and issues such as how anxiety, attention, distraction and anticipation alter pain perception better understood. In short, we can unravel the workings of pain perception at a neural level.

Relating specific neurophysiological markers to the perceptual changes induced by pharmacological agents and identifying their site of action within the human nervous system has been a major goal for drug discovery. Recently, pharmacological functional magnetic resonance imaging (phMRI) methods have been developed and applied to the field of pain research. Studies in our laboratory have identified novel pharmacodynamic information spatially across the human brain using remifentanil (short acting opoid analegesic drug) in combination with a painful stimulus. FMRI has also been used in combination with 'gold-standard' drugs that are currently used in the treatment of chronic pain. We have applied phMRI to gain a better understanding of action of these compounds in both human models of clinical pain as well as neuropathic and inflammatory pain patients. More recently, simultaneous recording of EEG and FMRI measurements have been used to directly examine pain in human subjects and in addition will be further applied to patients and the investigation of drug effects targeted for pain alleviation.

My talk will address all these issues and hopefully illustrate how innovative methods and ways of tackling this global problem are needed but also in development.

Randy Buckner, Ph.D.
Professor in Psychology at Harvard University
Wednesday, 09/13/06, noon, MGH bldg 149

Is There A Common Brain Network for Remembering the Past, Envisioning the Future, and Conceiving the Viewpoints of Others?
We think about the future to explore the consequences of upcoming events before they happen. Converging data from functional MRI studies suggest envisioning the future is one member of a larger class of related forms of self-projection that includes remembering, conceiving the viewpoint of others (theory of mind), and navigation. All emerge in development at about the same age and share a similar core functional anatomy. In particular, frontal and medial temporal systems traditionally associated with planning and episodic memory are components of the shared network. Recent imaging data further suggest that activity increases during passive states and lapses in attention (the default mode) likely associate with the same core anatomy raising the possibility that we spend much of our time considering the future and conceiving alternative perspectives. This bias to be in a processing mode that shifts our perception from the present to alternative, simulated perspectives has far-reaching cognitive and physiological consequences including (possibly) setting the stage for the development of Alzheimer's disease.

Daphne Holt, M.D., Ph.D.
Instructor of Psychiatry at Harvard Medical School
Wednesday, 09/20/06, noon, MGH bldg 149

Abnormal Salience in Psychosis: fMRI Evidence for Medial Temporal Lobe and Cingulate Cortex Dysfunction
Patients with schizophrenia tend to misinterpret ambiguous signals in their environment during psychotic states; a delusion frequently incorporates trivial coincidences or benign events which a healthy person can easily ignore. Because of these clinical observations, it has been hypothesized that psychosis is a state of "aberrant salience;" in a heightened state of arousal, psychotic patients misassign emotional salience to incidental external stimuli and internal representations, leading to delusions and hallucinations.

In support of this theory, we have shown that delusional patients with schizophrenia demonstrate a specific response bias to misclassify affectively neutral words as negatively-valenced. In this talk, I will describe fMRI studies of emotional processing in schizophrenia which aim to identify the neural basis of abnormal salience evaluations in psychosis. Our hypothesis is that patients with schizophrenia exhibit abnormal activity in limbic-evaluative systems, specifically medial temporal-cingulate circuitry, during the evaluation of ambiguous social information. Abnormal interactions between emotional and cognitive operations in psychotic patients may account for the emotional valence and the faulty reasoning that is characteristic of delusional thinking.

Avi Snyder, M.D., Ph.D.
Res. Assoc. Professor of Radiology, Washington University School of Medicine
Wednesday, 09/27/06, noon, MGH bldg 149

The 4dfp Suite of Image Processing Programs
The 4dfp suite of image processing programs is designed to for the analysis of structural images of any modality with an emphasis on functional MRI. 4dfp functionality includes minimization of fMRI artifacts (e.g., intensity inhomogeneity) and compensation for head movement within and across scans. The image registration components handle alignment of common modal as well as multimodal contrast mechanisms. Other modules perform spatial and temporal filtering. Applications include conventional fMRI as well as correlation anaysis, which is central to "functional connectivity" studies (see attached pdf). The 4dfp suite is maximally modular. All procedures are invoked using UNIX command lines (with options). Provision is made to automatically maintain a history of all processing steps so that final results (regardless of complexity) can be re-generated if the antecedent data and the necessary history file (*.4dfp.img.rec) are present. The rec file concept may be usefully extended to any data modality (e.g., MEG) that is sequentially processed ultimately generating multi-subject statistical results. Selected algorithms will be discussed according to interest.

Steve Jones, Ph.D.
Professor of Anesthesiology, Drexel University College of Medicine
Wednesday, 10/04/06, noon, MGH bldg 149

Sodium and Rubidium Magnetic Resonance Imaging Investigations of Experimental Ischemic Stroke
Our studies of the brain tissue electrolytes, Na+ and K+, in early experimental stroke have shown: 1. The sodium increase after stroke can be used to estimate the time from stroke onset. 2. Potassium changes occur that suggest a more definitive timing event that could potentially be used to avoid hemorrhagic transformation. Initial results using flame photometry and punch sampling for analysis of sodium (Wang et al, Stroke, 2000, 31: 1386), potassium, and rubidium (as a congener of potassium) have been (Jones et al, Stroke, 2006, 37: 883) or will be extended to MRI methods with the advantage of continuous measurements in individual subjects. This work is based on many years of studies using brain sampling, including ours, but is made relevant in today's world by the availability of MRI techniques to monitor dynamic changes of sodium and rubidium in a single animal.

Marty Sereno, Ph.D.
Professor, Departments of Cognitive Science and Neurosciences, UCSD
Wednesday, 10/11/06, 4pm, Harvard WJH 001

Mapping Higher Level Cortical Areas in Humans
There has traditionally been a dichotomy between research groups studying mapping (especially in the visual system) and groups studying higher level cognitive function. To some extent, this has been regarded as a division of labor with the former using surface-based methods and the latter primarily using 3D (Talairach) methods. I will summarize my recent work showing that maps -- retinotopic and otherwise -- extend well into areas activated by cognitive experimental paradigms (e.g., the precuneus). To understand how these different kinds of maps contribute to cognition, it is crucial that 'low level' mapping experiments and 'high level' cognitive experiments are done in the same groups of subjects.

Ole Jensen, Ph.D.
F.C. Donders Centre for Cognitive Neuroimaging, Radboud Universiteit
Wednesday, 10/25/06, 4pm, MIT 46-3189

The Role of Posterior Alpha Activity in Neuronal Processing and the Generation of Event-Related Fields and Potentials
Human recordings of ongoing brain activity by electroencephalography (EEG) and magnetoencephalography (MEG) are dominated by posterior oscillations around 10 Hz termed the alpha rhythm (Berger, 1929). While these oscillations are produced by large ensembles of neurons oscillating in synchrony, their role for neuronal processing and the generation of event-related fields/potentials remains unclear. It has been proposed that alpha activity reflects `cortical idling'; however, this notion has been contradicted by findings showing an increase in alpha activity with task difficulty during retention of working memory. This suggests that alpha activity reflects active inhibition of the visual stream: a top-down drive serves to suppress visual inputs in order to allocate resources to regions engaged in working memory processing. This hypothesis directly predicts that visual processing should decrease with increasing alpha power. This was recently demonstrated in an MEG experiment where an inverse relationship was observed between pre-stimulus alpha power and visual detection ability. It is highly conceivable that ongoing oscillatory activity plays an important role in the generation of event-related potentials and fields (ERPs and ERFs). In particular is has been suggested that phase-resetting of ongoing oscillations can account for some components of the ERPs/ERFs; however, recent findings contradict this notion. Nevertheless, this does not mean that alpha activity does not modulate the generation of evoked responses. If one assumes that the amplitude modulation of alpha activity is `DC offset', one can show that event-related synchronization and desynchronization will produce strong modulations in the slower components of event-related fields/potentials. Consistent with new data, this implies that ongoing alpha activity in general will modulate various components of visually evoked fields. The principle can be generalized to other frequency bands as well. In short: oscillatory alpha activity does play a strong role both in the generation of evoked activity and neuronal processing. Alpha activity mainly serves to inhibit the dorsal visual stream by a top-town mechanism exercising gain control.

Polina Golland, Ph.D.
Assistant Professor at MIT CSAIL
Wednesday, 11/01/06, noon, MGH bldg 149

Functional Region Hierarchy: Representation and Modeling of Spatial Activation Patterns in fMRI
In this talk, I will present a novel approach to computational modeling of spatial activation patterns observed through fMRI. Functional connectivity analysis is widely used in fMRI studies for detection and analysis of large networks that co-activate with a user-selected `seed' region of interest. In contrast, our method is based on clustering; it simultaneously identifies interesting seed time courses and associates voxels with the respective networks. This generalization eliminates the sensitivity to the threshold used to classify voxels as members of a network and enables discovery of co-activated networks without user selection of seed regions.

Based on the empirical observation that the detected patterns of co-activation are inherently hierarchical, we propose a new representation for spatial patterns of functional organization. Just like the anatomical hierarchies represent the structure of the brain as a tree of increasingly simple systems, we believe that the functional description of the brain should also be of a hierarchical nature. We introduce Functional Region Hierarchy, a top-down representation that encapsulates the notion that functionally defined regions should be viewed at different resolutions, as dictated by the observed activation pattern. We construct the functional region hierarchy through an iterative decomposition that utilizes clustering for network subdivision at each step.

The experimental results demonstrate that the functional region hierarchy provides a robust and anatomically meaningful model for spatial patterns of co-activation in fMRI. The hierarchical representation leads to insights into the structure of the functional networks that are not immediately apparent from flat representations that segment the brain into a large number of small regions. In addition, subject-specific region hierarchies tend to share common tree structure, further confirming the validity of this representation for modeling group-wise patterns of co-activation.

Martha Shenton, Ph.D.
Professor of Psychology and Professor of Radiology, Harvard Medical School
Wednesday, 11/08/06, noon, MGH bldg 149

Neurobiology of Schizotypal Personality Disorder
 

Kevin M. Spencer, Ph.D.
Assistant Professor of Psychiatry, VA Boston Healthcare System and Harvard Medical School
Wednesday, 11/15/06, noon, MGH bldg 149

EEG and Computational Modeling Studies of Gamma Oscillations in Schizophrenia
The precise synchronization of neuronal activity mediated by oscillations in the gamma band may play a key role in conscious perception by 1) linking features coded by individual neurons into coherent objects, and 2) enhancing the downstream impact of these cell assemblies coding task-relevant, attended objects. Thus, abnormal gamma oscillations in schizophrenia patients (SZ) could be related to hallucinations, in which auditory and visual objects are perceived which are not physically present. This hypothesis is supported by several findings:

1) The early visual-evoked gamma oscillation was reduced in SZ compared to healthy controls (HC) in visual perception tasks. This effect was most prominent in SZ with a history of visual hallucinations. A similar relationship was found for the early auditory-evoked gamma oscillation and auditory hallucination symptoms. 2) A gamma oscillation that was phase-locked to subjectsâ~Y reaction times for perceived objects occurred at a lower frequency in SZ than HC and was correlated with visual hallucination symptom scores. In contrast, visual evoked potential amplitude was reduced in SZ with vs. without a history of visual hallucinations. 3) The onset period of the auditory 40 Hz steady-state response to right ear (left hemisphere) stimulation was negatively correlated with auditory hallucination scores.

In sum, hallucination symptoms were associated with reduced sensory-evoked activity but enhanced perception-related activity, consistent with other evidence for hyperexcitability in the sensory cortex of hallucinators. Cortical hyperexcitability could be due to NMDA receptor hypofunction and impaired function of inhibitory interneurons, both of which have been implicated in schizophrenia and are crucial for the generation of gamma oscillations. This hypothesis is supported by computational models of gamma oscillations that mimic the effects of neural circuit abnormalities found in schizophrenia.

Joanna Fowler, Ph.D.
Senior Chemist, Brookhaven National Lab
Wednesday, 11/29/06, 4pm, MIT 46-3189

Imaging Addiction in the Human Brain
Molecular imaging utilizing positron emission tomography (PET) and labeled compounds continue to provide new knowledge of brain circuits which are altered in disease and by drugs. This knowledge, in turn, generates new opportunities to design and develop better drugs and to monitor treatment. The full utilization of PET in drug discovery and in drug research and development and its translation to humans requires advances in radiotracer chemistry, particularly in the development of rapid synthetic methods for introducing the short-lived isotopes, carbon-11 (t1/2: 20.4 min) and fluorine-18 (t1/2: 110 min) into a wide variety of chemical compounds. In this seminar, we highlight some examples of the use of PET and selective radiotracers to measure the pharmacokinetics and pharmacodynamics of drugs of abuse on the human brain. We will also highlight some new radiotracers, new animal models and imaging instrumentation which will serve as scientific tools to advance our knowledge of the human brain. Supported by the U. S. Department of Energy, Office of Biological and Environmental Research and the National Institutes of Health (NIDA, NIBIB) and the Office of National Drug Control Policy.

Ted Huppert, M.Sc.
PhD Candidate, Harvard University Graduate Programs in Biophysics,
MGH Photon Migration Imaging lab
Wednesday, 12/06/06, noon, MGH bldg 149

Inferring Baseline and Functional Blood Flow, Volume, and Cerebral Metabolism from the Dynamics of Blood Oxygenation Changes
During functional activation, cerebral blood oxygenation varies as a result of both an increase in regional blood flow and a localized change in cerebral oxygen consumption to support glycolysis. Hemodynamic measurements depending on blood oxygenation changes, such as the fMRI-BOLD signal or optical imaging, can be used to infer not only direct information about oxygen extraction, but also indirect information about the blood flow and volume responses based on the effect that these states have on the relative temporal dynamics of evoked changes. In this work, we describe and apply a multi-compartment model of the vascular network, which simultaneously considers both the vascular and oxygen transport biophysics in the brain, it is possible to separate the functional changes in both arteriole dilation and cerebral oxygen metabolism, which fundamentally drive the hemodynamic response. In addition, baseline oxygen metabolism, blood flow, and blood volume are determined by parameter estimation within the model. In this work, we demonstrate that the estimation of these two states and the structural and baseline parameters can be uniquely achieved from non-invasive optical measurements of oxy-, deoxy-, and total-hemoglobin changes. Similarly, we show using pulsed arteriole spin labeling techniques, which provides both a relative blood flow and an uncalibrated BOLD signal, that these states can also be estimated from fMRI methods alone. By comparing these estimates obtained from optical data or fMRI data alone to estimates using the complete multimodality data, which provides the full compliment of hemodynamic variables, we verify that both individual modalities are sufficient to estimate oxygen metabolism and arteriole dilation changes. This finding is supported by both numerical simulations and experimental data acquired simultaneously with optical and fMR imaging.

David J. Freedman, Ph.D.
Postdoctoral Fellow, Department of Neurobiology, Harvard Medical School
Wednesday, 12/13/06, noon, MGH bldg 149

Exploring the Roles of the Frontal, Temporal and Parietal Lobes in Visual Learning and Categorization
Our perception of the environment is not a faithful registration of its physical attributes.  Instead, we carve the world into meaningful groupings, or categories. For example, knowing that a new gadget is a "camera" instantly and effortlessly provides a great deal of information about its relevant parts and functions. The ability to categorize stimuli is a cornerstone of complex behavior. Categories are evident in all sensory modalities and range from relatively simple (e.g., color perception) to the most abstract human concepts. While much is known about how the brain processes simple sensory features (i.e. color, orientation, and direction of motion), less is known about the neuronal processes that encode the category membership, or meaning, of stimuli.

This talk will review a series of experiments aimed at understanding the respective roles of several interconnected brain areas during visual categorization. By recording the activity of individual neurons in monkeys trained to categorize visual stimuli, we found that activity in two brain areas, the lateral prefrontal cortex (PFC) and lateral intraparietal area (LIP), robustly encoded the category membership of visual stimuli. This suggests that both the PFC and LIP may be important stages for the transformation of visual information to more abstract representations of the categorical meaning of visual stimuli.

Anastasia Yendiki
Postdoctoral fellow, Athinoula A. Martinos Center for Biomedical Imaging, Harvard Medical School
Friday, 01/05/07, 1pm, MGH bldg 149

Bayesian approach to tensor estimation for diffusion MRI
Diffusion tensor imaging (DTI) is a growing application of MRI, in which Fourier-domain samples of the imaged volume are used to estimate a tensor map that represents the dominant direction of water diffusion at each voxel. In the brain, diffusion MRI provides anatomical information about white-matter fiber bundles that cannot be obtained by other non-invasive in vivo imaging methods.

Typical in vivo diffusion-weighted MR data suffers from low signal-to-noise ratio and significant imaging artifacts due to effects such as magnetic field inhomogeneities and eddy currents. Common DTI processing streams obtain distorted images from the MR data without modeling these effects, then attempt to reduce the distortions by post-processing the images, and finally estimate the diffusion tensors from the post-processed images.

In this talk, I will propose a penalized-likelihood (Bayesian) framework for estimating the unknown diffusion tensor map directly from the raw MR data. This framework incorporates a detailed model of the MR physics, including field inhomogeneities, eddy currents and imaging noise. It also allows for the smoothing of the tensor map via a prior that penalizes noisy solutions. The estimated tensors are constrained to be symmetric positive definite, thus ensuring non-negative eigenvalues. An iterative algorithm is used to compute the solution to this non-linear optimization problem. I will show results from preliminary evalution of the algorithm and discuss directions for future research.

John D. E. Gabrieli, Ph.D.
Grover Hermann Professor of Health Sciences and Technology and Cognitive Neuroscience, Harvard-MIT Division of Health Sciences and Technology, MIT Dept. of Brain and Cognitive Sciences
Wednesday, 01/17/07, noon, MGH bldg 149

Development of Declarative Memory Systems in the Human Brain
Little is known about the normal development of memory systems in the human brain. In three fMRI studies, we compared activations between children, ages 8-12, and adults. Relative to adults, children exhibited smaller brain regions that were selectively activated by faces and places, and worse recognition memory for faces and places. Children exhibited adult-like hemispheric specialization for verbal and spatial working memory tasks, but failed to exhibit the growth of activations adults showed in response to increasing working memory loads. During successful encoding of long-term memory for scenes, children exhibited adult-like magnitudes of activation in medial temporal lobe regions, but less activation than adults in prefrontal cortex. These studies reveal differential developmental trajectories for brain functions associated with working memory and long-term memory.

Doug Munoz, Ph.D.
Professor of Physiology and Psychology, Queen's University
Wednesday, 01/24/07, noon, MGH bldg 149

Combining Eye Movement Recording with Neurophysiology, Neuroimaging and Clinical Investigation to Probe Brain Function and Dysfunction
There is now considerable understanding of the neural circuitry controlling saccadic eye movements so that the eye movement system can now be used as a model system for investigating sensory, motor, cognitive, and clinical problems. In this seminar, Munoz will use a simple oculomotor task, the anti-saccade task, in combination with neurophysiological recording in animals, neuroimaging in humans, and behavioural assessment in patients in order to probe function and dysfunction in the eye movement system. Examples will be shown where research findings can be translated between approaches to gain further understanding of how the circuit controls behaviour.

Bruce Fischl, Ph.D.
Associate Professor of Radiology, Harvard Medical School
Wednesday, 01/31/07, 4pm, MIT 46-3189

Building Neuroanatomical Models with MRI
Models of brain structures generated from magnetic resonance imaging (MRI) data have grown in complexity in recent years, evolving from simple models with few classes such as gray matter, white matter and cerebrospinal fluid (CSF), into more complex ones representing multiple neural structures separately. This evolution has been possible due to developments in MR data acquisition technology that has yielded finer resolution, higher signal-to-noise ratio (SNR) images and an increasing number of contrast mechanisms, all of which have been used by increasingly sophisticated analysis tools to improve and extend classification. Nevertheless, despite these important advances, a critical unmet goal of this type of modeling is the generation of representations of myelo- and cytoarchitectonic boundaries from in vivo imaging data.

Brain cytoarchitecture remains the "Gold Standard" of neuroanatomy, as the cellular characteristics that define it are strongly linked to the functional properties of the brain, and the effects of many neurological disorders such as Alzheimer's disease (AD) are distinguished from healthy aging based on the specific cytoarchitectonic structures that are affected. Thus, being able to accurately quantify cytoarchitectural areas in vivo both in population studies, and in individuals, would be a major advance for basic systems neuroscience, as well as a tool for the evaluation of novel therapies, and even potentially for clinical diagnostic imaging. In this talk I will present some of our recent research to develop and validate methods that allow investigators to quantitatively and probabilistically model neuroanatomical structures as well as make inferences regarding the locations of architectonic boundaries from in vivo MRI data.

Stephanie Jones, Ph.D.
Instructor in Radiology at Harvard Medical School
Wednesday, 02/07/07, noon, MGH bldg 149

Neural Correlates of Tactile Detection: A Combined MEG and Biophysically Based Computational Modeling Study
Cortical correlates of sensory perception have typically been observed in higher order cortical areas, while there are conflicting reports as to their presence in primary sensory cortex. In this talk, I will present results from an investigation of the cortical correlates of somatosensory perception in human SI using a 2-fold approach that combined MEG brain imaging and computational neural modeling.

We recorded 306-channel MEG signals during a tactile detection task and extracted somatosensory evoked current dipoles that localized to the anterior bank of the post-central gyrus, area 3b, on structural MR images. Perception of the threshold level stimuli was predicted from ongoing state properties and from post-stimulus evoked responses in the SI signal. Perceived trials at threshold showed lower power in alpha (7-12Hz) and/or beta frequency bands (15-35Hz) in the signal immediately prior to stimulus onset (-500ms to onset). In the evoked response, the magnitude and timing of peaks in the early SI ECD waveform predicted perception beginning at ~70ms post-stimulus.

To make a direct and principled connection between the observed phenomena and the underlying neural dynamics, we developed a biophysically realistic computational neural model of a laminar SI cortex. The model incorporated the dendritic morphology and physiology of large pyramidal neurons known to be the primary generators of MEG ECDs. Results from the model led to the novel hypothesis that polarity and magnitude of peaks in the SI evoked response were induced by a sequence of lamina-specific excitatory drive to the local SI network. The sequence consisted of output from the granular layer (~25ms), followed by exogenous input to the supra-granular layers a (~70ms), and a second wave of granular output (~135ms), potentially reflecting initial input from the thalamus, followed by input from a higher order cortex or non-specific thalamus and a second wave of granular thalamic input. The model predicted further that correlates of perception in SI reflect subtle differences in the latency and magnitude of the longer latency inputs. Taken together, our findings strongly support the view that signatures of tactile detection are present in human SI, and are mediated by local neural dynamics induced by lamina-specific synaptic drive.

Armin Schwartzman, Ph.D.
Research Fellow, Harvard School of Public Health and Dana-Farber Cancer Institute
Wednesday, 02/14/07, noon, MGH bldg 149

A Log-Normal Distribution and Two-Sample Tests for the Full Diffusion Tensor
Diffusion Tensor Imaging (DTI) data differ fundamentally from most brain imaging data in that values at each voxel are not scalars but 3-by-3 positive definite matrices, also called diffusion tensors. Scalar summaries of the tensor do not capture all the information available in the data. For example, fractional anisotropy (FA) ignores spatial orientation. As a result, new statistical methods are needed for tensor-valued data. In this talk, I propose a new log-normal probability model for the full diffusion tensor. Based on this model, I derive notions of geometric mean and variance of diffusion tensors. Most importantly, I propose test statistics for testing whether the means of two samples of tensors (e.g. same voxel across two groups of subjects) are equal, have the same eigenvalues or have the same eigenvectors. These methods are shown in the context of a DTI study of reading ability in children.

Vince Calhoun, Ph.D.
Associate Professor of Electrical Engineering and Computer Engineering, University of New Mexico
Adjunct Associate Professor of Psychiatry, Yale University
Wednesday, 02/21/07, noon, MGH bldg 149

Data-Driven Brain Image Analysis: Algorithms & Applications
The analysis of structural and functional brain imaging data is challenging since, among other reasons, our understanding of the brain is far from complete and the contrast to noise of the data is typically quite low. In addition the data sets collected are typically very high in dimension and thus computational approaches are needed to reduce the problem to a tractable solution. One approach which is being used with increasing frequency is independent component analysis (ICA). Independent component analysis is a statistical method used to discover hidden factors (sources or features) from a set of measurements or observed data such that the sources are maximally independent. Typically, it assumes a generative model where observations are assumed to be linear mixtures of independent sources, and unlike principal component analysis (PCA) which uncorrelates the data, ICA works with higher-order statistics to achieve independence. ICA has recently demonstrated considerable promise in characterizing functional magnetic resonance imaging (fMRI) data, primarily due to its intuitive nature and ability for flexible characterization of the brain function. As typically applied, spatial brain networks are assumed to be systematically non-overlapping. Often temporal coherence of brain networks is also assumed, although convolutive and other models can be utilized to relax this assumption. ICA has been successfully utilized in a number of exciting fMRI applications including the identification of various signal-types such as task and transiently task-related and physiology-related signals in the spatial or temporal domain. Additional applications include the analysis of multi-subject fMRI data, the incorporation of a priori information, and the analysis of complex-valued fMRI data. In this talk, I will introduce ICA and its application to fMRI data analysis, and then review various applications of ICA to fMRI data. I will also discuss work in algorithm development. Finally, I will show some recent examples of joint analyses of multimodal data (e.g. combining fMRI with ectroencephalography (EEG) or structural imaging). The application of ICA to brain imaging data has proved to be quite fruitful. However there is still much work to be done in order to take full advantage of these rich data sets. One of the strengths of ICA is its ability to characterize the high-dimensional data in a concise manner. Continuing to do this and developing ways to mine the unexpected information in brain imaging data may lead to new breakthroughs in our understanding of the healthy and diseased brain.

ICA tutorial
Wednesday, 02/21/07, 3-5 pm, MGH bldg 149

Carl-Fredrik Westin, Ph.D.
Associate Professor of Radiology, Harvard Medical School
Wednesday, 02/28/07, 4pm, MIT 46-3189

White Matter Architecture from Diffusion MRI
In this talk I will present on-going activities in analysis of data from diffusion MRI at the Laboratory of Mathematics in Imaging (LMI), Department of Radiology, Brigham and Women's Hospital and Harvard Medical School.

The overarching goal of our diffusion analysis work is to pursue technological developments that improve the understanding of white matter anatomy. Such understanding, particularly due to the small dimension of the neural pathways relative to current imaging resolution, is vital to developing novel methods and techniques for the analysis of anatomical structures, and for the application of that analysis to understanding neural diseases. This information naturally complements information obtained from standard structural analysis. A driving clinical application for our work is to detect and localize white matter abnormalities in schizophrenia. We will discuss recent developments in spatial normalization of diffusion data (the first step in group analysis of brain imaging data is often to put the data in a common coordinate system), diffusion tractography, and methods fiber tract grouping for atlas generation. Compared to functional MRI analyses, group analysis paradigms for diffusion MRI data is currently a relatively unexplored area.

For details on our diffusion analysis work, see http://lmi.bwh.harvard.edu/papers/all.html

Joe Culver, Ph.D.
Assistant Professor of Radiology, Washington University School of Medicine
Wednesday, 03/07/07, noon, MGH bldg 149

Retinotopic Mapping of the Adult Human Visual Cortex with High-Density Diffuse Optical Tomography
Our lab explores ways of leveraging non-invasive optical measurements for both functional- and molecular- biological imaging. For application in humans we are developing non-invasive diffuse optical imaging (DOI) techniques. DOI methods have unique portability and hemodynamic contrast capabilities for mapping brain function in young subjects and subjects in enriched or clinical environments. However, DOI of human brain activity is challenging due to the concurrent requirements of high-dynamic range, low crosstalk, high channel counts, and sufficient temporal resolution. We have developed new instrumentation with improved performance characteristics that permits use of high-density imaging arrays and tomographic image synthesis. Current studies in the adult visual cortex are underway to evaluate imaging performance. Initial results show that the system can image functional features of the human visual cortex that were previously inaccessible to optical imaging such as eccentricity within quadrants of the visual field. The functional brain images have high contrast-to-noise, allowing visualization of individual activations and highly repeatable mapping within and across subjects. With the improved spatial resolution and localization, functional responses of 1.5 cm in extent and shifts of less than 1 cm are imaged. Cortical maps of angle and eccentricity in the visual field are consistent with retinotopic studies using functional magnetic resonance imaging (fMRI) and positron emission tomography (PET). These results demonstrate that high-density DOT is a practical and powerful tool for mapping function in the human cortex.

Susie Huang, Ph.D.
Harvard Medical School
Friday, 03/09/07, noon, MGH bldg 149

Avalanching Amplification of Sensitivity and Contrast in Magnetic Resonance Imaging by Nonlinear Feedback
Contrast in magnetic resonance imaging (MRI) is often limited by the dependence of the MRI signal on parameters that may not vary significantly throughout a sample. Here I describe a conceptually new approach to enhancing MR sensitivity and contrast that manipulates the intrinsic spin dynamics in the presence of nonlinear feedback fields induced by the spins themselves. This avalanching amplification scheme is shown to improve contrast-to-noise ratios by 15–24 times in epileptogenic lesions and malignant brain tumors, tissues with minimal contrast differences in routine MRI. Under the feedback interactions, the spins play an active role in determining and differentiating their subsequent evolution, thereby generating greater sensitivity to small changes in physiologically important MR properties not easily visualized by other MRI methods. Principles of rational pulse sequence design and the development of an active electronic feedback device for adapting feedback-enhanced contrast to small animal and ultimately human imaging at low field strengths will be discussed. The improved sensitivity and differentiation of local susceptibility variations by nonlinear feedback may aid in the distinction of changes in blood oxygenation level by functional MRI or local dipole fields induced by superparamagnetic iron oxide contrast agents for molecular imaging.

Steven M. Stufflebeam, M.D.
Director, Clinical Magnetoencephalography Service, Athinoula A. Martinos Center for Biomedical Imaging
Wednesday, 03/14/07, noon, MGH bldg 149

The Importance of Time in Clinical Neuroimaging
Timing is critical in normal brain function. In the laboratory, we are developing methods of integrating advanced imaging technologies, all of which have different spatial and temporal resolutions, such as Magnetoencephalography (MEG), diffusion tensor MRI, and functional MRI. Abnormal neural timing may contribute to the pathophysiology of many brain disorders including schizophrenia, epilepsy and reading disorders. Measuring brain activity at high temporal resolution is therefore critical. Using examples from the Clinical MEG Service at the Martinos Center, the lecture will illustrate how precise timing measurements are used in clinical functional neuroimaging, including epilepsy, schizophrenia, and brain tumors. Using auditory processing in schizophrenia as an example, a computational modeling based on neuroimaging data of the auditory steady state response is presented that could lead to a better treatment of this neuropsychiatric disease. Different measures of cortical connectivity will be contrasted using clinically relevant examples: structural connectivity, functional connectivity and effective connectivity. Web: http://www.nmr.mgh.harvard.edu/~sms

Daryl Hochman, Ph.D.
Assistant Professor of Surgery, Duke University Medical Center
Wednesday, 03/21/07, noon, MGH bldg 149

Optical Imaging of Cortical Activity
It has long been known that neuronal activity alters the way in which brain tissue absorbs and scatters light. These dynamic optical changes are known as 'intrinsic optical signals' (IOS) since they can be detected and imaged without the use of dyes or contrast-enhancing agents. IOS-imaging has recently been adapted for studies on human subjects undergoing neurosurgical treatment for drug-resistant epilepsy. Such intraoperative investigations offer a unique opportunity for studying the human brain, since data can be safely acquired directly from the exposed cortices of subjects. My presentation will focus on three issues: i) the physiological mechanisms underlying the generation of optical changes in cortical tissue, ii) a comparison of IOS-imaging to other imaging modalities such as BOLD-fMRI, and iii) how this technique is being used to map functional and epileptic activity in human subjects.

Elena Rykhlevskaia, Ph.D.
University of Illinois at Urbana-Champaign
Friday, 03/23/07, noon, MGH bldg 149

Modeling Anatomical and Functional Connectivity in the Brain
Over the last years a number of novel neuroimaging technologies have emerged, examples include diffusion tensor imaging and optical imaging. These new technologies require methods for statistical modeling and analysis that are uniquely tailored to the properties of the data. We developed a framework for statistical modeling of the dynamic cooperation among brain areas while exploiting probabilistic information about the presence of pathways that physically connect these areas. This framework includes lagged covariance structure modeling of functional connectivity, probabilistic tractography for the analysis of diffusion tensor imaging data, and approaches to combining anatomical and functional connectivity information in one model. We illustrate various components of the proposed statistical framework on experimental anatomical (collected with diffusion tensor imaging) and functional data (collected with optical imaging, as well as simulated).

F. DuBois Bowman, Ph.D.
Associate Professor of Biostatistics, Emory University
Wednesday, 04/04/07, noon, MGH bldg 149

Spatial Modeling of Functional Neuroimaging Data
Functional neuroimaging studies attempt to identify spatially localized brain regions that drive the execution of particular experimental tasks targeting, for example, cognition, emotion, or behavior. These studies also provide neural representations of the pathophysiology associated with psychiatric, addiction, and neurological disorders. Neuroimaging studies produce massive data sets comprised of serial scans on each subject, with each scan containing hundreds of thousands of spatially localized measurement sites (voxels). The data pose challenges for statistical analyses due to the enormity as well as to the complex patterns of spatial and temporal correlations. Temporal correlations are often addressed using variants of autoregressive models or other repeated measures covariance structures. Attempts to incorporate spatial correlations frequently involve pooling information from contiguous voxels in the brain. In this talk, we explore methods for modeling spatial correlations in neuroimaging data that extend beyond first-order neighbors, while also capturing temporal correlations across repeated scans for each subject. We illustrate the use of our methods on fMRI and PET functional neuroimaging data, from studies of schizophrenia and cocaine addiction, to determine task-related changes in measured brain activity and to assess functional connectivity within the brain.

Mark Mintun, M.D.
Professor of Radiology, Psychiatry and Neurobiology, Washington University School of Medicine
Wednesday, 04/11/07, noon, MGH bldg 149

Imaging Amyloid Plaques with PET: Where do we go from here?
Neurofibrillary tangles and beta-amyloid (Aβ) plaques are the pathological hallmark of Alzheimers disease (AD). The PET imaging tracer, N-methyl-[11C]2-(4'-methylaminophenyl)-6-hydroxybenzothiazole (or [11C]PIB for "Pittsburgh Compound-B"), has been shown to have significantly higher binding in subjects diagnosed with AD compared to nondemented controls. We have previously proposed that Aβ plaques exist in AD prior to the onset of dementia and have now confirmed that [11C]PIB PET scans can detect this proposed preclinical AD pathology with the same prevalence as seen in post-mortem. In this presentation, we will review the quantitative methodology associated with PIB imaging, and then introduce directions for future research with PET amyloid plaque imaging. In particular, examples will be presented of how we can use [11C]PIB imaging to substantially improve our understanding of early amyloid plaque deposition by correlating [11C]PIB data to other imaging and non-imaging biomarkers.

Andre van der Kouwe, Ph.D.
Instructor in Radiology at Harvard Medical School
Wednesday, 04/18/07, noon, MGH bldg 149

MRI Acquisition Methods for Brain Morphometry
A variety of sophisticated methods exist for improving MRI data quality and reducing artifacts offline after acquisition. However, there is still a lot of room to improve data quality at acquisition time before the images ever leave the scanner. In this talk I will review recent progress on three aspects of data acquisition. The first is a bandwidth-matched multiecho anatomical imaging protocol for use with the FreeSurfer analysis stream. The second is a framework for automatic prospective positioning for whole brain and subcortical structures. I will conclude with a review of recent progress on real-time motion correction. All of the sequences described in the talk are being developed on the Martinos Center scanners.

Nancy Kopell, Ph.D.
Professor of Mathematics, Boston University
Wednesday, 04/25/07, 4pm, MIT 46-3189

Gamma Rhythms: From Biophysics to Cognition
The nervous system produces electrical activity that can be parsed into frequency bands crudely associated with different cognitive states. The focus of this talk is on the gamma rhythm (30-90 Hz), which is associated with attention, learning, and early sensory processing. Different versions of the gamma rhythm can be produced in vitro, and there is evidence that there are also different versions in vivo. The talk discusses biophysical versions of gamma produced in vitro, and some case studies concerning gamma rhythms and function, including attention and plasticity.

Van Wedeen
Associate Professor in Radiology at Harvard Medical School
Wednesday, 05/02/07, noon, MGH bldg 149

Diffusion Tractography: New Vistas, New Issues
As tomographic imaging was an advance over earlier projection x-rays, so diffusion MRI tractography is a new way to represent and understand anatomy, and most of all, neuroanatomy. The need for such a method was announced in the classic Nature essay of Francis Crick [1993:361,109-10], pointing out that human connectivity was effectively unknown, and so it remains today. DTI tractography, a first response, was a step in the right direction, but proves too weak and inaccurate, as it is dependent on an infeasible assumption of a uniform fiber population in each voxel. To address this, high angular resolution diffusion MRI methods were invented, including DSI, QBI, HARDI and similar approaches, differing as to resolution and SNR. Recent comparison of DSI tractography with tracer anatomy in the macaque shows its capacity to accurately identify major cortical association pathways. Further studies show its potential to map fundamentals of cortical organization: overlapping cortical connection (at least) to association, striatal, commissural and projection systems. Note, however, that where tracer studies took years to show brain connectivity one region at a time, DSI gets all at once, in the same single brain, nondestructively, in a few days. Compared with DSI ex vivo, in vivo human DSI is 1000x lower in quality and looks it. This technological gulf presents a dilemma: How to draw useful inferences from existing data? Quite fundamentally, we lack models of tractography error and uncertainty, in part because the underlying microanatomy of the brain is both diverse and largely uncharted. We argue whats needed next is a reference atlas of the human to show for the first time what the human brain really looks like. We conclude that the problem of neural architecture is immensely greater than initially we could have imagined, and our power to reach for it is too.

Rahul Desikan, J.A.
Department of Anatomy and Neurobiology, Boston University School of Medicine
Wednesday, 05/16/07, noon, MGH bldg 149

MRI-Based Morphometric Measures of Regional Brain Atrophy in Prodromal Alzheimer's Disease
The initial stages of Alzheimer's disease (AD) are characterized by pathological changes in select brain areas. Structural MRI techniques can be used to measure volumetric alterations in individual brain regions and may provide an in-vivo characterization of these degenerative processes.

In the first study, an MRI-based automated atlas that subdivides the cerebral cortex into 34 neuroanatomic regions, per hemisphere, was developed from 40 MRI scans. Intraclass correlation coefficients, a measure of validity, demonstrated a mean value of .84, and mean distance maps, a measure of reliability, showed mean error of less than 1mm. These findings suggest that this atlas is both an anatomically valid and reliable method of subdividing the cortex into neuroanatomic regions of interest.

In the second study, this atlas was applied to 129 MRI scans obtained at baseline in subjects who were then followed clinically for an average of 6.2 years. Proportional hazards analyses were used to identify which brain regions predict time to progression from mild cognitive impairment (MCI) to AD and demonstrated that two regions, the entorhinal cortex (p < .004, hazard ratio = 0.57) and inferior parietal lobule (p < 0.05, hazard ratio = 0.68), were significant predictors.

In the third study, the atlas was applied to 66 sets of MRI scans obtained at two points in time (mean follow-up interval was 3.36 years). Annualized percent atrophy rates were differentially greater for the hippocampus, temporal pole, entorhinal cortex, middle and inferior temporal gyri (p-values < 0.01) in the MCI subjects who progressed to AD (converters) in comparison to the MCI subjects who did not progress to AD (non-converters). Comparisons between converters and elderly controls showed significant difference in the above-mentioned regions (except the inferior temporal gyrus) as well as in the inferior parietal lobule (p < 0.05).

These findings suggest that MRI-based morphometric methods may be useful in identifying those individuals with MCI at greatest risk for progression to AD. These findings further suggest that these methodologies may be useful for tracking the progression of disease over time.

Zhi-Pei Liang, Ph.D.
Professor of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign
Wednesday, 05/30/07, 4pm, MIT 46-3189

A Model-Based Approach to Fast Imaging with Sparse Sampling
A classical problem with tomographic imaging is the so-called limited data reconstruction problem, which occurs when physical and temporal constraints prevent sufficient coverage of the data space in the Nyquist sense. Traditionally, image reconstruction is performed using Fourier transform-based models, which often results in significant image artifacts (e.g., spurious ringing and loss of spatial resolution). To address this problem, numerous methods have emerged in the past two decades to incorporate a priori information into the imaging process. This talk will provide a systematic discussion of model-based imaging. I will begin with a brief review of the limited data problem in MRI and then discuss in detail how a priori information can be effectively utilized to achieve high spatial and temporal resolution in various biomedical applications (e.g., cardiac imaging, functional neuroimaging, and molecular cancer imaging).

Brad Dickerson, M.D.
Assistant Professor of Neurology, Harvard Medical School
Wednesday, 06/06/07, noon, MGH bldg 149

The Cortical Signature of Alzheimer's Disease: Structural and Functional Abnormalities Associated with and Predictive of Dementia
The symptoms of Alzheimer's disease and other neurodegenerative dementias relate to pathologic changes in specific vulnerable brain regions, including entorhinal cortex, hippocampus, and neocortical association areas. Much of what is known about these relationships has been learned from histologic studies of post-mortem tissue. MRI offers a window into brain anatomy and function in living individuals. New MRI methods, as well as a variety of analytic tools, are enabling us to make measurements of brain anatomy at an unprecedented level of detail. An improved understanding of the relationships between neurodegenerative changes in particular brain regions and specific clinical symptoms and signs may assist with early diagnosis or treatment monitoring. In this talk, I will review some of our latest findings on MRI-based measures of medial temporal lobe anatomy in very early AD, cerebral cortical thickness measures in AD and non-AD dementias, and regional brain function during memory tasks in aging and prodromal AD.

Simona Temereanca, Ph.D.
Research Fellow in Radiology at Harvard Medical School
Tuesday, 06/19/07, noon, MGH bldg 149

Eye Movement Effects on Word Processing during Reading: A Combined Approach Using MEG, EEG, High Resolution MRI and Psychophysics
Reading and visual exploration of the environment are essential human behaviors that involve rapid eye movements (saccades) to bring the visual target onto the fovea, the retinal area of highest visual acuity. Perceptual and cognitive processes are modified before, during and presumably after saccades, although the underlying neural mechanisms remain unknown. In this talk, I will present ongoing research that investigates the influence of saccades on post-saccadic neuronal processes devoted to word recognition, using a novel approach that integrates MEG, EEG, high resolution MRI and psychophysics methods. Previous work has used extensively eye movements and event-related potentials to words during stationary fixation as separate techniques, and much less in combination. Here, the electrooculogram is employed to detect saccades online and to trigger the foveal word presentation at different post-saccadic latencies (e.g. 0 ms, 50 ms, 100 ms, 150 ms and 600 ms).

One set of experiments examines the time-course of post-saccadic effects on behavioral and neuronal responses to foveally presented words. Observed post-saccadic modulation may reflect visual influences due to the rapid image motion over the retina during saccades, as well as influences attributed to motor processes or to interactions between eye movements and attention. To separate these visual, motor, and attentional components, in a second set of experiments we compare brain activations to words presented following saccades vs. following external background motion that mimics that during an eye movement. Correlations between cortical activation measures (latency and magnitude) and behavioral measures (reaction times and accuracy) are used to test the facilitatory or supressive nature of the effects. Results suggest that saccades may facilitate word processing by compensating for the suppressive visual effects following an eye movement.

Jin Hyung Lee, Ph.D.
Postdoctoral Scholar, Stanford University
Thursday, 06/21/07, 3:30pm, MGH bldg 149

Functional and Molecular Magnetic Resonance Imaging Technologies and Applications
Major challenges in functional brain imaging include spatial distortion and signal dropout. The BOLD contrast mechanism which relies on signal dephasing due to the presence of deoxyhemoglobin inherently sensitizes images to other sources of off-resonance that lead to image distortion with signal dropout regions. Furthermore, such distortions also limit spatial resolution. By using steady-state methods that enable robust contrast with short readout time and 3D imaging compatibility, full brain (including the susceptibility prone areas) coverage and high resolution sub-mm isotropic functional brain imaging can be achieved.

In molecular imaging using MRI, some of the key issues include the low sensitivity and control of the bio-distribution. A new contrast material that has very high sensitivity and potential for flexible bio-distribution control has been developed. The material poses an added advantage of having therapeutic effects when exposed to near IR radiation. These features can be combined to ultimately allow targeted visualization of diseases combined with therapy.

In this talk, with a focus on the two approaches outlined above, some of the technical developments and applications for functional and molecular imaging will be discussed.

Miguel Castelo-Branco, M.D., Ph.D.
Assistant Professor of Biophysics, Universidade de Coimbra
Wednesday, 07/11/07, noon, MGH bldg 149

Neural Mechanisms Underlying Motion Integration and Perceptual Decision: Psychophysics, fMRI and Multimodal Imaging Approaches
The cluttered nature of the visual world often imposes difficult perceptual decision problems which may even lead to paradoxical dissociations of perceived motion from the physical properties of the stimulus. Here, we have studied neural mechanisms that trigger visual segmentation of moving surfaces under variable conditions of stimulus ambiguity, and presence or absence of textured noise. The mechanisms triggering decision were studied using perceptually bi-stable stimuli: during fMRI experiments subjects were asked to report, by means of button presses, their interpretation of two superimposed gratings moving in different directions (plaid stimuli). Plaids may be perceived either as two surfaces, one being transparent and sliding on top of the other (transparent or component motion) or as a single coherent pattern whose direction of motion is intermediate to the component vectors (non-transparent or pattern motion). We found that the hMT+ complex directly encodes and triggers decision processes related to surface segmentation, even when perception is dissociated from veridical stimulus properties.

In this talk we will also address the question whether the perceptual interpretation of contextual surrounds can bias visual surface segmentation processes. We asked observers to report whether they perceived transparent or non-transparent surface motion under multiple contexts with varying luminance/contrast conditions and graded levels of baseline perceptual ambiguity. Surprisingly, only transparently perceived surrounds induced consistent (across all subjects) and significant perceptually congruent central bias, thereby vetoing motion integration. Furthermore, only surround transparency suppressed the durations of single incongruent percepts, as compared to non-transparent surrounds. These results are consistent with findings suggesting partially separable and hierarchically distinct neurophysiological substrates of pattern (non-transparent) and component (transparent) motion processes.

Finally, we will discuss a strategy that allowed us to directly correlate neural activity patterns with neuronal cell loss and psychophysical performance in Williams Syndrome, a genetic neurodevelopmental model to understand magnocellular processing, motion integration and dorsal stream dysfunction.

Vijay Parthasarathy, Ph.D.
Johns Hopkins University
Friday, 07/27/07, noon, MGH bldg 149

Optimal Use of Harmonic Phase (HARP) MRI: A Method to Measure Regional Cardiac Function Using MR Tagging
MR tagging is an imaging method that is used to encode tissue motion. Clinically, it is primarily used to measure regional cardiac motion from which cardiac mechanical function can be measured. One of the ways of extracting tissue motion from tagged MR images is the Harmonic Phase (HARP) method. HARP is as an image processing method, which uses only a small portion of k-space to compute the tissue motion. Special pulse sequences have been developed to acquire only the necessary amount of data for HARP processing, which has led to the possibility of real time cardiac functional imaging.

Although it is known that HARP needs only a small amount of data in k-space, it is not completely understood how much data is necessary and sufficient to describe cardiac motion completely in different clinical situations. Being a low resolution technique, it suffers from certain drawbacks, but the trade off is the speed at which data can be acquired and processed. In this talk, I will discuss ways to optimize the imaging and tagging parameters for optimal HARP processing. To do this, we will first study the sources of error in HARP processing, and we will use simulations to arrive at optimal tagging and imaging parameters for different clinical scenarios.

Britta Hahn, M.D.
NIH/NIDA
Wednesday, 08/08/07, noon, MGH bldg 149

Attention, Nicotine, and the Default Network of Resting Brain Function
A series of functional MRI studies will be presented designed to (a) dissociate different processes of visuospatial attention, and (b) characterize the effects of nicotine thereon. Parametric manipulation of bottom-up and top-down processes of attentional resource allocation revealed distinct networks of brain regions engaged by these functions. A different analysis approach based on trial-by-trial variation in reaction time identified rostral anterior and posterior cingulate cortices as activated with faster responding under conditions of stimulus unpredictability. These structures are part of the "default" network that is active in the absence of goal-directed activity, and appear to subserve a broad and spontaneous information gathering function. Nicotine improved attentional performance and enhanced task-induced deactivation of such default activity. Greater performance enhancement was associated with greater default deactivation, controlling for nicotine plasma levels. Control experiments did not suggest any confounds by non-specific vascular effects of nicotine.

Randy Gollub, M.D., Ph.D.
Assistant Professor in Radiology at Harvard Medical School
Wednesday, 08/15/07, noon, MGH bldg 149

Introduction to and tour of the GCRC Biomedical Imaging Core
Randy Gollub and Kashawna Harling will be giving a brief presentation and tour of the General Clinical Research Center (GCRC) Biomedical Imaging Core (BIC) facility. Conveniently located directly across the hall from the Brain Map Conference Room, this facility and its staff is here to serve the clinical (human subjects) research needs of the Martinos Imaging Center and the other clinical research groups in the Navy Yard.

If you ever need:

• trained personnel to help with blood draws to obtain samples for genetic studies, 
• access to a laboratory and/or trained personnel to process blood or other biological speicimens
• expert nursing or nurse practitioner assistance in performing routine or specific physical exams for inclusion, exclusion or follow up of subjects in your study
• an engineer with expertise to build or modify equipment to make it magnet compatible for imaging studies
• physiological monitoring equipment for a behavioral or imaging study
• statistical consultation for an imaging study design, power analysis, manuscript or grant preparation
• ancillary funding to cover the costs of screening or diagnostic laboratory tests
• electrically shielded research space for conducting sensitive EEG, optical or other data

Or if you ever just wondered what happens behind those doors!

Please attend this presentation so we can answer your questions directly.  We will review the resources available, introduce our staff, and explain the procedures for using the facility.  There will be time to address all questions and a tour of the facility so you can check it out for yourselves.

Please see our web site and contact Randy Gollub (rgollub at partners dot org) if you have any questions.