Brain Mapping seminar abstracts 2007-08

Andrew Maudsley, Ph.D.
Professor of Radiology, University of Miami
Wednesday, 09/12/07, noon, MGH bldg 149

Metabolic Mapping in Brain by Volumetric Proton MRSI
Distributions of magnetic-resonance observable metabolites in the brain can be efficiently mapped using volumetric MR Spectroscopic Imaging (MRSI) with echo-planar readout. The efficient processing of this data can, however, represent a challenge in light of the large amounts of data and the non-ideal characteristics of in vivo spectra. In this presentation a MRSI data processing environment will be described that integrates and automates multiple processing and analysis functions for imaging of proton metabolite distributions in the normal human brain. The image reconstruction methods make use of several sources of prior information, including MRI-derived tissue segmentation, morphological analysis, and spectral characteristics of the observed metabolites. Additional capabilities include normalization of metabolite signal intensities and transformation into a common spatial reference frame, thereby enabling the formation of a database of MR-measured human metabolite values as a function of acquisition, spatial, and subject parameters. This development is carried out under the MIDAS project (Metabolite Imaging and Data Analysis System). The analysis of multiple normal subject data demonstrate considerable heterogeneity of metabolite distributions in the brain, and preliminary results for clinical studies indicate that considerable advantage can be gained by performing voxel-based comparisons of individual subject data against normal metabolite distributions.

Penelope Kostopoulos, Ph.D.
McGill University
Wednesday, 09/19/07, noon, MGH bldg 149

The Ventrolateral Prefrontal Cortex and Memory Disambiguation
Although it is widely known that the prefrontal cortex plays a role in memory, the specific contribution of particular prefrontal regions in mnemonic functions remains controversial. A series of fMRI experiments were conducted to examine the specific role of the mid-ventrolateral prefrontal cortex in memory retrieval. It was hypothesized that, when stimuli in memory are related to each other with multiple associations and familiarity, strong stimulus-stimulus links or contextual cues are not sufficient for successful retrieval, control processing emanating from the mid-ventrolateral prefrontal cortex is required to disambiguate and select the appropriate information among memory traces. This type of retrieval is referred to as active retrieval to contrast it to automatic retrieval that depends on the simple re-activation of memory traces. Specific increases in activity were reported within the mid-ventrolateral prefrontal areas 45 and 47/12 during the performance of active retrieval tasks using visual, tactile, and auditory stimuli. These results indicate that the mid-ventrolateral prefrontal cortex plays a specific role in the disambiguation of information in memory for the active retrieval of relevant mnemonic information.

Amit Anand, M.D.
Associate Professor of Psychiatry and Radiology, Indiana University School of Medicine
Wednesday, 09/26/07, 4pm, MIT bldg 46

Novel Techniques for Imaging the Mood Regulating Circuit: Implications for the Diagnosis and Treatment of Depression
Recent neuroscience research indicates that brain functions, particularly higher brain functions such as emotion, which are implicated in psychiatric illnesses, are represented in circuits and distributed networks rather than in localized brain regions. In a series of studies, for which data will be presented, we have developed fMRI methods to measure local activation and distributed connectivity within the putative cortico-striatal-thalmaic-amygdalar mood regulating circuit (MRC). We have shown, with open-labeled antidepressant treatment, that decreased cortico-limbic connectivity and increased limbic activation is present in depression and that these abnormalities are reversed with antidepressant treatment. These techniques could potentially be used to predict and monitor the effects of both pharmacological and non-pharmacological treatments for depression.

Robert Handa, Ph.D. & Stuart Tobet, Ph.D.
Professors of Biomedical Sciences, Colorado State University
Wednesday, 10/03/07, noon, MGH bldg 149

Brain Sexual Differentiation: Inside stories...
Steroid hormones dramatically influence the development of numerous sites in the nervous system. Basic mechanisms in neural development provide foci for understanding how factors related to sex can alter the ontogeny of these regions. Sex differences in neurogenesis, cell migration, cell differentiation, cell death, and synaptogenesis are being addressed. Any and all of these events serve as likely targets for genetic or gonadal steroid-dependent mechanisms throughout development. Sex differences in the positioning of cells in the developing hypothalamus further suggest that cell migration may be one key target for early gene actions that impact long-term susceptibility to brain sexual differentiation. Estrogens have numerous effects on the brain, both in adulthood and during development. These actions of estrogen are mediated by at least two distinct estrogen receptor (ER) systems, ER alpha (ERα) and ER beta (ERβ). In brain, ERα plays a critical role in regulating reproductive neuroendocrine function and behavior, however, a definitive role for ERβ in any neurobiological function has been slow in forthcoming. Clues to the function of ERβ in the central nervous system can be gleaned from the neuroanatomical distribution of ERβ and the phenotypes of neurons that express ERβ. ERβ immunoreactivity has been found in populations of GnRH, CRH, vasopressin, oxytocin and prolactin containing neurons in the hypothalamus. Utilizing subtype-selective estrogen receptor agonists can help determine the roles for ERβ in a number of sex-dependent behaviors in rat models.

Nicholas Ayache, Ph.D.
Research director, INRIA/ASCLEPIOS, Sophia-Antipolis, France
Wednesday, 10/10/07, noon, MGH bldg 149

Computational Models of the Human Body for Medical Image Analysis
Medical image analysis brings about a revolution to the medicine of the 21st century, introducing a collection of powerful new tools designed to better assist the clinical diagnosis and to model, simulate, and guide more efficiently the patient's therapy. A new discipline has emerged in computer science, closely related to others like computer vision, computer graphics, artificial intelligence and robotics.

In this talk, I describe the increasing role of computational models of anatomy and physiology to guide the interpretation of complex series of medical images, and illustrate my presentation with three applications: 1) statistical modeling and analysis of sulcal lines on the brain cortex; 2) modeling and simulation of brain tumors evolution; 3) electro-mechanical modeling of the heart function for therapy planning and simulation. I conclude with some promising trends, including the analysis of in vivo confocal microscopic images.

Scott Hoge, Ph.D.
Instructor of Radiology, Brigham and Women's Hospital and Harvard Medical School
Wednesday, 10/17/07, noon, MGH bldg 149

Combining GRAPPA and SENSE to Improve Parallel MR Imaging
This talk will present a review of parallel MR imaging to illustrate the similarities and differences between the two dominant reconstruction algorithms: GRAPPA and SENSE. The inherent strengths of these methods are complementary, which allows one to combine them to great effect. We will demonstrate two such combinations: the use of GRAPPA to enable self-referenced SENSE reconstructions using very few auto-calibration lines, and the use of each algorithm in different temporal-frequency bands associated with UNFOLD. These examples demonstrate that, in certain cases, the combination of SENSE and GRAPPA will produce higher image quality than either method can provide independently.

Bruce Rosen, M.D., Ph.D.
Director, Martinos Center for Biomedical Imaging
Wednesday, 10/24/07, noon, MGH bldg 149

Functional Magnetic Resonance Imaging: An Historical Perspective

Paul M. Matthews, M.D., D.Phil.
Vice-President for Imaging, Genetics and Neurology, GlaxoSmithKline, UK
Wednesday, 10/31/07, noon, MGH bldg 149

Unravelling Single Gene Contributions to Brain Structure: Hope or Hype
The likely heritability of structure has been well-defined and is substantial. There has been hope that, as a "endophenotype", structural brain imaging could enhance the power of genetic association studies. Defining specific genes that contribute to brain structure should allow more detailed appreciation for mechanisms contributing to the heritability of intelligence and other quantitative traits, as well as providing insight into mechanisms of some brain disorders. In this lecture, the potential for integrated genetics and imaging to provide insight into genes that contribute to neurodegeneration is illustrated by a study of ApoE4 in Alzheimer’s disease. Evidence for novel genetic associations with brain structural variation in a population without neurological disease also is presented. Current problems with study power and validation will be highlighted.

Chun-yu Yip, Ph.D.
Electrical Engineering and Computer Science, University of Michigan
Monday, 11/12/07, 12:30pm, MGH bldg 149

RF Pulse Designs for Signal Recovery in T2*-Weighted Functional MRI
In blood-oxygenation-level-dependent (BOLD) functional magnetic resonance imaging (fMRI) using T2* contrast, images suffer from loss of signals at brain regions close to the air-filled cavities in the human head. The artifact arises from magnetic field distortion caused by the magnetic susceptibility difference between air and brain tissues, and hampers functional studies of important brain regions such as the orbito-frontal cortex. In this research project, I investigate two methods of designing radio-frequency (RF) pulses that can recover the signal loss. In addition to slice selective excitation, both pulse designs ``precompensate'' the through-plane dephasing that occurs between excitation and data acquisition. One method, which utilizes ``three-dimensional tailored RF pulses'', achieves these goals via three-dimensional spatially selective excitation. The other method uses spectral-spatial selective excitation, and relies on the assumption that through-plane dephasing is correlated with resonance frequency offset. All these sophisticated pulses are numerically designed using the iterative conjugate gradient method. To facilitate those design methods, I also propose new techniques applicable to general pulse designs, such as frameworks for pulse computation acceleration and joint design of excitation k-space trajectory and RF pulse. With phantom and human experiments, I demonstrate that the methods are efficacious in signal recovery, but not without costs and hurdles to overcome.

Peter Caravan, Ph.D.
Instructor in Radiology at Harvard Medical School
Wednesday, 11/14/07, noon, MGH bldg 149

Molecular MRI of Myocardial Collagen
Molecular MR imaging is attractive from a number of standpoints. It combines biochemical information with the high spatial resolution, deep tissue penetration, and excellent soft tissue contrast inherent in MRI. The molecular information yielded can be combined with anatomical and functional data acquired in the same exam. Molecular MRI uses stable isotopes which means the probes can be prepared in advance, used at any time, and exhibit no ionizing radiation. The major drawback is sensitivity of the probe. Typically micromolar (~nmol per gram tissue) concentrations are required to elicit contrast. To overcome this sensitivity issue, large assemblies of metal ions have been used, e.g. MION, polymeric agents, but the large size of these agents impacts their pharmacokinetics and tissue penetration. An alternate approach is to identify high concentration molecular targets of relevance. In this presentation myocardial collagen is targeted in order to assess myocardial perfusion.

First-pass contrast enhanced MRI for the study of myocardial perfusion and the assessment of coronary artery disease is a powerful technique, but the imaging tradeoffs necessitated by ultrafast imaging limit the degree of resolution and spatial coverage that can be obtained. Moreover the patient must be stressed while inside the magnet, two injections of contrast are required, and a time consuming data analysis is required. We sought an MR contrast agent that could be administered during stress and that provided persistent high resolution images reflective of blood flow at stress, i.e. combine the ease of nuclear cardiology with the much higher resolution of steady-state MRI. We hypothesized that interstitial myocardial collagen, present at levels compatible with detection by gadolinium enhanced MR, could act as a reservoir for a targeted MR contrast agent that distributed as a function of perfusion.

This presentation describes design, synthesis, and characterization of a gadolinium-based, type I collagen-targeted molecular MR contrast agent and reports its proof of concept in a porcine model of critical coronary stenosis.

Richard Leahy, Ph.D.
Professor of Electrical Engineering, Biomedical Engineering and Radiology,
University of Southern California
Wednesday, 11/28/07, noon, MGH bldg 149

Combining Cortical and Volumetric Analysis of the Human Brain
Intersubject registration of anatomical MR images is a necessary step in group analysis of structural and functional neuroimaging studies. Statistical power for detecting cortical effects can be enhanced by ensuring alignment of sulcal features, either through the use of explicit landmarks or through curvature matching. I will describe the approach we have taken to registration of cortical surfaces based on a set of sulcal curves, and the interactive computational tools we have developed for this purpose. I will also describe an extension of this approach to volumetric registration in which we use an initial surface alignment to ensure a one-to-one mapping between cortices while aligning subcortical structures through a combination of harmonic mapping and elastic energy minimization.

Ona Wu, Ph.D.
Instructor in Radiology at Harvard Medical School
Wednesday, 12/05/07, 4pm, MIT bldg 46

MRI-Based Imaging Models of Infarction after Acute Stroke
Stroke is a leading cause of death and morbidity in the US. Currently, the only FDA approved pharmaceutical intervention for the treatment of acute human ischemic stroke is thrombolysis with intravenous recombinant tissue plasminogen activator (rt-PA) administered within 3 h of stoke onset. Beyond this time frame, reduced clinical efficacy along with an increased risk of worsened outcome has been observed, and therefore rt-PA has been used cautiously. On the other hand, this 3h time window has been argued to be too narrow, unnecessarily excluding patients. An imaging tissue signature that can identify patients who will maximally benefit from treatment, or alternatively, who are at high risk of developing intracerebral hemorrhage, would provide a tremendous advance in patient management. Towards this goal, using supervised learning algorithms and acute MRI, we (1) developed tissue signature algorithms to predict infarction under natural history conditions, e.g. no rt-PA, (2) assessed efficacy of these natural history models for identifying treatment-induced changes in patients given rt-PA, and (3) developed treatment models to predict response to different therapies prior to administration. Using a generalized linear model (GLM), we combined multiple acute MRI data, such as diffusion-weighted and perfusion-weighted MRI. We were able to predict tissue outcome for natural history acute stroke patients and for experimental rat stroke models with high accuracy. Demonstrating the sensitivity of the algorithms for detecting an alteration in the natural progression of the ischemic cascade, the performance of the models was significantly reduced for the rt-PA group compared with the non-thrombolysed group. Using the GLM to predict outcome for different therapies, e.g. rt-PA or normobaric hyperoxia, the treatment models were fond to perform more accurately in patients who received treatment, with the estimated risk of infarction reduced compared to the risk maps produced by the natural history model. These results show that tissue signature maps generated with MRI-based statistical algorithms are promising metrics for diagnosis, prognosis and therapeutic evaluation after acute stroke that can translate readily from preclinical to clinical settings and a key step towards personalized medicine.

Mike Fox, Ph.D.
Department of Radiology, Washington University
Wednesday, 12/12/07, noon, MGH bldg 149

Spontaneous brain activity and variability in human behavior
The majority of functional neuroscience studies have focused on the brain's response to a task or stimulus. However, the brain is very active even in the absence of explicit input or output. Spontaneous fluctuations in the blood oxygen level dependent (BOLD) signal of functional magnetic resonance imaging are a potentially important and revealing manifestation of spontaneous neuronal activity. These studies have provided insight into the intrinsic functional architecture of the brain, variability in measured brain responses, and most recently variability in human behavior.

Chris Cocosco, M.Eng.
Image Sciences Institute, Utrecht University
Monday, 12/17/07, 10am, MGH bldg 149, research affairs room A

Automatic Image-Driven Segmentation of Medical Magnetic Resonance Images Without Prior Appearance Models
The focus of the presentation will be on original contributions in methods for the automatic segmentation of Magnetic Resonance Images (MRI) of the human head and heart. These methods achieve the segmentation without using previously trained image appearance or object shape models, and employ imperfectly supervised non-parametric classifiers and graph algorithms. The presentation will also include a brief overview of other past research.

Daniel C. Javitt, M.D., Ph.D.
Professor of Psychiatry and Neuroscience, New York University School of Medicine and Nathan Kline Institute for Psychiatric Research
Wednesday, 01/16/08, noon, MGH bldg 149

Bottom up contributions to cognitive dysfunction in schizophrenia
Schizophrenia is associated with cognitive deficits that represent a core component of the disorder. Deficits have been extensively documented in domains such as executive processing, working memory, and learning/declarative memory formation, which localize to high-level cortical regions such as frontal cortex or hippocampus. In contrast, integrity of sensory processing in schizophrenia is less well appreciated. This presentation will present recent work demonstrating impairments in early stage visual and auditory processing in schizophrenia, and will highlight contributions of low level sensory processing deficits to more holistic processes such as perceptual closure, reading and voice emotion recognition. Overall patterns of neurocognitive dysfunctions will be discussed relative to predictions of the PCP/NMDA model and other glutamatergic conceptualizations of schizophrenia.

Diego Contreras, M.D., Ph.D.
Associate Professor of Neuroscience, University of Pennsylvania
Wednesday, 01/16/08, 3pm, MGH bldg 149

Input Integration in Different Network States Studied with Intracellular and Optical Recordings in Vivo
Intracellular recordings from thalamus and cortex in the rat whisker-barrel system reveal that synaptic inputs originating from separate whiskers may add sublinearly or suppress each other depending on the stimulation interval. The suppression is due to lateral inhibition in thalamus and not in cortex. Current source density analysis reveals that the rules of summation do not change after cortical inactivation but the time course of suppression is reduced when anesthesia level is reduced. Voltage sensitive dyes provide support to the integration rules revealed with intracellular recordings and compare the differences between the depolarizing and hyperpolarizing phases of the slow oscillation.

Sylvain Baillet, Ph.D.
Head, Neuroimaging Group, Centre National de la Recherche Scientifique
Wednesday, 01/23/08, noon, MGH bldg 149

Mapping and Tracking the Flow of Brain Activity using MEG/EEG: Hypothesis and Methods
The exquisite temporal resolution of electromagnetic measurements using MEG and EEG combined with anatomical MRI leads to high-density sequences of images of brain activity. The analysis and appraisal of such data are high-dimensional and not deprived of subjectivity.

We develop the concept of 'cortical flow' as a methodological extension to the empirical description of fast activity that unfolds onto the cortical surface, as revealed by MRI-constrained electromagnetic brain imaging. The methods involved include the computation of the optical flow of 3D manifolds and the detection of metastable episodes in the large-scale dynamics of cortical currents. We show how this concept relates to empirical investigations on 'microstates' and computational models of neural ensembles.

Time permitting, we would also like to discuss our recent developments on anatomical coregistration based on the diffeomorphic matching of brain surfaces based on their respective sulcal imprint.

David Salat, Ph.D.
Assistant Professor in Radiology at Harvard Medical School
Wednesday, 01/30/08, 4pm, MIT bldg 46

A Review of Diffusion Tensor Imaging Studies of Aging and Alzheimer's Disease
There is strong interest in the use of diffusion tensor imaging (DTI) for the study of brain anatomy and pathology. DTI has been particularly important in the characterization of changes in brain tissue integrity with aging and age associated neurological disease. Recent data demonstrate the potential for DTI to provide useful metrics in the basic and clinical aspects of a variety of aspects of these conditions; however, controversies exist about the biological significance of diffusion measures as well as the technical aspects of DTI analysis. Thus, a variety of questions remain about how DTI can best be utilized to understand age associated neural change.

This Brainmap will review the current literature utilizing DTI in the study of aging and Alzheimer's disease, and will present some recent data on this topic. Specifically, we utilized DTI to examine tissue changes in individuals with Alzheimer.s Disease and find a network of tissue pathology in that exceeds and complements more commonly employed prior imaging methods. Overall, although limitations remain, this review demonstrates that a wide range of information can be obtained from current DTI procedures, and that these various metrics could provide important information about disease mechanisms.

Mary Phillips, M.D.
Co-Director, Brain Imaging Research Center, Carnegie Mellon University and University of Pittsburgh
Professor of Clinical Affective Neuroscience, Department of Psychological Medicine, Cardiff University
Wednesday, 02/06/08, noon, MGH bldg 149

Imaging Emotion in the Brain: from Normal Emotion to Abnormal Mood States
My long-term research goal is to identify specific functional abnormalities in neural systems underlying the different subprocesses important for emotion perception and emotion regulation that may be present as biomarkers of major psychiatric disorders, including bipolar disorder, unipolar depression, schizophrenia and anxiety disorders. The identification of these neural markers of psychiatric disorder is, I believe, a crucially important step toward the long-term goal of improving diagnostic accuracy with the use of reliable biomarkers in individuals presenting in the early stages of psychiatric illness. In the presentation, I will describe data from my research groups in Pittsburgh and the UK indicating functional and structural abnormalities in neural systems underlying different component processes of emotion perception that may distinguish individuals with different mood disorders, such as bipolar disorder and unipolar depression. I will also present data from my research group in Pittsburgh, where we have begun to examine the extent to which these neuroimaging markers may be present in individuals at high genetic risk of subsequently developing bipolar disorder and unipolar depression to aid the identification of potential risk markers of disorder development. I will finally present data from a provisional, naturalistic treatment study of unipolar depression, where the focus has been examination of the extent to which neuroimaging markers of dysfunction in neural systems underlying the different component processes of emotion processing may act as predictors of response to specific treatments.

Lisa Feldman Barrett, Ph.D.
Professor of Psychology, Boston College
Wednesday, 02/13/07, noon, MGH bldg 149

What is an emotion?

Hesheng Liu, Ph.D.
Research Fellow in Radiology at Harvard Medical School
Wednesday, 02/20/08, noon, MGH bldg 149

Functional Mapping Using fMRI, fcMRI, iEEG, and MEG/EEG -- Preliminary Findings with Motor Mapping, Visual Selectivity and Memory Retrieval
Several functional mapping techniques were employed on surgical patients and normal subjects. In the first part of this talk, I evaluate the feasibility of resting state fMRI (fcMRI) for task-free pre-surgical functional mapping. This new approach may be useful for the patients who are unable to perform tasks in the scanner, or who require several functional maps in one session. In the second part, I will show preliminary results from a visual object recognition study using intracranial EEG (iEEG). Our findings indicate that the temporal lobe shows selectivity to multiple object categories. The selectivity is invariant when the size and orientation of the objects change. The selectivity also remains when patients are viewing dynamic pictures. In the final part, I will discuss a fMRI, MEG, EEG study of episodic memory retrieval.

Carl Schwartz, M.D.
Director, Developmental Neuroimaging and Psychopathology Research Laboratory, Harvard Medical School Department of Psychiatry at MGH/Martinos Center
Wednesday, 02/27/08, noon, MGH bldg 149

Infant temperament, Adolescent psychiatric disorder, and Neuroimaging: Pitfalls and Possibilities
Temperament refers to the individual differences in emotions, cognitions, and behaviors based on biology that are first observed in infancy or early childhood. Temperament is important because of its role in both normative development and as a risk factor for psychiatric disorder. We review our earlier studies of the relationship between childhood temperament and adolescent anxiety disorder. We then discuss fMRI studies from our lab that offer insights into the neural bases of differences in temperament. We will discuss inhibited temperament as a putative endophenotype for anxiety disorder, and the potential benefit of such an approach to other psychiatric illnesses including depression and autism. Possible pathways from temperament to disorder will be discussed. Limitations and potential fallacies in common experimental designs for fMRI, and as well as implications for psychiatric diagnosis and treatment will be discussed.

Chris B. Schaffer, Ph.D.
Assistant Professor of Biomedical Engineering, Cornell University
Wednesday, 03/05/08, noon, MGH bldg 149

All-Optical Animal Models of Small-Scale Stroke
Clinical evidence shows that ischemic and hemorrhagic microvascular lesions in the brain play an important role in elderly dementia, but few effective treatment or preventative strategies exist. This deficit is due, in part, to a lack of good animal models of these small-scale strokes that would allow the progression of disease to be studied and would provide a platform for the evaluation of therapeutics. Here, we use advanced optical techniques to induce single-vessel occlusions and hemorrhages in the cortex of live, anesthetized rodents, as a means to provide a comprehensive animal model of small-scale stroke. A tightly-focused femtosecond laser pulse is used to deposit laser energy into the endothelial cells that line a specifically targeted vessel causing an injury that triggers clotting or causes hemorrhage, but only in the targeted vessel. This technique allows any blood vessel, including individual arterioles, capillaries, and venules, in the top 0.5 mm of the cortex of a rodent to be selectively lesioned. We also use optical techniques, such as two-photon excited fluorescence microscopy, to study the physiological consequences of these occlusions. To date, we have investigated how blood flow reroutes after the occlusion of a single vessel for clots located at different positions in the vascular hierarchy. This study has pointed to the extreme vulnerability of cortical perfusion to occlusions in penetrating arterioles and to the remarkable robustness of the cortical surface arteriole network. Ongoing work aims at determining the effect of microvascular lesions on the health and functionality of nearby brain cells.

Earl Miller, Ph.D.
Picower Professor of Neuroscience, The Picower Institute for Learning and Memory and Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology
Wednesday, 03/12/08, noon, MGH bldg 149

Rules, Concepts, and Executive Brain Functions
What controls your thoughts? How do you focus attention? How do you know how to act while dining in a restaurant? This is cognitive control, the ability to organize thought and action around goals. I will present theories and evidence on how this arises from interactions between cortical areas (especially the prefrontal cortex) and the basal ganglia. Results from our laboratory have in these areas have properties commensurate with a role in "executive" brain function. They are involved in directing attention, in recalling stored memories, predicting reward value, and they integrate the diverse information needed for a given goal. Perhaps most importantly, they transmit acquired knowledge. Their activity reflects learned task contingencies, concepts and rules. In short, they seem to underlie our internal representations of the "rules of the game" needed for goal-directed behavior.

Leonid Perlovsky, Ph.D.
Harvard University and Air Force Research Lab
Wednesday, 03/19/08, noon, MGH bldg 149

Language, Cognition, and Cultural Evolution
What are the mechanisms of interaction between language and cognition? The talk briefly reviews past mathematical difficulties of modeling the mind and new mathematical techniques of dynamic logic and neural modeling fields, which overcome these difficulties. Mechanisms of concepts, emotions, instincts, imaginations, intuitions are described; they are inseparable from perception and cognition. Engineering applications illustrate orders of magnitude improvement in pattern recognition, data mining, fusion, financial predictions. Dynamic logic is extended to language and mechanisms of joint operations of language and cognition. It turned out those human abilities could only evolve jointly.

The last part of the talk moves to future research directions: roles of beautiful, music, sublime in the mind, cognition, and evolution. I relate dynamic logic to the knowledge instinct, which drives the mind to understand the world, and argue that instinct is even more important than sex or food. Two mechanisms of the knowledge instinct, differentiation and synthesis determine evolution of cultures, and explain two types of cultural evolutions. The knowledge-acquiring cultures leading to science and technology, but doubting their values, and stagnating cultures investing their limited amount of knowledge with ever increasing emotional value. English language leads to the first type of culture, and Arabic language leads to the second type. The Nobel Prize in Economics, 2003 is related to the Biblical story of the fall. Mathematical models of the mind and cultures bear on contemporary world, and may be used to improve mutual understanding among peoples around the globe and reduce tensions between cultures.

James Gee, Ph.D.
Associate Professor of Radiologic Science and Computer and Information Science, University of Pennsylvania
Wednesday, 03/26/08, 4pm, MIT bldg 46

Structure-Specific Normalization and Analysis for Multi-Modality Brain Imaging
In multi-subject structural, functional or diffusion-weighted MRI studies, it is common to analyze the imaging data using whole-brain approaches, such as voxel-based or deformation-based morphometry. Whole-brain analysis is a powerful tool for exploratory research, but in studies where there are a priori hypotheses about individual anatomical structures, a more structure-centric approach to image analysis may be appropriate. This is especially true for complex structures such as the hippocampus, which tend to not be adequately normalized by whole-brain techniques. This talk will focus on recent work at Penn that uses deformable geometrical models as a framework for shape-based normalization, smoothing and analysis of individual anatomical structures. Several applications will be discussed, including a new detailed atlas of the hippocampus derived from high-field postmortem MRI; hippocampus-specific analysis of fMRI activation in memory tasks, as applied to predicting memory outcome in epilepsy surgery; and a new surface-based framework for DTI analysis.

Anna Devor, Ph.D.
Assistant Adjunct Professor, Department of Neurosciences and Department of Radiology, University of California San Diego
Friday, 04/04/08, noon, MGH bldg 149 (conf. room A, 1st floor)

Evidence that Neuronal Signaling and not Energy Consumption Controls the Hemodynamic Response
We examined bilateral neuronal and hemodynamic changes, and glucose uptake in response to a unilateral somatosensory stimulus in rat primary somatosensory cortex (SI). In contrast to the contralateral forepaw area where neuronal activity, blood oxygenation/flow and glucose uptake increased in unison, we observed blood oxygenation/flow decrease and arteriolar vasoconstriction in presence of increased glucose uptake in the ipsilateral SI. Voltage-sensitive dyes measurements revealed sequential de- and hyperpolarization with a comparable amplitude bilaterally. Ipsilateral hyperpolarization was stronger, suggesting stronger recruitment of inhibitory interneurons in ensemble response. Laminar electrophysiological recordings revealed an increase in ipsilateral spiking consistent with the observed increase in glucose uptake. The vasoconstriction and decrease in blood flow in presence of an increase in both neuronal spiking and glucose uptake in ipsilateral SI argues against feed back signaling by energy metabolites. Rather, our results are consistent with feed forward neuronal control of vasodilation and vasoconstriction.

Dmitriy Yablonskiy, Ph.D.
Professor of Radiology and Physics, Washington University in St. Louis
Wednesday, 04/09/08, noon, MGH bldg 149

Quantitative BOLD: Mapping of Human Cerebral Deoxyhemoglobin-Containing Blood Volume and Oxygen Extraction Fraction
Since fundamental discovery by Ogawa and co-workers of BOLD contrast in MRI substantial progress has been achieved in understanding the nature and mechanisms underlying the BOLD signal. However, most work has been directed at the study of dynamic BOLD temporal changes in the MRI signal during functional activation. At the same time, very little attention has been paid to studying the nature and mechanisms of BOLD contrast in the resting (baseline) state of brain activity which can be described as a default mode of the brain activity.

Quantitative evaluation of brain hemodynamics in the baseline state is important for understanding normal human brain functioning as well as understanding the pathophysiology of disorders such as stroke, Alzheimers disease, Huntingtons disease, Parkinsons disease and other neurological disorders. It can also be of great importance for evaluation of hypoxia within tumors of the brain and other organs. Currently, clinically accepted methods for measuring cerebral blood flow (CBF) and oxygen extraction fraction (OEF) main parameters characterizing brain hemodynamics rely on PET techniques.

In my presentation I will describe a technique for in vivo quantitative evaluation of brain hemodynamic parameters that has been under development in our laboratory. The technique quantitative BOLD (qBOLD) is based on BOLD phenomena and allows quantitative measurements of OEF and deoxyhemoglobin containing blood volume (DBV).

Phillip Zhe Sun, Ph.D.
Assistant Professor in Radiology at Harvard Medical School
Wednesday, 04/16/08, noon, MGH bldg 149

Imaging pH Using Chemical Exchange-Dependent Saturation Transfer (CEST) MRI
Chemical exchange saturation transfer (CEST) MRI, first proposed by Ward and Balaban, offers a remarkably sensitive detection mechanism that allows characterization of dilute proteins/peptides and microenvironment properties. For instance, CEST agents that are sensitive to tissue energy metabolites/byproducts including glucose, glycogen and lactate as well as pH and temperature have been reported. Given that CEST imaging is capable of capturing cellular and molecular events that are often inaccessible by conventional MRI, it may significantly enhance our understanding of tissue pathophysiology and thus, holds great promise for developing therapeutics and guiding treatments. In fact, in vivo amide proton transfer (APT) MRI, a variant of CEST MRI, has been shown sensitive to ischemic tissue acidosis and tumor proteins/peptides content change, and thus, complementing the conventional MRI approach. While on the other hand, it is also known that CEST MRI is susceptible to concomitant RF irradiation and field inhomogeneity artifacts and varies with experimental conditions, and consequently, the sensitivity and specificity of CEST MRI is not optimal. To overcome these limitations, we have developed analytical and numerical models to quantitatively describe the CEST contrast, proposed concise correction algorithms for minimizing measurement errors, and ultimately demonstrated absolute pH imaging using CEST MRI. Recently, we have further exploited in vivo CEST MRI and showed that tissue acidosis, assessed by APT MRI, serves as a metabolic imaging marker for disrupted oxygen/glucose metabolism and can help predict final tissue infarction, augmenting the commonly used perfusion and diffusion MRI methods. Principles of CEST MRI and its in vivo applications will be presented.

Better Images of the Thinking Brain: Advances in Image Reconstruction and Motion Correction
Magnetic susceptibility, the material property of becoming magnetized in response to an applied magnetic field, plays a fundamental role in the generation of the desired signal in the most common form of using magnetic resonance imaging (MRI) to image human brain function (functional MRI or fMRI). Unfortunately, magnetic susceptibility is also a major source of static image distortions as well as dynamic signal variations the limit functional MRI. In this talk, I will describe how susceptibility leads to both spatially distorted images, signal losses in certain regions of the brain, and increased image variability due to head motion. We have developed MRI acquisition and image reconstruction approaches that individually and together can reduce these effects. The image reconstruction approaches follow a general framework of modeling the physical processes and iteratively estimating the object as well as the distorting processes. This provides dynamic estimates that qualitatively improve overall image appearance and quantitatively more robust corrections for head motion.

Martin Styner, Ph.D.
Research assistant professor of Computer Science, University of North Carolina
Wednesday, 04/30/08, 4pm, MIT bldg 46

DTI Analysis in the Developing Brain of Humans and Non-Human Primates
Diffusion tensor imaging (DTI) has become increasingly important as a means of investigating the structure and properties of neural white matter, with many applications in our own research in the developing brain of humans and macaque monkeys. Several analysis frameworks have been proposed, such as region based, voxel-based as well as fiber tract based framework. I will present our current work in this field using both fiber tract and region based analysis. These methods will be illustrated in applications of normal brain development in rhesus macaques, as well as neonatal white matter pathology in Krabbe's disease. I will further describe two ongoing development efforts: a) towards an automated analysis of atlas based, parametrized DTI fiber tracts and b) the use of DTI based connectivity information as the basis for a novel, population based cortical correspondence computation.

David Poeppel, Ph.D.
Professor of Biology and Linguistics, Neuroscience and Cognitive Science Program, University of Maryland College Park
Friday, 05/02/08, noon, MGH bldg 149

Temporal Primitives in Auditory Cognition and Speech Perception
Generating usable internal representations of speech input, or an auditory scene more generally, requires, among other operations, fractionating the signals into temporal units/chunks of the appropriate granularity. Adopting (and adapting) Marr's approach to vision, a perspective is outlined that formulates linking hypotheses between specific neurobiological mechanisms (for example cortical oscillations and phase-locking) and the representations that underlie auditory cognition (for example syllables). Focusing on the implementational and algorithmic levels of description, I argue that the perception of sound patterns requires a multi-time resolution analysis. In particular, recent experimental data from psychophysics, MEG (Luo & Poeppel, 2007), and concurrent EEG/fMRI (Giraud et al., 2007) suggest that there exist two privileged time scales that form the basis for constructing elementary auditory percepts. These "temporal primitives" permit the construction of the internal representations that mediate the analysis of speech and other acoustic signals.

Helen Mayberg, M.D.
Professor of Psychiatry, Emory University
Wednesday, 05/14/08, noon, MGH bldg 149

Modulating Putative Depression Circuits Using Deep Brain Stimulation
Critical to development of deep brain stimulation (DBS) as a novel therapy for intractable depression, has been the evolving understanding of brain systems mediating normal and abnormal mood states and the ongoing, systematic characterization of neural substrates mediating successful and unsuccessful response to other antidepressant interventions. Based on previous work implicating the subcallosal cingulate (Brodmann Area 25) and its functional connections to specific paralimbic, cortical and subcortical regions in the pathophysiology of depression and antidepressant response mechanisms, we have piloted the use of chronic high frequency DBS to modulate BA25 connectivity in patients with treatment resistant depression. The theoretical and data-driven foundation for developing this new procedure as well as long-term clinical, neuropsychological and imaging findings from two ongoing experimental studies will be presented.

Wolf-Dieter Heiss, M.D.
Professor of Neurology University of Cologne
Wednesday, 05/21/08, noon, MGH bldg 149

Multimodal Functional Imaging by PET in Gliomas: Pathophysiology and Treatment Strategies
Imaging in patients with brain tumors aims toward the determination of the localization, extend, type, and malignancy of the tumor. Imaging is being used for primary diagnosis, planning of treatment including placement of stereotaxic biopsy, resection, radiation, guided application of experimental therapeutics, and delineation of tumor from functionally important neuronal tissue. After treatment, imaging is being used to quantify the treatment response and the extent of residual tumor. At follow-up, imaging helps to determine tumor progression and to differentiate recurrent tumor growth from treatment-induced tissue changes, such as radiation necrosis. A variety of complementary imaging methods are currently being used to obtain all the information necessary to achieve the abovementioned goals. Computed tomography and magnetic resonance imaging (MRI) reveal mostly anatomical information on the tumor, whereas magnetic resonance spectroscopy and positron emission tomography (PET) give important information on the metabolic state and molecular events within the tumor. Functional MRI and functional PET, in combination with electrophysiological methods like transcranial magnetic stimulation, are being used to delineate functionally important neuronal tissue, which has to be preserved from treatment-induced damage, as well as to gather information on tumor-induced brain plasticity. In summary, imaging in patients with brain tumors plays a central role in the management of the disease and in the development of improved imaging-guided therapies.

John Ashburner, Ph.D.
University College London
Wednesday, 05/28/08, noon, MGH bldg 149

Multivariate Analysis of Anatomical MRI
This talk will describe the methods used in the Functional Imaging Lab (London, UK) for multivariate analysis of grey matter maps. The procedure begins by segmenting the anatomical images into their different tissue classes, using a generative modelling approach similar to that of Fischl or van Leemput. The model combines tissue classification, bias correction and nonlinear registration of pre-generated tissue probability maps. The segmentation results are used to generate approximately rigidly aligned gray and white matter images for each subject. More precise inter-subject registration is then performed by repeatedly aligning the grey and white matter images to their own mean, using a high-dimensional warping approach that preserves the one-to-one mapping among brains. Warped versions of the gray matter images are generated, which are locally scaled in order to compensate for expansion/contraction during warping. These pre-processed data are then used as features for multivariate kernel classification or regression.