Yale

Seminar Series in

Bioimaging Sciences



Please note that these talks are held in N135 TAC which is the conference room at the MRRC located at 300 Cedar Street (corner of Congress and Cedar).

Refreshments served 15 minutes prior to start of seminar.

Please call 5-6622 for directions.


Previous Seminars

Fall 2003


Dec 16 (Tuesday), 4:15 pm
John Detre, M.D. (University of Pennsylvania)
ASL perfusion fMRI

The majority of fMRI studies have used blood oxygenation level dependent (BOLD) contrast as a surrogate marker for regional neural activation. An alternative contrast mechanism for visualizing regional brain function provides direct measurement of cerebral blood flow (CBF) using MRI. This class of techniques is termed arterial spin labeled (ASL) perfusion MRI, and utilizes magnetically labeled arterial blood water as endogenous tracer for measuring CBF. ASL methods demonstrate clinically meaningful alterations in baseline CBF in a variety of CNS disorders. ASL methods are also useful for imaging changes in CBF in response to task activation or pharmacological challenges. Existing data demonstrate that 1) CBF measured using ASL perfusion MRI is accurate and reproducible, 2) ASL contrast shows stable noise characteristics over the entire frequency spectrum, making it suitable for studying sequential changes in resting or activated brain function over long periods, 3) CBF-based task-activation measured by ASL shows less intersubject variability than BOLD, and 4) ASL perfusion fMRI can be measured using pulse sequences that are insensitive to susceptibility effects. ASL methods may be particularly suitable for populations in which knowledge of baseline CBF contribute to the interpretation of activation results. This presentation will summarize methodological considerations in ASL perfusion MRI and illustrate its applications in basic and clinical neuroimaging.

Dec 9 (Tuesday), 4:15 pm
Michael Garwood, Ph.D. (University of Minnesota)
Magnetic resonance spectroscopy and imaging of human breast at 4 Tesla

Improved diagnostic tools are needed to reliably detect breast cancer and to assess therapeutic response early in the course of treatments. At the University of Minnesota's Center for Magnetic Resonance Research, we are exploring the use of high magnetic field (4 Tesla) MR imaging and spectroscopy of breast. As part of this work, we have developed new MR coils, scanning techniques, and processing methods to exploit the advantages of high field. Results to be presented show that high field MRS provides valuable information which improves diagnostic accuracy and allows early identification of therapeutic response vs. non-response. The latter capability will ultimately allow oncologist to individualize treatments to achieve maximal benefit and to speed the transfer of new drugs from the laboratory to clinic.

Jianhui Zhong, Ph.D. (University of Rochester)
Oct 23 (Thursday), 4:15 pm
Imaging from intermolecular multiple-quantum coherence: Do we need another contrast mechanism?

Dr. Zhong is Professor in the Departments of Diagnostic Radiology and Biomedical Engineering, Schools of Medicine and Engineering at the University of Rochester, NY. The list of different MR techniques and related contrast mechanisms continues to grow even after more than 50 years of development and four Nobel Prizes. One of the reasons is that with the introduction of every new technique, there is hope (or hype) of achieving a new level of understanding of underlying physio-/pathology revealed by the measurement. We have done some preliminary works with intermolecular multiple-quantum coherence (iMQC) imaging both in animals and in humans. In this talk I will demonstrate that iMQC signals possess some characteristics different from conventional MRI signal, and can be used potentially to study functions and microstructures of biological interests. Fundamental issues including low intrinsic SNR in iMQC and possible remedies will also be discussed.

Spring 2003


Robert Mach, Ph.D.
Receptor imaging studies in nonhuman primate models of aging and substance abuse
Tues, Apr 29, 4:00 PM (refreshments at 3:45 pm)
CAB N135 MRRC, CAB Building 300 Cedar Street

Dr. Mach is Professor of Division of Radiological Sciences at the Mallinckrodt Institute of Radiology in Washington University School of Medicine, St. Louis, MO.

Prof. Mach's laboratory is actively participating in synthesis and biological evaluation of ligands for imaging the activity of various receptors in vivo. He will discuss some recent PET work on primate models.

Derek Toomre, Ph.D.
Cellular highways, traffic jams, explosions and the (Q-) bridge: new insights with advanced live cell imaging approaches
Tues, Feb 18, 2003, 4:00 PM (refreshments at 3:45 pm), MRRC, CAB Building, N135 300 Cedar Street

Dr. Derek Toomre is a new faculty with the Department of Cell Biology and is also the Director of a new light microscopy imaging center called "The CINEMA laboratory" (Cellular Imaging using New Microscopy Approaches) located in the Yale School of Medicine. As a post-doc he trained with Dr. Kai Simons at the EMBL in Heidelberg, Germany.

The application of new advanced live cell imaging (and related) approaches are providing new insight to the dynamics and organization of important cellular processes. For instance, the use of evanescent wave or Total Internal Reflection Fluorescent Microscopy (TIRFM) allows one to observe and analyze membrane and cytoskeleton dynamics in the first 100nm of the cell with incredible signal-to-noise - permitting visualization of fusion of single vesicles with the plasma membrane. Automated image processing and Monte Carlo simulations have revealed that these exocytic fusion events occur non-randomly in cellular hotspots. 4D multicolor imaging of membrane traffic will also be discussed.

Vince Calhoun, PhD
Independent Component Analysis of Functional MRI Data in the Complex Domain
Tues Jan 21 2003 4:00 pm (refreshments at 3:45 pm), MRRC Conference Room N135 CAB

Dr. Calhoun is an Assistant Clinical Professor in the Department of Psychiatry, Yale University School of Medicine and is the Director of Medical Image Analysis Laboratory at the Olin Neuropsychiatry Research Center, Institute of Living, Hartford.

Virtually all fMRI studies analyze only magnitude images even though the acquired data are complex image pairs. We recently proposed a synthesis/analysis model for fMRI data that can be used to optimize independent component analysis (ICA), a flexible modeling technique, and extended it to multiple subjects. In this talk, we discuss the extension of the model to processing of complex images. Using a flexible approach such as ICA is especially important in this case since it is unclear how to model the complex fMRI data. We will derive and apply several complex ICA approaches to fMRI data and discuss their relative advantages and disadvantages. We will also discuss certain challenges when using the phase information, such as increased sensitivity to certain artifacts. Through simulations and fMRI data we show that utilizing complex data appears to provide an empricial sensitivity increase. Additionally, the phase information may be useful for data interpretation and enable the separation of smaller and larger blood vessels. We will thus demonstrate that it makes both intuitive and empirical sense to develop methods to more fully utilize the available data in an fMRI experiment.

Fall 2002


Rikki Waterhouse, PhD
Development of PET Radiotracers for Glutamate Receptors
Tues Dec 17 2002 4:00 pm, CAB N135 MRRC Conference Room, 300 Cedar Street

Dr. Waterhouse is an Assistant Professor in the Departments of Neuroscience and Psychiatry and is also is a faculty in the Division of Biological Psychiatry at Columbia University.

The presentation will cover various aspects of the development of metabotropic and ionotropic glutamate receptor radiotracers for PET and SPECT, with a paricular focus on the NMDA ion channel, and group II mGluRs. The speaker will present her work in this area carried out over the past 3 years.

Arne Hengerer, PhD
Molecular biology for medical imaging
Mon, Dec 9, 4:00 PM CAB N135 MRRC 300 Cedar Street

Dr. Hengerer is presently Director for New Business Development in Molecular Imaging at Siemens Medical Solutions in Erlangen, Germany. He received his ph.D. in Molecular Biology, and his thesis focussed on recombinant proteins for biosensors.

Molecular imaging is currently in its early stages. Encouraging advances achieved in clinical PET and animal research indicates that this technique evolves into an indispensable diagnostic tool. When employed complementary to morphological imaging procedures, molecular imaging will result in a substantially improved diagnostic power. With molecular imaging earlier diagnosis of diseases compared to morphological imaging will come within reach, since anatomical structures always follow changes at the molecular level. Thereby it is possible to detect diseases much earlier, preferentially at their onset. Finally the focus from diagnostics might change from simple diagnosis to disease prediction and prevention. Consequently a conversion of the healthcare systems aimed at "sickness repair" to one focused on maintaining wellness might evolve. Nuclear imaging is the only clinically established molecular imaging modality so far. Among the emerging therapies, which will create further applications for medical imaging are targeted drug delivery, gene therapy and stem- and immuno-cell therapy. All these therapy schemes heavily depend on localized molecular information, i.e. on molecular imaging.

David A. Jaffray, Ph.D.
Flat-Panel Cone-Beam CT: An Adaptable Technology for Image-Guidance Applications
Wed, Dec 4, 4:00 PM CAB N135 MRRC CAB 300 Cedar Street

Dr. Jaffray is the Fidani Chair of Radiation Physics and Head of Radiation Therapy Physics at the Departments of Radiation Oncology and Medical Biophysics in University of Toronto.

Localized therapies are often compromised by the lack of peri-therapeutic imaging data. Additional imaging data acquired in the therapy setting allows intra-operative planning as well as monitoring and feedback. Many investigators have pursued the development of such imaging systems with varying degrees of success. Flat-panel cone-beam CT (FP-CBCT) promises to be a powerful and adaptable new imaging technology for applications that demand volumetric soft-tissue imaging in a variety of settings. FP-CBCT has been made viable through the development of large, robust 2D x-ray detectors (amorphous-silicon photodiode arrays) and high- speed reconstruction hardware for cone-beam CT. Investigations over the past three years have demonstrated the potential of this technology to generate high-resolution (sub-millimeter), low-dose (~0.5 cGy) volumetric ([25cm]3) CT images with soft-tissue contrast sensitivity comparable to conventional CT. Many of the physical challenges of cone-beam CT have been examined and the hypothesis that flat-panel cone-beam CT is a powerful technology for image-guided therapy remains strong. The real test of this technology is beginning, as application specific embodiments of the technology are being explored. At present, two image-guidance system are being constructed based upon this technology - a linear accelerator with on-board cone-beam CT, and an isocentric, mobile C-arm capable of intra-operative cone-beam CT. Clinical applications for this new technology will be presented, ranging from high-precision radiation therapy of the prostate to C-arm based image-guided brachytherapy and vertebroplasty.

Spring 2002


Donald G. Buerk, Ph.D.
Physiological measurements and modeling of nitric oxide biotransport in vivo
Thurs, Jan 10, 4:15 PM Hope 110 Jane Ellen Hope Building 315 Cedar Street

Dr. Buerk is a Professor in the Departments of Physiology, Bioengineering, and the Institute for Environmental Medicine at the University of Pennsylvania, Philadelphia, PA.

In vivo measurements of tissue oxygen partial pressure (pO2) and nitric oxide (NO) have been made with recessed electrochemical microsensors in the cerebral cortex and other tissues in rodents and cats. Blood flow was measured by laser Doppler flowmetry in most experiments. The role of NO in coupling blood flow with increased neuronal activity has been studied in the rat somatosensory cortex during electrical stimulation of the forepaw, and in the cat optic nerve during photic stimulation of the eye with flickering light. Hyperbaric O2 exposure has been studied in rats, and in knock-out mice lacking genes for neuronal NOS (nNOS) or endothelial NOS (eNOS). Increases in cortical tissue NO with hyperbaric O2 at 2.8 atmospheres absolute (ATA) were associated with increased nNOS activity. Vascular endothelial growth factor (VEGF) and growth of new blood vessels into collagen gels implanted over the cortex has been studied with NO microsensors. Mathematical models for NO biotransport including effects on O2 metabolism and O2 transport to tissue are under development. Although NO is a simple molecule, it has complex biological interactions that are intriguing to investigate and difficult to model. Tissue NO measurements using recessed NO microsensors provide insight into the complex behavior of NO in physiological systems.

Robin A. de Graaf, Ph.D.
Nuclear Spin Gymnastics. Optimization of In Vivo NMR Studies on Brain Energy Metabolism.
Tues, Jan 22, 4:00 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street

Dr. de Graaf is an Assistant Professor in the Department of Diagnostic Radiology and is an important member of the Section of Bioimaging Sciences as well as the Magnetic Resonance Center.

NMR is becoming a dominant tool to study anatomy, function, and metabolism non-invasively in vivo. MR imaging (MRI) and in particular functional MRI is ideally suited to study brain function in response to physiological stimulation, while MR spectroscopy allows the quantitative detection of important metabolic pathways, like the TCA cycle. Even though NMR is, in principle, a quantitative technique, the acquisition of reliable, robust and quantitative NMR data is not straightforward. Nuclear spin gymnastics, both theoretical and experimental, are at the heart of designing optimal NMR sequences. Proton-observed, carbon-edited NMR spectroscopy sequences will be used as a central example to illustrate the requirements to convert induced electrical current ("the NMR signal") to quantitative metabolic fluxes.

Mark D. Does, Ph.D.
Tissue characterization using sub-voxel MRI studies
Tues, Jan 29, 4:00 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street

Dr. Does is an Assistant Professor in the Department of Diagnostic Radiology and is an important member of the Section of Bioimaging Sciences.

One of the great attributes of magnetic resonance imaging (MRI) as a probe of biological samples is the array of different contrast mechanisms available (e.g., T1 and T2 relaxation, diffusion, flow, magnetization transfer, ?). Each unique preparation of magnetization can render image contrast that can distinguish specific anatomical or physiological characteristics. In this manner, and with contemporary MRI hardware, tissues are typically characterized with voxel dimensions on the order of about 1 mm3. However, in many tissues, complex NMR signal characteristics can distinguish water compartments that are several orders of magnitude smaller. For example, careful measurement of transverse relaxation (T2) in nerve and white matter reveals multiple relaxation rates, which are thought to be derived from unique micro- anatomical compartments: myelin, axoplasm, and extra-axonal space. Using such approaches with T2 and other contrast mechanisms, it should be possible to better understand MRI tissue characteristics on a sub-voxel level.

Graeme F. Mason, Ph.D.
MRS studies of neurotransmitter metabolism in depression and other applications: 13C metabolic modeling
Thurs, Feb 7, 4:00 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street
Dr. Mason is an Assistant Professor in the Department of Psychiatry and Biomedical Engineering Program.

1H magnetic resonance spectroscopy (MRS) is used routinely to measure the concentrations of amino acid neurotransmitters and other compounds in the brain, and concentration differences have been seen between resting control states and conditions of altered function, pharmacological treatment, and various disease states. To explore those differences, kinetic measurements can be made using MRS of the carbon isotope 13C. 13C is not radioactive but can be detected with MRS. 13C-labeled substrates are supplied for use by the brain, where they are converted to labeled products. The appearance of those products is detected over time. To evaluate those data qualitatively requires mathematical modeling of the metabolic pathways that lead to the product labeling. This presentation will cover basic procedures for simulating and fitting isotopic flows, with some applications to illustrate the method at work.

Rodolfo Llinas
"Imaging the functional substrate for cognition from single neuron to global brain function: Is it possible?"
Tues, Feb 12, 4:00 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street
Dr. Llinas is a Professor in the Departments of Physiology and Neuroscience at New York University School of Medicine.

Attempting to understand how the brain, as a whole, might be organized seems, for the first time, to be a serious topic of inquiry. One aspect of its neuronal organizational that is particularly central to global function is the rich thalamocortical interconnectivity, and most particularly the reciprocal nature of this enormous neuronal loop. Moreover, the interaction between the specific and non-specific thalamic loops suggests that, rather than a gate into the brain, the thalamus represents a hub from which any site in the cortex can communicate with any other such site or sites. The goal of the presentation is to explore the basic assumption that large-scale, temporal coincidence of specific and non- specific thalamic activity generates the functional states that characterize human cognition.

Peter Bandettini
"The hemodynamic response and more: Advances and prospects for fMRI"
Thurs, Mar 14, 3:00 PM Fitkin Amphitheatre (LMP 1094) Fitkin Building 330 Cedar Street

Dr. Bandettini is a Biophysicist at the Laboratory of Brain and Cognition (NIMH, NIH) and directs the research efforts at the Unit on Functional Imaging Methods as well as the Functional MRI Core Facility.

My research at the NIH has been essentially focused on developing fMRI methodology. This includes better understanding the origins of fMRI contrast, developing methods to extract more meaningful information at higher spatial and temporal resolution, working on specific applications based on what we have learned about fMRI contrast and contrast dynamics, and lastly, to look into other potential sources of fMRI contrast.
In this lecture, I will describe some of this ongoing work. Specifically, I will focus on our recent efforts to a) better understand and make use of the dynamics of blood oxygenation level contrast, b) better understand the time series fluctuations and how they relate to the "optimal resolution," and c) develop methods for extracting and mapping neuronal activity directly.

Fall 2001


Richard P. Kennan
Transcranial Near Infrared Spectroscopy of Brain Function
Thurs, Dec 6, 4 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street

Dr. Richard P. Kennan is an Assistant Professor in Diagnostic Radiology at Albert Einstein School of Medicine. His lab has a strong focus towards improving functional imaging methods (fMRI and NIRS).

It has long been known that light in the near infrared region is sensitive to changes in hemoglobin oxygenation state and concentration in living tissue. This effect has been exploited in the development of peripheral pulse oximetry and tissue oximeters. More recently, transcranial near infrared spectroscopy (referred to as optical topography) has been developed for visualizing brain function by mapping optical absorption changes on the cortical surface. Near infrared spectroscopy yields complementary information to other hemoglobin sensitive methods, such as fMRI, and can thus provide a more complete understanding of the underlying phsyiology. Furthermore, the facility of optical methods allows non-invasive measurement of human brain function under a variety of conditions without subject restriction . Data shall be presented on the MRI based validation of optical topography techniques in motor, auditory, and cognitive tasks as well as current work in progress towards monitoring tissue oxygenation in sickle cell disease.

Seong-Gi Kim
Limitation of fMRI resolution: How finely is CBF regulated?
Thurs, Dec 13, 4 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street

Dr. Seong-Gi Kim is an Associate Professor in Diagnostic Radiology at University of Minnesota, Minneapolis. His lab has a strong focus towards improving the temporal and spatial resolutions of MRI-based neuroimaging methods.

Functional MRI has been widely utilized for imaging brain functions. However, the extent of the fMRI hemodynamic response around active cortical columns remains poorly understood and controversial. Thus, we evaluated BOLD and CBF responses using a well-established cat orientation column model and a rat somatosensory model. As expected, conventional positive BOLD signals contained large draining vessels, and were not localized only to active neuronal active sites. Activation maps obtained using CBF fMRI were devoid of large draining vein contamination, and CBF responses were spatially localized to cortical columns/layers. These results suggest that hemodynamic-based fMRI can indeed be used to map high-resolution functional structures if large vessel contributions can be minimized.

Kamil Ugurbil
Functional neuroarchitecture investigated using nuclear spins
Tues, Dec 18, 4 PM Hope 103 Jane Ellen Hope Building 315 Cedar Street

Dr. Kamil Ugurbil is a Professor in the Department of Diagnostic Radiology at University of Minnesota, Minneapolis. He is also the Director of the Center for Magnetic Resonance Research at University of Minnesota, Minneapolis. His lab has played a pivotal role in the development and application of 13C MRS and 1H MRI-based neuroimaging methods towards functional imaging of the mammalian brain.

The history of Nuclear Magnetic Resonance is marked by ever increasing number of innovations that have produced novel and surprising uses of this phenomenon for probing biological processes. In the last decade, this approaches has been increasingly used for the acquisition of physiological and biochemical information non-invasively in intact systems including the human brain. One of the most notable accomplishments in this general effort has been the introduction of the magnetic resonance approaches to map brain function (fMRI). fMRI is based on the sensitivity of MR signals to secondary metabolic and hemodynamic responses that accompany increased neuronal activity. Despite this indirect link to neurotransmission, recent studies from our laboratory have demonstrated in animal models that under appropriate conditions, these fMRI maps have accuracy at the scale of submillimeter neuronal organizations such as the orientation columns of the visual cortex, and are directly proportional in magnitude to electrical signals, such as single unit spiking activity, generated by the neurons. High magnetic fields have been critical in achieving such specificity in functional maps because they provide advantages through increased signal-to-noise ratio (SNR), diminishing blood-related contributions to mapping signals, and enhanced sensitivity to microvasculature. These gains were recently utilized for the first time in human brain studies at 7 Tesla yielding information on functional parcellation in the millimeter scale in the fusiform face area (FFA) for object recognition and faces. Understanding brain function also requires information on connectivity among different brain regions. Techniques based on anisotropy of water diffusion have recently been pursued for this purpose. Limitations encountered by such techniques can be alleviated significantly due to the enhanced SNR available at high magnetic fields. Using this approach, we have recently generated simultaneous connectivity and activation maps in the cat brain at 9.4 Tesla. These recent high field human and animal brain studies will be presented.

Spring 2001


James S. Duncan, Ph.D.
"Model-Based Medical Image Analysis"
Wed, Feb 7, 4 PM, Brady Auditorium (B 131), Cedar Street, Yale University School of Medicine.

Dr. Duncan is a Professor in the Departments of Diagnostic Radiology and Electrical Engineering and will speak about the model-based approaches used in medical image analysis.

He will overview some of the recent work in his laboratory that is representative of the variety of ideas and efforts being pursued in the field of medical image analysis. The presentation will include a discussion of algorithm development in the areas of: deformable models for segmentation of anatomical structure, rigid and nonrigid image registration, approaches for structural and functional quantitative analysis, and physical/biomechanical modeling in image analysis. Some of the application areas for these approaches that will be overviewed include: the analysis of neuroanatomical structure, the study of the deformation of the left ventricle of the heart and image-guided intervention in neurosurgery and radiotherapy.

Douglas L. Rothman, Ph.D.
In vivo Magnetic Resonance Spectroscopy Studies of the Glutamate and GABA Neurotransmitter Cycles and Functional Neuroenergetics
Wed, Mar 14, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine.

Dr. Rothman is an Associate Professor in the Department of Diagnostic Radiology and the Director of Magnetic Resonance Center for Research in Metabolism and Physiology.

Recent developments in magnetic resonance spectroscopy ( MRS ) have wllowed the study neuronal glutamate and GABA metabolism and the relationship of amino acid metabolism to functional neuroenergetics. The brain pools of GABA, glutamate, and glutamine have been shown to be localized within glutamatergic neurons, GABAergic neurons, and glia respectively (under non pathological conditions). Under non fasting conditions glucose is the almost exclusive source of energy for the brain. By following the flow of 13C label from glucose into these metabolites MRS has been used to determine the separate rates of glucose oxidation in these cell types. The metabolism of glutamatergic neurons, GABAergic neurons and glia are coupled by neurotransmitter cycles. In the glutamate/glutamine cycle, glutamate released from nerve terminals (by either vesicular release or transport reversal) is transported into surrounding glial cells , and converted to glutamine. Glutamine in then transported out of the glia and transported in to the neurons, where it is converted back to glutamate thereby completing the cycle By following the flow of 13C label from glutamate into glutamine the rate of the glutamate/glutamine cycle may be determined using MRS. Through a similar strategy the GABA/glutamine cycle may be measured.
The application of MRS to study brain glutamate and GABA metabolism, and the coupling of neurotransmitter cycling to neuroenergetics , has provided several new, and controversial, insights into the relationship of brain metabolism and function. Contrary to the previous view of a separate metabolic and neurotransmitter pool of glutamate, glutamate release and recycling has been shown to be a major metabolic pathway. MRS studies of GABA metabolism in the rodent and human brain have suggested that there is also an important role of the metabolic pool of GABA in inhibitory function. Another key finding is that the glutamate/glutamine cycle in the cerebral cortex is coupled in a close to 1:1 ratio to neuronal (primarily glutamatergic) glucose oxidation above isoelectricity. This finding in combination with cellular studies has led to a model for the coupling between functional neuroenergetics and glutamate neurotransmission. The coupling between neurotransmission and neuroenergetics provides a linkage between the functional imaging signal and specific neuronal processes.

George I. Zubal
Nuclear Medicine Imaging of Human Brain Function
Wed, Mar 28, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Zubal is an Associate Professor in the Department of Diagnostic Radiology and hold numerous other appointments throughout the medical school; Chair of the Radioactive Drug Research Committee, Chair of the Radiation Safety Committee, Director of the SPECT Epilepsy Group, Senior Physicist in the NeuroSPECT Imaging Center.

Nuclear Medicine utilizes trace amounts of radioactive materials bound to metabolitically active molecules, which transport the radioisotope through functional pathways in the human organs. We can image these isotopes (using SPECT or PET) to create 3-dimensional representations of the organs' normal activity and/or its states of disease. This information is combined with the anatomy obtained from CT for MRI scans. The pharmacokinetics of the radiotracer 99m-Tc-HMPAO allow it to be used to image the brain blood flow at the time of injection. In epilepsy, we use SPECT for obtaining brain images immediately after the onset of a seizure. By registering the seizure and non-seizure SPECT data and applying a normalization and subtraction, we calculate a functional image is calculated demonstrating describing changes in brain perfusion at the time of the seizure. Using the radiotracer 18-FDG, the metabolism in the patient's brain can be measured using PET techniques. We apply image processing techniques to interictal PET and SPECT brain images to further aid in the localization of epileptogenic foci by calculating a functional image which represents the degree of un-coupling between perfusion and metabolism. Un-coupling of these two functions is characteristices of epileptogenic tissue in temporal lobe epilepsy and has the potential to serve as a diagnostic measure for localization in other areas as well. Similar to the SPECT perfusion analysis above, PET metabolism and perfusion images are spatially registered in three dimensions and a functional ratio-image is computed. These functional maps are overlayed onto a 3D rendering of the same patient's MRI anatomy. Such noninvasive imaging techniques help to minimize surgical intervention during the work-up and ultimate resection of tissue in curing epilepsy patients.

R. Todd Constable
Rapid imaging in the Presence of Magnetic Field Inhomogeneities: Problems and Solutions impacting Functional MRI
Wed, Apr 4, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Constable is an Associate Professor in the Departments of Diagnostic Radiology and Neurosurgery. He runs an active research program in the development of new MR imaging methods with applications to functional MR imaging of the brain and the heart.

Rapid imaging has many benefits the most important of which is the ability to rapidly produce images with intensities that reflect changes in local tissue blood oxygenation. The contrast mechanism exploited in most fMRI applications is the BOLD effect (Blood Oxygen Level Dependent contrast) wherein the image intensity in a gradient echo image is made sensitive to local changes in magnetic susceptibility which occur as diamagnetic oxygenated blood replaces paramagnetic deoxygenated blood upon tissue activation.
Rapid imaging allows either time-course studies of brain activation to be obtained (the so called event-related fMRI studies) or it allows the collection of many samples within a block design paradigm in order to produce robust statistical maps reflecting brain activation. The combination of rapid imaging and the need for sensitivity to local susceptibility effects (BOLD contrast) unfortunately also makes these sequences highly sensitive to static field inhomogeneities. The problems these field inhomogeneities cause and possible solutions to these problems will be discusssed, and specific examples of these issues in fMRI experiments described.

Steven W. Zucker
Computer Vision and Primate Vision
Wed, Apr 18, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Steven W. Zucker is the David and Lucile Packard Professor of Computer Science and Electrical Engineering at Yale University. Before moving to Yale in 1996, he was Professor of Electrical Engineering at McGill University, Director of the Program in Artificial Intelligence and Robotics of the Canadian Institute for Advanced Research, and the Co-Director of the Computer Vision and Robotics Laboratory in the McGill Research Center for Intelligent Machines. He was elected a Fellow of the Canadian Institute for Advanced Research, a Fellow of the IEEE, and (by) Fellow of Churchill College, Cambridge.

We consider the analysis of visual information by brains and by computers. A theory of computational vision, from edge detection to shape recognition, is sketched. The theory is geometrical, and at the early level is realized by interpreting edge elements as signaling a local approximation to tangents. These tangents are then grouped into global curves, which are classified by the Whitney theorem, and which support a description of shape as the singularities of a curve-evolution equation. A biologically-relevant theory of stereo reconstruction for space curves emerges from this formulation, which could be relevant for diagnostic applications. Other applications in computational neuroscience will be included.

Hal Blumenfeld
Network Inhibition Hypothesis for Loss of Consciousness During Seizures
Wed, May 2, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Blumenfeld is an Assistant Professor of Neurology and Neurobiology at Yale University School of Medicine

Epileptic seizures serve as a useful model system for exploring the difficult question of how consciousness occurs and is impaired through alterations in neuronal activity. Some epileptic seizures cause loss of consciousness, while others do not, allowing questions to be asked about the anatomical regions and physiological activity patterns that are important for consciousness. We have found through a convergence of data from functional neuroimaging and electrophysiological studies in both animal and human models that specific neuronal networks are disrupted during certain epileptic seizures, involving fronto-parietal association cortex, medial thalamus, and the pontomesencephalic reticular formation. These findings may have important functional consequences for both the behavioral manifestations of seizures, and normal brain function.

Anna W. Roe
Optical imaging and electrophysiological studies of sensory cortical function
Wed, May 9, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Anna W. Roe is an assistant professor in the Section of Neurobiology at Yale University. Her graduate work was done at M.I.T. and postdoctoral studies at the Rockefeller University and Baylor College of Medicine

In the cerebral cortex, each sensory modality (vision, audition, somatosensation) is represented in multiple cortical areas (e.g. over 30 visual areas). We would like to know why so many areas are required for the computation that allows us to see, hear, and touch. What does one area do that another does not? How do they work together? Previous studies have focussed on understanding the role of single cortical areas. The goal of our research is to understand how multiple areas work cooperatively at the scale of single functional domains (e.g. 50um scale). We are now examining the concurrent activation of pairs of closely related cortical areas during the processing of : 1) visual contours in visual areas V1 and V2, and 2) visual brightness in visual areas V1 and V2, and 3) tactile information in primary somatosensory cortex (Areas 3b and 1). In an effort to relate such cortical activations with behavior, we have developed a chronic optical chamber method that permits long-term study of cortical function in the awake, behaving animal. These studies have shown that each cortical area provides a specific (and uniquely abstracted) view of the visual world, each of which by itself is insufficient, but when considered together provides unique identification.

Lawrence B. Cohen
Voltage and Calcium imaging of brain activity
Wed, May 23, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Lawrence B. Cohen is a professor in Physiology at Yale University.

Optical measurements of brain activity can be divided into two types; measurements of intrinsic optical properties (intrinsic imaging) and measurements using extraneously added probes (of membrane potential or calcium concentration). This talk will focus on measurements of the second type.
These optical measurements have certain attractive features as well as certain severe limitations. Among the attractive features are the speed of the signal and the clear relationship to brain activity. Among the severe limitations are their invasiveness and the difficulty in measuring activity from structures deeper than 500-1000 microns from the surface.
Calcium imaging. A number of dyes are available whose absorption or fluorescence are sensitive to the concentration of calcium. Because calcium influx into a neuron is a frequent correlate of activity as well as an important regulator of the consequences of activity, measurement of the intracellular calcium concentration is a relatively direct measure of activity. We have used calcium signals to measure the input to the olfactory bulb in response to application of odorants to the nose.
Voltage-imaging. A different class of dyes have optical signals that follow membrane potential. We have used these dyes to follow the oscillations that occur in the olfactory bulb following application of odorants to the nose.

Robert G. Shulman
The Glycogen Shunt in Brain and Muscle
Wed, May 30, 4 PM Brady Auditorium (B 131) Cedar Street Yale University School of Medicine

Dr. Robert G. Shulman is the Sterling professor in MB&B at Yale University.

It is becoming clear that the tradtional picture of lactate generation, which depends upon a scarcity of oxygen, has no experimental support. In muscle it is well known that blood is far from depleted of oxygen when lactate is generated. In the brain the very existence of the BOLD functional imaging signal reflects an excess of oxygen during stimulation, rather than a depletion, and yet it is under these condtions that lactate is generated.
The explanation proposed then of lactate generation during stimulation is that in order to supply the power (time rate of energy consumption) needed during the millisecond neuronal stimulations in muscle and brain, glycogen is mobilized, via the activation of the fast enzyne, glycogen phosphorylase, the so callled fight or flight enzyme. This generaes more lactate than is needed by the slower oxidatiove processes so that the lactate concentration builds up and thereby stimulates the efflux. Lactate so generated need not be wasted, but as Brooks had shown can be transported to another site where it can be used efficiently. Hence lactate is not a sign of energy mismatch but of the temporal lack of synchrony between the millisecond needs for energy and the slower steady state supply. It is a time buffer.
October 2003