neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 1 of 15 NEUROINFORMATICS – REVIEW ARTICLES1: Cell Mol Biol (Noisy-le-grand). 2007 Jan 20;52(6):16-23.The increasing influence of medical image processing in clinical neuroimaging.Barillot C.CNRS VisAGeS U746 Unit/Project, IRISA UMR 6074 Campus de Beaulieu, Rennes France.Christian.Barillot@irisa.fr This paper review the evolution of clinical neuroinformatics domain in the passed andgives an outlook how this research field will evolve in clinical neurology (e.g. Epilepsy,Multiple Sclerosis, Dementia) and neurosurgery (e.g. image guided surgery, intra-operativeimaging, the definition of the Operation Room of the future). These different issues, asaddressed by the VisAGeS research team, are discussed in more details and the benefits ofa close collaboration between clinical scientists (radiologist, neurologist and neurosurgeon)and computer scientists are shown to give adequate answers to the series of problems whichneeds to be solved for a more effective use of medical images in clinical neurosciences.2: Neuroinformatics. 2007 Winter; 5(1):79-94.Web-based method for translating neurodevelopment from laboratory species tohumans.Clancy B, Kersh B, Hyde J, Darlington RB, Anand KJ, Finlay BL.Department of Biology, University of Central Arkansas, Conway, Arkansas; Dept of Pediatrics,Neurobiology & Developmental Sciences, University of Arkansas for Medical Sciences, ArkansasChildrens Hospital Research Institute, Little Rock, Arkansas. Biomedical researchers and medical professionals are regularly required to compare avast quantity of neurodevelopmental literature obtained from an assortment of mammalswhose brains grow at diverse rates, including fast developing experimental rodent speciesand slower developing humans. In this article, we introduce a database-driven website,which was created to address this problem using statistical-based algorithms to integratehundreds of empirically derived developing neural events in 10 mammalian species(http://translatingtime.net/). The site, based on a statistical model that has evolved over thepast decade, currently incorporates 102 different neurodevelopmental events obtained from10 species: hamsters, mice, rats, rabbits, spiny mice, guinea pigs, ferrets, cats, rhesusmonkeys, and humans. Data are arranged in a Structured Query Language database, whichallows comparative brain development measured in postconception days to be convertedand accessed in real time, using Hypertext Preprocessor language. Algorithms applied to thedatabase also allow predictions for dates of specific neurodevelopmental events whereempirical data are not available, including for the human embryo and fetus. By designing aweb-based portal, we seek to make these comparative data readily available to all those whoneed to efficiently estimate the timing of neurodevelopmental events in the human fetus,laboratory species, or across several different species. In an effort to further refine andexpand the applicability of this database, we include a mechanism to submit additional data.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 2 of 153: Neuroinformatics. 2007 Winter;5(1):35-58.Toward a workbench for rodent brain image data: systems architecture and design.Moene IA, Subramaniam S, Darin D, Leergaard TB, Bjaalie JG.Neural Systems and Graphics Computing Laboratory, Centre for Molecular Biology and Neuroscience& Institute of Basic Medical Sciences, University of Oslo, P.O. Box 1105 Blindern, N-0317 Oslo,Norway. We present a novel system for storing and manipulating microscopic images from sectionsthrough the brain and higher-level data extracted from such images. The system is designedand built on a three-tier paradigm and provides the research community with a web-basedinterface for facile use in neuroscience research. The Oracle relational databasemanagement system provides the ability to store a variety of objects relevant to the imagesand provides the framework for complex querying of data stored in the system. Further, thesuite of applications intimately tied into the infrastructure in the application layer provide theuser the ability not only to query and visualize the data, but also to perform analysisoperations based on the tools embedded into the system. The presentation layer uses extantprotocols of the modern web browser and this provides ease of use of the system. Thepresent release, named Functional Anatomy of the Cerebro-Cerebellar System (FACCS),available through The Rodent Brain Workbench (http:// rbwb.org/), is targeted at thefunctional anatomy of the cerebro-cerebellar system in rats, and holds axonal tracing datafrom these projections. The system is extensible to other circuits and projections and to othercategories of image data and provides a unique environment for analysis of rodent brainmaps in the context of anatomical data. The FACCS application assumes standard animalbrain atlas models and can be extended to future models. The system is available both forinteractive use from a remote web-browser client as well as for download to a local servermachine.4: Neuroinformatics. 2007 Winter; 5(1):1-2.A second look back.De Schutter E.Theoretical Neurobiology, University of Antwerp, Belgium. This issue opens the fifth volume of Neuroinformatics, which is a good time to look at howthe journal is doing, as it has evolved quite a bit as I wrote a similar editorial for the secondvolume (De Schutter, 2004). What has not changed is that we are very proud about oureditorial work. Our impact factor is excellent for a journal with a strong emphasis oninformatics and methods, we started at 3.0 for 2004 and are now at 3.9. This puts us headsand shoulders above all computational neuroscience, machine learning, and neurosciencemethods journals. We rank in the top-half of neuroscience journals, better than many classicneuroscience titles, and do even better in informatics in which we are ranked fourth ininterdisciplinary computer science. This high impact factor is supported by two trends, apositive and a negative one. Rather negative is that we publish relatively few articles, in fact,the third and fourth volumes contained a quarter less articles than the first two. This helps ofcourse with the impact factor but also reflects a rather low article submission rate. We expectthat the good impact factor will help to solve this problem but will also make sure that ahigher influx of manuscripts will not lead to a lowering of the quality of the journal.Nevertheless, this volume will still include only four issues; the increase to six volumes hasbeen postponed till we get a permanent increase in article submission.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 3 of 155: Neurotoxicology. 2007 Feb 15; [Epub ahead of print]Extrapolating brain development from experimental species to humans.Clancy B, Finlay BL, Darlington RB, Anand KJ.University of Central Arkansas, AR, United States; University of Arkansas for Medical Sciences LittleRock, AR, United States. To better understand the neurotoxic effects of diverse hazards on the developing humannervous system, researchers and clinicians rely on data collected from a number of modelspecies that develop and mature at varying rates. We review the methods commonly used toextrapolate the timing of brain development from experimental mammalian species tohumans, including morphological comparisons, "rules of thumb" and "event-based" analyses.Most are unavoidably limited in range or detail, many are necessarily restricted to rat/humancomparisons, and few can identify brain regions that develop at different rates. We suggestthis issue is best addressed using "neuroinformatics", an analysis that combinesneuroscience, evolutionary science, statistical modeling and computer science. A current useof this approach relates numeric values assigned to 10 mammalian species and hundreds ofempirically derived developing neural events, including specific evolutionary advances inprimates. The result is an accessible, online resource (http://www.translatingtime.net/) thatcan be used to equate dates in the neurodevelopmental literature across laboratory speciesto humans, predict neurodevelopmental events for which data are lacking in humans, andhelp to develop clinically relevant experimental models.6: Methods Inf Med. 2007;46(2):142-6.Neural engineering--a new discipline for analyzing and interacting with the nervoussystem.Durand DM.Neural Engineering Center, Wickenden 112, Department of Biomedical Engineering, Case WesternReserve University, Cleveland, OH 44106, USA. firstname.lastname@example.org OBJECTIVES: The field of neural engineering focuses on an area of research at theinterface between neuroscience and engineering. The area of neural engineering was firstassociated with the brain machine interface but is much broader and encompassesexperimental, computational, and theoretical aspects of neural interfacing, neuroelectronics,neuromechanical systems, neuroinformatics, neuroimaging, neural prostheses, artificial andbiological neural circuits, neural control, neural tissue regeneration, neural signal processing,neural modelling and neuro-computation. One of the goals of neural engineering is todevelop a selective interface for the peripheral nervous system. METHODS: Nerve cuffselectrodes have been developed to either reshape or maintain the nerve into an elongatedshape in order to increase the circumference to cross sectional ratio. It is then possible toplace many electrodes around the nerve to achieve selectivity. This new cuff (flat interfacenerve electrode: FINE) was applied to the hypoglossal nerve and the sciatic nerve in dogsand cats to estimate the selectivity of the interface. RESULTS: By placing many contactsclose to the axons, three different types of selectivity were achieved: 1) The FINE couldgenerate a high degree of stimulation selectivity as estimated by the individual fasciclerecording. 2) Similarly, recording selectivity was also demonstrated and blind sourcealgorithms were applied to recover the signals. 3) Finally, by placing arrays of electrodesalong the nerve, small fiber diameters could be excited before large fibers thereby reversingthe recruitment order. CONCLUSION: Taking advantage of the fact that nerves are not round
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 4 of 15but oblong or flat allows a novel design for selective nerve interface with the peripheralnervous system. This new design has found applications in many disorders of the nervoussystem such as bladder incontinence, obstructive sleep apnea and stroke.7: Neuroinformatics. 2006 Winter;4(4):319-20.Brain maps and connectivity representation.Vercelli A.Department of Anatomy, Pharmacology, and Forensic Medicine, University of Torino, Corso M.DAzeglio 52, 10126 Torino, Italy. email@example.com: Neuroinformatics. 2006 Winter;4(4):299-317.Neuroanatomical affiliation visualization-interface system.Palombi O, Shin JW, Watson C, Paxinos G.POWMRI, The University of New South Wales, Randwick NSW, Australia. firstname.lastname@example.org A number of knowledge management systems have been developed to allow users tohave access to large quantity of neuroanatomical data. The advent of three-dimensional (3D)visualization techniques allows users to interact with complex 3D object. In order to betterunderstand the structural and functional organization of the brain, we presentNeuroanatomical Affiliations Visualization-Interface System (NAVIS) as the original softwareto see brain structures and neuroanatomical affiliations in 3D. This version of NAVIS hasmade use of the fifth edition of "The Rat Brain in Stereotaxic coordinates" (Paxinos andWatson, 2005). The NAVIS development environment was based on the scripting languagename Python, using visualization toolkit (VTK) as 3D-library and wxPython for the graphicuser interface. The following manuscript is focused on the nucleus of the solitary tract (Sol)and the set of affiliated structures in the brain to illustrate the functionality of NAVIS. Thenucleus of the Sol is the primary relay center of visceral and taste information, and consistsof 14 distinct subnuclei that differ in cytoarchitecture, chemoarchitecture, connections, andfunction. In the present study, neuroanatomical projection data of the rat Sol were collectedfrom selected literature in PubMed since 1975. Forty-nine identified projection data of Solwere inserted in NAVIS. The standard XML format used as an input for affiliation data allowsNAVIS to update data online and/or allows users to manually change or update affiliationdata. NAVIS can be extended to nuclei other than Sol.9: Neuroinformatics. 2006 Winter;4(4):275-98.A new module for on-line manipulation and display of molecular information in thebrain architecture management system.Bota M, Swanson LW.The Neuroscience Research Institute, University of Southern California, Los Angeles, California90089-2520, USA. email@example.com A new "Molecules" module of the Brain Architecture Management System (BAMS;http://brancusi.usc.edu/bkms) is described. With this module, BAMS becomes the first onlineknowledge management system to handle central nervous system (CNS) region and celltypechemoarchitectonic data in the context of axonal connections between regions and celltypes, in multiple species. The "Molecules" module implements a general knowledgerepresentation schema for data and metadata collated from published and unpublishedmaterial, and allows insertion of complex reports about the presence of molecules collated
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 5 of 15from the literature. For different CNS neural regions and cell types, the modules databasestructure includes representation of molecule expression revealed by various techniquesincluding in situ hybridization and immunohistochemistry, molecule coexpression and time-dependent level changes, and physiological state of subjects. The metadata representationallows online comparison and evaluation of inserted experiments, and "Molecules"structureallows rapid development of data transfer protocols enabling neuroinformatics visualizationtools to display gene expression patterns residing in BAMS, in terms of levels of expressedmolecules and in situ hybridization data. The modules web interface allows users toconstruct lists of CNS regions containing a molecule (depending on physiological state),retrieve further details about inserted records, compare time-dependent data within andacross experiments, reconstruct gene expression patterns, and construct complex reportsfrom individual experiments.10: Neuroinformatics. 2006 Winter;4(4):271-3.Wheres the beef? Missing data in the information age.Kennedy D.11: Neuroinformatics. 2006 Summer;4(3):213-6.The ups and downs of neuroscience shares.Ascoli GA.12: Neuroinformatics. 2006;4(2):199-212.A general XML schema and SPM toolbox for storage of neuro-imaging results andanatomical labels.Keator DB, Gadde S, Grethe JS, Taylor DV, Potkin SG; FIRST BIRN.University of California, Irvine, CA. With the increased frequency of multisite, large-scale collaborative neuro-imaging studies,the need for a general, self-documenting framework for the storage and retrieval of activationmaps and anatomical labels becomes evident. To address this need, we have developedand extensible markup language (XML) schema and associated tools for the storage ofneuro-imaging activation maps and anatomical labels. This schema, as part of the XML-based Clinical Experiment Data Exchange (XCEDE) schema, provides storage capabilitiesfor analysis annotations, activation threshold parameters, and cluster and voxel-levelstatistics. Activation parameters contain information describing the threshold, degrees offreedom, FWHM smoothness, search volumes, voxel sizes, expected voxels per cluster, andexpected number of clusters in the statistical map. Cluster and voxel statistics can be storedalong with the coordinates, threshold, and anatomical label information. Multiple thresholdtypes can be documented for a given cluster or voxel along with the uncorrected andcorrected probability values. Multiple atlases can be used to generate anatomical labels andstored for each significant voxel or cluter. Additionally, a toolbox for Statistical ParametricMapping software (http://www. fil. ion.ucl.ac.uk/spm/) was created to capture the results fromactivation maps using the XML schema that supports both SPM99 and SPM2 versions(http://nbirn.net/Resources/Users/ Applications/xcede/SPM_XMLTools.htm). Support foranatomical labeling is available via the Talairach Daemon (http://ric.uthscsa.edu/projects/talairachdaemon.html) and Automated Anatomical Labeling (http://www.cyceron.fr/freeware/).
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 6 of 1513: Neuroinformatics. 2006;4(2):139-62.NeuroScholars electronic laboratory notebook and its application toneuroendocrinology.Khan AM, Hahn JD, Cheng WC, Watts AG, Burns GA.Neuroscience Research Institute, Department of Biological Sciences, 3641 Watt Way, HedcoNeurosciences Building, University of Southern California, Los Angeles, CA 90089-2520, USA. Scientists continually relate information from the published literature to their currentresearch. The challenge of this essential and time-consuming activity increases as the bodyof scientific literature continues to grow. In an attempt to lessen the challenge, we havedeveloped an Electronic Laboratory Notebook (ELN) application. Our ELN functions as acomponent of another application we have developed, an open-source knowledgemanagement system for the neuroscientific literature called NeuroScholar (http://www.neuroscholar. org/). Scanned notebook pages, images, and data files are entered into theELN, where they can be annotated, organized, and linked to similarly annotated excerptsfrom the published literature within Neuroscholar. Associations between these knowledgeconstructs are created within a dynamic node-and-edge user interface. To produce aninteractive, adaptable knowledge base. We demonstrate the ELNs utility by using it toorganize data and literature related to our studies of the neuroendocrine hypothalamicparaventricular nucleus (PVH). We also discuss how the ELN could be applied to modelother neuroendocrine systems; as an example we look at the role of PVH stressor-responsive neurons in the context of their involvement in the suppression of reproductivefunction. We present this application to the community as open-source software and invitecontributions to its development.14: Neuroinformatics. 2006;4(2):131-8.A view of the digital landscape for neuroscience at NIH.Huerta MF, Liu Y, Glanzman DL.15: Neuroinformatics. 2006;4(2):129-30.On the future of the human brain project.De Schutter E, Ascoli GA, Kennedy DN.16: Proc Natl Acad Sci U S A. 2006 Jul 11;103(28):10775-80. Epub 2006 Jun 30.Arithmetic processing in the brain shaped by cultures.Tang Y, Zhang W, Chen K, Feng S, Ji Y, Shen J, Reiman EM, Liu Y.Institute of Neuroinformatics and Laboratory for Brain and Mind, Dalian University of Technology,Dalian 116023, China. The universal use of Arabic numbers in mathematics raises a question whether thesedigits are processed the same way in people speaking various languages, such as Chineseand English, which reflect differences in Eastern and Western cultures. Using functional MRI,we demonstrated a differential cortical representation of numbers between native Chineseand English speakers. Contrasting to native English speakers, who largely employ a
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 7 of 15language process that relies on the left perisylvian cortices for mental calculation such as asimple addition task, native Chinese speakers, instead, engage a visuo-premotor associationnetwork for the same task. Whereas in both groups the inferior parietal cortex was activatedby a task for numerical quantity comparison, functional MRI connectivity analyses revealed afunctional distinction between Chinese and English groups among the brain networksinvolved in the task. Our results further indicate that the different biological encoding ofnumbers may be shaped by visual reading experience during language acquisition and othercultural factors such as mathematics learning strategies and education systems, whichcannot be explained completely by the differences in languages per se.17: Neuroinformatics. 2006 Winter;4(1):51-64.Imaging genomics applied to anxiety, stress response, and resiliency.Xu K, Ernst M, Goldman D.Laboratory of Neurogenetics, National Institute on Alcohol Abuse and Alcohollism, National Institute ofHealth, Rockville, MD 20852, USA. firstname.lastname@example.org Anxiety and stress response/resiliency are heritable traits central to the etiology of multiplepsychiatric diseases, but efforts to identify genetic variation influencing this broad domain ofneurobiological function are hampered by the coarseness of the phenotypic measures andthe effects of environmental factors. Neuroimaging offers a powerful approach for assessingfunctional neuronal activity. Neurophysiological measures can serve as intermediatephenotypes more directly linked to small gene effects, compared with behavioral end pointsof neural dysfunction. Imaging genomics is a relatively new research area that is concernedwith linking functional gene variants and brain information processing. Here, we will focus onprocesses affected by anxiety and stress. Neuroimaging has been combined with geneticanalysis to reveal genetic effects of functional variants of the serotonin transporter (5-HTT)and catechol-O-methyltransferase (COMT) genes on brain response to stressful stimuli. Thelow-expressing allele of the 5-HTT promoter polymorphism (HTTLPR) is associated withanxiety and with greater amygdala and other regional responses to emotional. The COMTMet158 allele leads to lower COMT activity and has also been associated with anxiety, andthe effect of this gene is apparently additive with HTTLPR. Individuals with Met158genotypes are more sensitive to pain stress and, as shown by C11 Carfentanil imaging, havediminished ability to upregulate opioid release after pain/stress. These results suggest thatfunctional variants of 5-HTT and COMT impact brain functions involved in stress and anxiety.18: J Affect Disord. 2006 May;92(1):133-8. Epub 2006 Feb 20.Neuroinformatics: a new tool for studying the brain.Bloom FE, Morrison JH, Young WG.Neurome Inc., 11149 North Torrey Pines Road, La Jolla, CA 92037, USA. email@example.com BACKGROUND: Central nervous system diseases constitute a major target for drugdevelopment. Genes expressed by the nervous system may represent half or more of themammalian genome, with literally tens of thousands of gene products. METHODS: Bettermethods are therefore required to accelerate the pace of mapping gene expression patternsin the mouse brain and to evaluate the progressive phenotypic changes in genetic models ofhuman brain diseases. CONCLUSIONS: Recent studies of mouse models of AmyotrophicLateral Sclerosis and Alzheimers disease illustrate how such data could be used for drugdevelopment. Since these two diseases-- especially Alzheimers Disease-- entail disordered
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 8 of 15behavior, cognition and emotions, the framework and the methodology described in thisarticle might in the future find applications in research on affective disorders.19: Science. 2006 Jan 13;311(5758):176.Continuing progress in neuroinformatics.Gazzaniga MS, Van Horn JD, Bloom F, Shepherd GM, Raichle M, Jones E.20: Neuroinformatics. 2005;3(4):315-8.Dangerous phase.Schiff SJ.Krasnow Institute, George Mason University, Fairfax, VA, USA. firstname.lastname@example.org "Use a quiet reference." How many times have we heard this mantra during training orpractice, interpreting electroencephalogram (EEG) tracings, or implanting intracranialelectrodes? How many of us have used common reference EEG for synchrony studies inrecent years? Far too many. Perhaps one source of this problem is the number 104. This isthe relatively small number of citations to the reference Fein et al. (1988), which should haveput to rest any further use of referential EEG for coherence measurements. And inretrospect, a more careful reading by us of Nunezs (1981) text would have instructed us notto do this. How such warnings have managed to escape integration into common knowledgeand practice is troublesome. Electrical potentials are all measured with respect to otherpotentials. Technically, a potential difference is calculated by integrating the electrical fieldover a given path from one place to another in EEG terms, we mea sure a potential withrespect to another potential, measured at one or more electrodes. All EEG potentialmeasurements reflect the paths used to measure those potentials, and do not directly reflectlocalized regions of the brain beneath one electrode. Worse, in scalp EEG, the layers ofcerebrospinal fluid, dura, skull, and scalp serve to smooth, filter, spread out, and redirectcurrents generated within the brain so that the measured scalp potentials bear a rathertenuous relationship to the underlying (presumably dipole) current sources. In calculatingcoherence, it is easy to show that if the potential differences are all made with respect to acommon reference, then the amplitude of the reference can dominate the coherenceestimate (Fein et al., 1988). In recent years, phase synchronization has been increasinglyapplied to analyze the dynamics of nonlinear systems (Pikovsky et al., 2000). In Guevara etal. (in this issue), we see the extension of Feins results for phase coherency. The geometryof Fig. 1 in Guevara et al. should be imprinted on all of us the amplitude of a commonreference can dominate the calculated phase synchronization. There is far too muchliterature within the past decade that calculated phase synchronization from commonreferenced EEG.The good news is that the fix to remove common reference artifacts issimple. The bad news is that the interpretation of reference- free synchronization results frombrain signals requires considerable caution.21: Neuroinformatics. 2005;3(4):287-92.The impact of neuroinformatics.Kennedy DN.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 9 of 1522: Neurosurg Focus. 2005 Oct 15;19(4):E4.Bioinformatics and functional magnetic resonance imaging in clinical populations:practical aspects of data collection, analysis, interpretation, and management.Vincent DJ, Hurd MW.Department of Radiology, Medical University of South Carolina, Charleston, South Carolina 29425,USA. email@example.com In this paper the authors review the issues associated with bioinformatics and functionalmagnetic resonance (fMR) imaging in the context of neurosurgery. They discuss the practicalaspects of data collection, analysis, interpretation, and the management of large data sets,and they consider the challenges involved in the adoption of fMR imaging into clinicalneurosurgical practice. Their goal is to provide neurosurgeons and other clinicians with abetter understanding of some of the current issues associated with bioinformatics orneuroinformatics and fMR imaging. Thousands to tens of thousands of images are typicallyacquired during an fMR imaging session. It is essential to follow an activation task paradigmexactly to obtain an accurate representation of cortical activation. These images are theninteractively postprocessed offline to produce an activation map, or in some cases a series ofmaps. The maps may then be viewed and interpreted in consultation with a neurosurgeonand/or other clinicians. After this consultation, long-term archiving of the processed fMRactivation maps along with the standard structural MR images is a complex but necessaryfinal step in this process. The fMR modality represents a valuable tool in the neurosurgicalplanning process that is still in the developmental stages for routine clinical use, but holdsexceptional promise for patient care.23: Neuroinformatics. 2005;3(3):243-62.Methods for quantifying the informational structure of sensory and motor data.Lungarella M, Pegors T, Bulwinkle D, Sporns O.Department of Mechano-Informatics, School of Information Science and Technology, University ofTokyo, 113-8656 Tokyo, Japan. Embodied agents (organisms and robots) are situated in specific environments sampledby their sensors and within which they carry out motor activity. Their control architectures ornervous systems attend to and process streams of sensory stimulation, and ultimatelygenerate sequences of motor actions, which in turn affect the selection of information. Thus,sensory input and motor activity are continuously and dynamically coupled with thesurrounding environment. In this article, we propose that the ability of embodied agents toactively structure their sensory input and to generate statistical regularities represents amajor functional rationale for the dynamic coupling between sensory and motor systems.Statistical regularities in the multimodal sensory data relayed to the brain are critical forenabling appropriate developmental processes, perceptual categorization, adaptation, andlearning. To characterize the informational structure of sensory and motor data, we introduceand illustrate a set of univariate and multivariate statistical measures (available in anaccompanying Matlab toolbox). We show how such measures can be used to quantify theinformation structure in sensory and motor channels of a robot capable of saliency-basedattentional behavior, and discuss their potential importance for understanding sensorimotorcoordination in organisms and for robot design.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 10 of 1524: Neuroinformatics. 2005;3(2):163-6.A new era in computational neuroscience.Blackwell KT.School of Computational Sciences, and the Krasnow Institute of Advanced Studies, Rockfish CreekLane, MS 2A1 George Mason University, Fairfax, VA 22030, USA. firstname.lastname@example.org: Neuroinformatics. 2005;3(2):115-31.Comparison of vector space model methodologies to reconcile cross-speciesneuroanatomical concepts.Srinivas PR, Wei SH, Cristianini N, Jones EG, Gorin FA.Center for Neuroscience, UC Davis, Davis, CA, USA. Generating informational thesauri that classify, cross-reference, and retrieve diverse andhighly detailed neuroscientific information requires identifying related neuroanatomical termsand acronyms within and between species (Gorin et al., 2001) Manual construction of suchinformational thesauri is laborious, and we describe implementing and evaluating aneuroanatomical term and acronym reconciliation (NTAR) system to assist domain expertswith this task. NTAR is composed of two modules. The neuroanatomical term extraction(NTE) module employs a hidden Markov model (HMM) in conjunction with lexical rules toextract neuroanatomical terms (NT) and acronyms (NA) from textual material. The output ofthe NTE is formatted into collections of term- or acronym-indexed documents composed ofsentences and word phrases extracted from textual material. The second informationretrieval (IR) module utilizes a vector space model (VSM) and includes a novel, automatedrelevance feedback algorithm. The IR module retrieves statistically related neuroanatomicalterms and acronyms in response to queried neuroanatomical terms and acronyms.Neuroanatomical terms and acronyms retrieval obtained from term-based inquiries werecompared with (1) term retrieval obtained by including automated relevance feedback andwith (2) term retrieval using "document-to-document" comparisons (context-based VSM). Theretrieval of synonymous and similar primate and macaque thalamic terms and acronyms inresponse to a query list of human thalamic terminology by these three IR approaches wascompared against a previously published, manually constructed concordance table ofhomologous cross-species terms and acronyms. Term-based VSM with automatedrelevance feedback retrieved 70% and 80% of these primate and macaque terms andacronyms, respectively, listed in the concordance table. Automated feedback algorithmcorrectly identified 87% of the macaque terms and acronyms that were independentlyselected by a domain expert as being appropriate for manual relevance feedback. Context-based VSM correctly retrieved 97% and 98% of the primate and macaque terms andacronyms listed in the term homology table. These results indicate that the NTAR systemcould assist neuroscientists with thesauri creation for closely related, highly detailedneuroanatomical domains.26: Genet Mol Res. 2004 Dec 30;3(4):564-74.Bioinformatics: perspectives for the future.Costa Lda F.Cybernetic Vision Research Group, Institute of Physics at São Carlos, University of São Paulo, CaixaPostal 369, 13560-970 São Carlos, SP, Brazil. email@example.com
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 11 of 15 I give here a very personal perspective of Bioinformatics and its future, starting bydiscussing the origin of the term (and area) of bioinformatics and proceeding by trying toforesee the development of related issues, including pattern recognition/data mining, theneed to reintegrate biology, the potential of complex networks as a powerful and flexibleframework for bioinformatics and the interplay between bio- and neuroinformatics. Humanresource formation and market perspective are also addressed. Given the complexity andvastness of these issues and concepts, as well as the limited size of a scientific article andfinite patience of the reader, these perspectives are surely incomplete and biased. However,it is expected that some of the questions and trends that are identified will motivatediscussions during the IcoBiCoBi round table (with the same name as this article) andperhaps provide a more ample perspective among the participants of that conference andthe readers of this text.27: J Biomed Inform. 2004 Oct;37(5):380-91.Multivariate image analysis in biomedicine.Nattkemper TW.Applied Neuroinformatics Group, Faculty of Technology, Bielefeld University, P.O. Box 100131,D-33501 Bielefeld, Germany. firstname.lastname@example.org In recent years, multivariate imaging techniques are developed and applied in biomedicalresearch in an increasing degree. In research projects and in clinical studies as well m-dimensional multivariate images (MVI) are recorded and stored to databases for asubsequent analysis. The complexity of the m-dimensional data and the growing number ofhigh throughput applications call for new strategies for the application of image processingand data mining to support the direct interactive analysis by human experts. This articleprovides an overview of proposed approaches for MVI analysis in biomedicine. Aftersummarizing the biomedical MVI techniques the two level framework for MVI analysis isillustrated. Following this framework, the state-of-the-art solutions from the fields of imageprocessing and data mining are reviewed and discussed. Motivations for MVI data mining inbiology and medicine are characterized, followed by an overview of graphical and auditoryapproaches for interactive data exploration. The paper concludes with summarizing openproblems in MVI analysis and remarks upon the future development of biomedical MVIanalysis.28: Neuroinformatics. 2004;2(3):271-4.A gateway to the future of neuroinformatics.Gardner D, Shepherd GM.Laboratory of Neuroinformatics, Department of Physiology and Biophysics, Weill Medical College ofCornell University, NY, USA. email@example.com: Neuroinformatics. 2004;2(2):145-62.The small world of the cerebral cortex.Sporns O, Zwi JD.Department of Psychology, Indiana University, Bloomington 47405, USA. firstname.lastname@example.org While much information is available on the structural connectivity of the cerebral cortex,especially in the primate, the main organizational principles of the connection patterns linkingbrain areas, columns and individual cells have remained elusive. We attempt to characterize
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 12 of 15a wide variety of cortical connectivity data sets using a specific set of graph theory methods.We measure global aspects of cortical graphs including the abundance of small structuralmotifs such as cycles, the degree of local clustering of connections and the average pathlength. We examine large-scale cortical connection matrices obtained from neuroanatomicaldata bases, as well as probabilistic connection matrices at the level of small cortical neuronalpopulations linked by intra-areal and inter-areal connections. All cortical connection matricesexamined in this study exhibit "small-world" attributes, characterized by the presence ofabundant clustering of connections combined with short average distances betweenneuronal elements. We discuss the significance of these universal organizational features ofcortex in light of functional brain anatomy. Supplementary materials are at www.indiana.edu/~cortex/lab.htm.30: Neuroinformatics. 2004;2(2):127-44.Online retrieval, processing, and visualization of primate connectivity data from theCoCoMac database.Kötter R.C & O Vogt Brain Research Institute, Heinrich Heine University Düsseldorf, Moorenstr. 5, D-40225,Germany. Connectivity is the key to understanding distributed and cooperative brain functions.Detailed and comprehensive data on large-scale connectivity between primate brain areashave been collated systematically from published reports of experimental tracing studies.Although the majority of the data have been made easily available for online retrieval, themultiplicity of brain maps and the precise requirements of anatomical naming limit theintuitive access to the data. The quality of data retrieval can be improved by observing asmall set of conventions in data representation. Standardized interfaces open up furtheropportunities for automated search and retrieval, for flexible visualization of data, and forinteroperability with other databases. This article provides a discussion and examples in textand image of the capabilities of the online interface to the CoCoMac database of primateconnectivity. These serve to point out sources of potential confusion and failure, and todemonstrate the automated interfacing with other neuroinformatics resources that facilitateselection and processing of connectivity data, for example, for computational modelling andinterpretation of functional imaging studies.31: Annu Rev Neurosci. 2004;27:419-51.Neuronal circuits of the neocortex.Douglas RJ, Martin KA.Institute of Neuroinformatics, University/ETH Zurich, Zurich 8057, Switzerland. email@example.com We explore the extent to which neocortical circuits generalize, i.e., to what extent canneocortical neurons and the circuits they form be considered as canonical? We find that, ashas long been suspected by cortical neuroanatomists, the same basic laminar and tangentialorganization of the excitatory neurons of the neocortex is evident wherever it has beensought. Similarly, the inhibitory neurons show characteristic morphology and patterns ofconnections throughout the neocortex. We offer a simple model of cortical processing that isconsistent with the major features of cortical circuits: The superficial layer neurons withinlocal patches of cortex, and within areas, cooperate to explore all possible interpretations ofdifferent cortical input and cooperatively select an interpretation consistent with their variouscortical and subcortical inputs.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 13 of 1532: Nat Neurosci. 2004 May;7(5):467-72.E-neuroscience: challenges and triumphs in integrating distributed data frommolecules to brains.Martone ME, Gupta A, Ellisman MH.Department of Neurosciences, National Center for Microscopy and Imaging Research and The Centerfor Research in Biological Systems, The University of California San Diego, La Jolla, California92093-0608, USA. Imaging, from magnetic resonance imaging (MRI) to localization of specificmacromolecules by microscopies, has been one of the driving forces behindneuroinformatics efforts of the past decade. Many web-accessible resources have beencreated, ranging from simple data collections to highly structured databases. Although manychallenges remain in adapting neuroscience to the new electronic forum envisioned byneuroinformatics proponents, these efforts have succeeded in formalizing the requirementsfor effective data sharing and data integration across multiple sources. In this perspective,we discuss the importance of spatial systems and ontologies for proper modeling ofneuroscience data and their use in a large-scale data integration effort, the BiomedicalInformatics Research Network (BIRN).33: Neuroinformatics. 2003;1(1):81-109.Tools and approaches for the construction of knowledge models from theneuroscientific literature.Burns GA, Khan AM, Ghandeharizadeh S, ONeill MA, Chen YS.K-Mechanics Research Group, 3641 Watt Way, Hedco Neuroscience Building, University of SouthernCalifornia, Los Angeles, CA 90089-2520, USA. firstname.lastname@example.org Within this paper, we describe a neuroinformatics project (called "NeuroScholar,"http://www.neuroscholar.org/) that enables researchers to examine, manage, manipulate,and use the information contained within the published neuroscientific literature. The projectis built within a multi-level, multi-component framework constructed with the use of softwareengineering methods that themselves provide code-building functionality forneuroinformaticians. We describe the different software layers of the system. First, wepresent a hypothetical usage scenario illustrating how NeuroScholar permits users toaddress large-scale questions in a way that would otherwise be impossible. We do this byapplying NeuroScholar to a "real-world" neuroscience question: How is stress-relatedinformation processed in the brain? We then explain how the overall design of NeuroScholarenables the system to work and illustrate different components of the user interface. We thendescribe the knowledge management strategy we use to store interpretations. Finally, wedescribe the software engineering framework we have devised (called the "View-Primitive-Data Model framework," [VPDMf]) to provide an open-source, accelerated softwaredevelopment environment for the project. We believe that NeuroScholar will be useful toexperimental neuroscientists by helping them interact with the primary neuroscientificliterature in a meaningful way, and to neuroinformaticians by providing them with useful,affordable software engineering tools.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 14 of 1534: Neuroinformatics. 2003;1(1):65-80.A percolation approach to neural morphometry and connectivity.Costa Lda F, Manoel ET.Cybernetic Vision Research Group, IFSC-USP, Caixa Postal 369, 13560-970, São Carlos, SP, Brazil.email@example.com This article addresses the issues of neural shape characterization and analysis from theperspective of one of the main roles played by neural shapes, namely, connectivity. Thisstudy is oriented toward the geometry at the individual cell level and involves the use of thepercolation concept from statistical mechanics, which is reviewed in an accessible fashion.The characterization of the neural cell geometry with respect to connectivity is performed interms of critical percolation probability obtained experimentally while considering severaltypes of geometrical interactions between cells, therefore directly expressing the potential forconnections defined by each situation. Two basic situations are considered: dendrite-dendrite and dendrite-axon interactions. The obtained results corroborate the potential of thecritical percolation probability as a valuable resource for characterizing, classifying, andanalyzing the morphology of neural cells.35: Neuroinformatics. 2003;1(1):1-2.An information science infrastructure for neuroscience.Ascoli GA, De Schutter E, Kennedy DN.36: Neuroinformatics. 2003;1(2):149-65.Neuroscience data and tool sharing: a legal and policy framework forneuroinformatics.Eckersley P, Egan GF, Amari S, Beltrame F, Bennett R, Bjaalie JG, Dalkara T, De SchutterE, Gonzalez C, Grillner S, Herz A, Hoffmann KP, Jaaskelainen IP, Koslow SH, Lee SY,Matthiessen L, Miller PL, da Silva FM, Novak M, Ravindranath V, Ritz R, Ruotsalainen U,Subramaniam S, Toga AW, Usui S, van Pelt J, Verschure P, Willshaw D, Wrobel A, Tang Y;OECD Working Group on Neuroinformatics.Department of Computer Science & Software Engineering, Intellectual Property Research Institute ofAustralia, The University of Melbourne. firstname.lastname@example.org The requirements for neuroinformatics to make a significant impact on neuroscience arenot simply technical--the hardware, software, and protocols for collaborative research--theyalso include the legal and policy frameworks within which projects operate. This is not leastbecause the creation of large collaborative scientific databases amplifies the complicatedinteractions between proprietary, for-profit R&D and public "open science." In this paper, wedraw on experiences from the field of genomics to examine some of the likely consequencesof these interactions in neuroscience. Facilitating the widespread sharing of data and toolsfor neuroscientific research will accelerate the development of neuroinformatics. We proposeapproaches to overcome the cultural and legal barriers that have slowed these developmentsto date. We also draw on legal strategies employed by the Free Software community, insuggesting frameworks neuroinformatics might adopt to reinforce the role of public-sciencedatabases, and propose a mechanism for identifying and allowing "open science" uses fordata whilst still permitting flexible licensing for secondary commercial research.37: Neuroinformatics. 2003;1(2):145-7.From data to knowledge.Ascoli GA.
neuroinformaticsref1docdocdoc4105.doc 12/1/10 Page 15 of 1538: Neuroinformatics. 2003;1(4):397-410.The informatics of a C57BL/6J mouse brain atlas.MacKenzie-Graham A, Jones ES, Shattuck DW, Dinov ID, Bota M, Toga AW.Laboratory of Neuro Imaging, Department of Neurology, University of California, Los Angeles, CA. The Mouse Atlas Project (MAP) aims to produce a framework for organizing and analyzingthe large volumes of neuroscientific data produced by the proliferation of genetically modifiedanimals. Atlases provide an invaluable aid in understanding the impact of geneticmanipulations by providing a standard for comparison. We use a digital atlas as the hub ofan informatics network, correlating imaging data, such as structural imaging and histology,with text-based data, such as nomenclature, connections, and references. We generatedbrain volumes using magnetic resonance microscopy (MRM), classical histology, andimmunohistochemistry, and registered them into a common and defined coordinate system.Specially designed viewers were developed in order to visualize multiple datasetssimultaneously and to coordinate between textual and image data. Researchers cannavigate through the brain interchangeably, in either a text-based or image-basedrepresentation that automatically updates information as they move. The atlas also allowsthe independent entry of other types of data, the facile retrieval of information, and thestraight-forward display of images. In conjunction with centralized servers, image and textdata can be kept current and can decrease the burden on individual researchers computers.A comprehensive framework that encompasses many forms of information in the context ofanatomic imaging holds tremendous promise for producing new insights. The atlas andassociated tools can be found at http://www.loni.ucla.edu/MAP.39: Neuroinformatics. 2003;1(4):379-95.The cell-centered database: a database for multiscale structural and proteinlocalization data from light and electron microscopy.Martone ME, Zhang S, Gupta A, Qian X, He H, Price DL, Wong M, Santini S, Ellisman MH.Department of Neurosciences, University of California at San Diego, San Diego, CA, USA.email@example.com The creation of structured shared data repositories for molecular data in the form of web-accessible databases like GenBank has been a driving force behind the genomic revolution.These resources serve not only to organize and manage molecular data being created byresearchers around the globe, but also provide the starting point for data mining operationsto uncover interesting information present in the large amount of sequence and structuraldata. To realize the full impact of the genomic and proteomic efforts of the last decade,similar resources are needed for structural and biochemical complexity in biological systemsbeyond the molecular level, where proteins and macromolecular complexes are situatedwithin their cellular and tissue environments. In this review, we discuss our efforts in thedevelopment of neuroinformatics resources for managing and mining cell level imaging dataderived from light and electron microscopy. We describe the main features of our web-accessible database, the Cell Centered Database (CCDB; http://ncmir.ucsd.edu/CCDB/),designed for structural and protein localization information at scales ranging from largeexpanses of tissue to cellular microdomains with their associated macromolecularconstituents. The CCDB was created to make 3D microscopic imaging data available to thescientific community and to serve as a resource for investigating structural andmacromolecular complexity of cells and tissues, particularly in the rodent nervous system.