Computational neuroanatomy

• Julia Berzhanskaya, Krasnow Institute, George Mason University, Fairfax, VA

• Dr. Giorgio Ascoli, George Mason University, Fairfax, VA, USA

In the broadest sense, computational neuroanatomy is the application of computational techniques (e.g. analysis, visualization, modeling, and simulation) to the investigation of neural structure. Within the field of computational neuroscience, computational neuroanatomy is principally considered to aim at creating anatomically accurate models of the nervous system. At the cellular level, computational neuroanatomy is based on the quantitative description of the structure of individual neurons, the density of neuronal elements within specific brain areas, and their interconnectivity. An important goal in computational neuroanatomy is to relate the structure to the function, at both the single cell and network levels. Combining precise morphology and detailed biophysics of single cells within large-scale realistic and data-driven network models allows simulation of emergent properties of neural circuits. Development of databases of neuronal structures (e.g.NeuroMorpho.Org ) and biophysical models (e.g. SenseLab ModelDB) enables quantitative data transfer between laboratories and constitutes an essential area of scientific emphasis in contemporary computational neuroanatomy research.

Staining and imaging techniques for morphology and connectivity

For about a century the Golgi technique has been very successful in staining neurons enabling the semi-automatic reconstruction and the quantitative analysis of their neuronal branching patterns (e.g., Glaser and van der Loos, 65; Overdijk et al., 78). In combination with other classical staining methods it has been used to achieve a quantitative statistical description of brain tissue in terms of the density of neurons, synapses, and total length of axonal and dendritic arborizations (per cubic millimeter). For more details see the section on stereology and the book by Braitenberg and Schuez (1991). Further progress in computational neuroanatomy was enabled by the spread of modern imaging techniques that allow digital reconstructions of dendritic and axonal morpohology. In the last few decades, the development of intracellular labeling (Cullheim and Kellerth, 78, Snow et al., 76) using biocytin/HRP injections and various visualization methods, including HRP-DAB or fluorescent avidin reaction (Morozov et al., 02), has led to a large output of high resolution data about dendritic morphology. Tracer injections are typically performed in single neurons in vivo (anesthetized animal) or in vitro (slice conditions), and microscopy rendering often involves fixing and reslicing the tissue. New visualization approaches, such as GFP expression in specific cell types, allow direct morphology observation in vivo, both at the whole cell and subcellular levels (Oliva et al., 02, Trachtenberg et al., 02). In addition, cell-class specific GFP expression in conjunction with traditional in vitro microscopy reduces the sampling bias due to limited access to certain neuron types in vivo and variable cell survival during slice preparation. Finally, in vivo visualization enables observation of developmental and activity-dependent morphological changes, such as individual spine plasticity (Majewska et al., 06). McCormick (99) reconstructs neurons and their connections directly from stacks of thin sections, milled off from a block of tissue. Methods of automatic slice production, image capture and digitization play an important role in high-volume reconstructions (see Cryoplane Fluorescence Microscopy, Scope Controller and Neurolucida).

Quantitative analysis of dendritic and axonal morphology

Figure 1: Example of reconstructed neuron from NeuroMorpho.org.

A fundamental step for computational neuroanatomy is the tracing of the acquired neuronal images into a three-dimensional (3D) digital representation of the branching dendrites and/or axons (Figure 1). Traditionally performed manually, this process is very labor intensive severely limiting the number of available reconstructions. Both commercial (Neurolucida, MicroBrightField) and freeware software systems (Neuromantic,NeuroMorpho, or NeuronStudio) offer some level of automatization of the reconstruction process. These tools considerably simplify data acquisition thus contributing to the increase of the neuron morphology databases. In computer-aided neuromorphological reconstructions, the digital tracing files describe neuronal trees as a series of interconnected cylinders, each represented by diameter, spatial coordinates, and links to other cylinders in the arbor (Overdijk et al., 78, Uylings et al., 89, Ascoli et al., 01). The number of data points representing a single neuron can be quite large (see examples e.g. at NeuroMorpho.Org ).

Digital reconstructions enable quantitative analysis of neuronal shapes by means of morphometric parameters describing the metrical and topological properties and the spatial embedding of the three-dimensional structures (Uylings et al., 86, Hillman, 88, Burke et al, 92, Ascoli et al, 01, Uylings and Van Pelt, 02). These morphometric parameters make it possible to statistically describe the variability in neuronal morphologies. In more advanced approaches investigators have designed procedures for the computational generation of neuronal morphologies with realistic variabilities in their shapes based on intrinsic correlations between morphometric parameters (Hillman, 88, Burke et al., 92, Ascoli et al., 01) or based on neuronal growth principles (Van Pelt and Uylings, 07). Both the reconstruction and the growth-model approach are algorithmic procedures, using a limited set of statistical descriptors (basic parameters) or growth rules, respectively, to generate stochastic neuronal structures that are statistically indistinguishable from the real neurons of the same morphological class (e.g. Samsonovich and Ascoli, 05, Van Pelt and Schierwagen, 04). Freeware tools for the reconstruction approach are available to measure nearly unlimited structural features from digital reconstruction files (e.g., the L-Measure software program). The ability to characterize neurons with a compact code consisting of a few statistical descriptors represents a valuable compression tool of vast neuroanatomical data.

Potential synaptic connectivity between individual cells

The number of synapses between two given cells in a brain depends on multiple factors, including partitioning of 3D space by the dendrites of multiple neurons that share it, trajectories of the axons that navigate this 3D neuropil forest, and spatial distribution of chemical factors during axonal growth and neuronal activity. While the reconstruction of dendritic and axonal trees of selected pre- and postsynaptic cells is possible, in fixed adult tissue the neuronal activity pattern and the chemical makeup of the environment during development are obviously unknown. Computational neuroanatomists have developed both theoretical and numerical approaches to tackle this issue. One study suggested that intricate branching of axonal and dendritic trees are optimized to provide a high degree of connectivity between neurons, given the existing geometrical restrictions (Chklovskii, 04). In a different attempt to predict individual axon pathways through well defined structural landmarks, a system was built for simulation of growth and branching of thalamocortical projections on the basis of a voxel-based mouse brain atlas (Senft, 02).

The critical underlying question remains: how does the number of synapses between two neurons depend on the overlap between their dendritic and axonal trees? According to "Peters rule", synaptic contacts occur where dendrites and axons happen to be in apposition. These "potential synapses" are required but not sufficient for an actual synapse formation; and the expected number of connections between two neurons is proportional to the product of their dendritic and axonal trees densities (e.g. Peters et al., 91). However, a recent study of connectivity between two neurons using glutamate uncaging and dendritic/axonal reconstruction shows that there may be fewer functional synapses than geometrical appositions (Shepherd et al., 05). Furthermore, the ratio of the number of functional synapses to that of total possible synapses specifically depends on the laminar source of axonal projection as well as dendritic target (Shepherd et al., 05). In particular, in the somatosensory cortex, this ratio for synapses between layer 5 and layers 2/3 is higher than the ratio for the layer 4 to layer 2/3 projection (but both are significantly less than 1). Therefore, notwithstanding a baseline correlation between potential synaptic connectivity and axo-dendritic overlap, the set of functional synapses effectively expressed at any one time represents only one of a combinatorial multitude of possibilities. On the one hand, this may subserve a greater network capacity for representation diversity. On the other, there is the additional possibility to increase total synaptic strength by recruiting new synapses (for example, as a result of learning).

Structure/activity relationship: dendritic trees, input processing and firing patterns

The computational simulation of detailed brain activity at the cellular level requires quantitative knowledge of neuronal class specific distribution of ionic channel and other membrane properties, as well as synaptic connectivity between neurons. Digital reconstructions of neuronal morphology represent a central node for both of these requirements. Furthermore, they have been proven useful to investigate another crucial facet of neuronal activity, namely the effect of dendritic structure on cell function. Numerous modeling reports demonstrated that the neuronal firing patterns simulated with anatomically realistic compartmental representations of the neurons with a specified channel distribution heavily depend on morphology. Several known “cortical spiking patterns” were obtained by distributing the same membrane properties on cell classes such as pyramidal and stellate cells (Mainen and Sejnowski, 96), or only changing the topological asymmetry of the model dendrites (Van Ooyen et al., 02). Even more specifically, the whole range of diverse firing observed in hippocampal CA3 pyramidal cells was reproduced when implementing an identical biophysical model on a set of morphologies which reflected the naturally occurring variability (Krichmar et al., 02). These findings are particularly striking because they were obtained with simple stimulation corresponding to the experimental protocol of somatic current injection. A more natural activation of synaptic inputs, which are distributed on the dendrites, would be expected to exacerbate the influence of morphology on output patterns. Following these results, extensive efforts are ongoing to correlate dendritic structure and neuronal activity in various cell classes. At the same time, multiple studies indicate that channel distribution can change dendritic processing and neuronal firing properties (e.g. Migliore et al., 05, Poirazi and Mel, 01).

While a detailed overview of biophysical models is beyond the scope of this article, it is clear that static anatomical information may be necessary but not sufficient to predict individual neuron properties. At the very least, the study of dynamic functional properties of neurons in vivo is required. Recent attempts to integrate vast and diverse information on the anatomical, physiological, and molecular characteristics of neurons into a comprehensive, practical, and useful classification system led to the emerging consensus that dendritic morphology is only one of the factors that can aid specific neuronal identification (Somogyi and Klausberger, 05). Axonal projections, laminar distribution of neuronal inputs and outputs, and immunohistochemical markers are other prominent features, especially for interneurons. Noticeably, computational neuroanatomy techniques (and particularly digital reconstructions among them) are contributing substantially to foster progress in many if not all of these aspects.

Stereology, cell densities, and brain region volumes

Quantification of network connectivity presupposes a robust estimate of cell count. It is important to recognize that the number of neurons in a given brain region is not constant over time, but can change during development, as a consequence of neurodegenerative disease, or in the process of learning, as was suggested in particular for hippocampal areas such as the dentate gyrus (DG). To compute the convergence and/or divergence ratios between cell populations in a normal brain, or to test hypotheses about changes in a specific cell group count associated with a particular neurological disorder, total cell numbers and densities have to be calculated. Earlier anatomical studies counted Nissl-stained cells in two-dimensional (2D) slices. A switch to the 3D volume is not trivial. Stereology techniques developed in the last decades (Gundersen, 86, Baddeley et al., 86, West et al., 91, Guillery and Herrup, 97) provide unbiased estimates of such quantitative 3D parameters as the number of cells, volumes of brain areas, and cell density. Development of specialized software such as Stereoinvestigator and Neurolucida (MicroBrightField) combines modern stereological tools with a precise control of microscope lateral (in the plane of each slide) and depth displacements. These programs automate data collection from 2D slices and significantly speed up and simplify acquiring 3D estimates.

Recent application of stereological techniques together with new cell-specific imaging labels has led to considerable scientific achievements. These include demonstrating a quantitative link between aging in human brain and number of neocortical neurons (Pakkenberg and Gundersen, 98), changes in Alzheimer’s disease (Stark et al., 07), link between the drop in the number of cells in hippocampal areas CA1 and DG and cognitive decline in Alzheimer patients (West et al., 04), discovering that a significant proportion (6%) of new DG cells develop each month (Cameron and McKay, 01), finding that decreased amygdala cell numbers but not volumes are associated with autism (Shuman and Amaral, 06), alterations in neuronal numbers in Schizophrenia (Dorph-Petersen et al., 07; Stark et al., 04; Pakkenberg, 90), and no further doubling of number of neurons in Broca’s area after birth (Uylings et al, 05). An important aspect of testing hypotheses associating changes in brain volume with specific neurological disorders involves parcellation of image stacks (e.g. from MRI scans), into specific areas as a prerequisite for the reliable measurement of corresponding volumes. Significant inter-subject variability, especially in the gyral and sulcal patterns of the human cortex, corroborated the idea of probabilistic brain atlases (Thompson et al., 00). Automated techniques have been developed to aid extraction of cortical thickness and boundaries of specific areas (Fischl et al., 00, Fischl and Dale, 04). Moreover, computational methods of cortical inflation and flattening exist to switch to a surface-based coordinate system for a more accurate representation of cortical topography (Fischl et al., 99). An important component of computational neuroanatomy develops tools for quantitative comparisons between different brain atlases (see "The Brain Architecture Project")

Connectivity between brain areas and network topology

Synaptic connectivity data between pairs of individual cells can be summarized into cell class-specific statistics that will generally be location dependent within a brain area. Another important aspect of computational neuroanatomy involves the analysis of connectivity between different brain regions. Structural connectivity can be assessed by tracer injections in a given area of interest and the subsequent study of fixed tissue (Van Essen, 85). In addition, the recently developed structural MRI technique, Diffusion Tensor Imaging (Shimony et al., 04, Mori and Zhang, 06), can be used in vivo. Computational analysis of cortico-cortical connections pointed out the similarities with the small-world network architecture (Watts and Strogatz, 98) which has a short average path length and relatively high clustering (Braitenberg and Schuez, 91, Hilgetag et al., 00, Sporns et al., 00). Network modeling studies demonstrated that changes in the fundamental pattern of network connectivity can contribute to the manifestation of pathological population activity such as epilepsy (Netoff et al., 04, Dyhrfjeld-Johnsen et al., 07). Structural connectivity between areas reflects the topological proximity of different areas, but not the total number and strength of the synaptic connections (which, in turn, can be modified during development or in a task-dependent manner). To address that, new modeling techniques estimate functional connectivity between different areas based on the correlation of activity as detected by fMRI, EEG or MEG activation (Lee et al., 06, David et al., 06).

Development and activity-dependent plasticity of neural connectivity

Neuroanatomical studies in fixed tissue produce a snapshot of neuronal structure and connectivity at a given moment in time. In the living brain, this architecture changes as it develops, or as a result of learning. A key challenge is to characterize the mechanisms contributing to these plastic changes. Sophisticated imaging techniques have been recently established to assess the dynamics of dendritic spine turnover in the mature sensory cortex in vivo (Majewska et al., 06). Meanwhile, methods for observing axonal extension, branching and maturation of synaptic contacts, as a consequence of or even during manipulation of the experimental environment, are still being developed (Alsina et al., 01, Mumm et al., 06). Biophysical models of molecular processes underlying axonal and dendritic elongation and branching (Janulevicius et al., 06, Graham and van Ooyen, 06) give an insight into intracellular and extracellular determinants of neuronal morphology. Computational techniques that relate developmental changes in neural architecture to the functionality of mature neural circuits provide a useful tool and allow testing hypothesized metabolic or activity-dependent constraints.

Numerous computational models explain how columnar structure and feature representation develops in the visual cortex (Swindale, 96) based on cross-correlation and Hebbian learning (Miller et al., 89, Nakagama et al., 06), elastic self-organizing nets (Goodhill and Willshaw, 90) or efficient information coding (Olshausen and Field, 96). These models typically use simplified neural elements, generalized cell populations with uniform properties, and connections with average weights from one cell class to another. The fate of an individual axon distributing synaptic contacts across a complex dendritic tree during development or experience-dependent plasticity remains to be fully explored. A separate class of models addresses effects of chemoattractors (Goodhill and Xu, 05), and homeostatic plasticity (Rabinovich and Segev, 06) on neural circuit organization and functionality.

References

• Alsina B,Vu T,Cohen-Cory S. Visualizing synapse formation in arborizing optic axons in vivo: dynamics and modulation by BDNF. Nat Neurosci. 2001; 4(11):1093-101.
• Ascoli GA, Krichmar JL, Nasuto SJ, Senft SL. Generation, description and storage of dendritic morphology data. Philos Trans R Soc Lond B Biol Sci. 2001; 356(1412):1131-45.
• Baddeley AJ, Gundersen HJG, Cruz-Orive LM. Estimation of surface area from vertical sections. J Microsc. 1986; 142:259-276.
• Braienberg and Schuez. Cortex: Statistics and Geometry of Neuronal Connectivity. Springer, Berlin, New York. 1st edition 1991, 2nd edition 1998.
• Burke RE, Marks WB, Ulfhake B. A parsimonious description of motoneuron dendritic morphology using computer simulation. J Neurosci. 1992; 12(6):2403-16.
• Cameron HA, McKay RD. Adult neurogenesis produces a large pool of new granule cells in the dentate gyrus. J Comp Neurol. 2001; 435(4):406-17.
• Chklovskii DB. Synaptic connectivity and neuronal morphology: two sides of the same coin. Neuron. 2004;43(5):609-17.
• Cullheim S, Kellerth JO. A morphological study of the axons and recurrent axon collaterals of cat sciatic alpha-motoneurons after intracellular staining with horseradish peroxidase. J Comp Neurol. 1978;178(3):537-57.
• David O, Kiebel SJ, Harrison LM, Mattout J, Kilner JM, Friston KJ. Dynamic causal modeling of evoked responses in EEG and MEG. Neuroimage. 2006; 30(4):1255-72.
• Dorph-Petersen, K-A, Pierri JN, Wu Q, Sampson AR, Lewis DA. Primary visual cortex volume and total neuron number are reduced in schizophrenia. J Comp Neurol. 2007; 501:290–301.
• Dyhrfjeld-Johnsen J, Santhakumar V, Morgan RJ, Huerta R, Tsimring L, Soltesz I. Topological determinants of epileptogenesis in large-scale structural and functional models of the dentate gyrus derived from experimental data. J Neurophysiol. 2007;97(2):1566-87.
• Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A. 2000; 97(20):11050-5.
• Fischl B, Sereno MI, Dale AM. Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system. Neuroimage. 1999; 9(2):195-207.
• Fischl B,van der Kouwe A, Destrieux C,Halgren E,Segonne F,Salat DH,Busa E,Seidman LJ, Goldstein J, Kennedy D, Caviness V, Makris N, Rosen B, Dale AM.Automatically parcellating the human cerebral cortex. Cereb Cortex. 2004;14(1):11-22.
• Glaser EM, van der Loos H. A semiautomatic computer microscope for the analysis of neuronal morphology. IEEE Trans Biomed Eng. 1965; 12:22-31.
• Goodhill GJ, Willshaw DJ. Application of the elastic net algorithm to the formation of ocular dominance stripes. Network Comput Neural Syst 1990; 1:41–59.
• Goodhill GJ, Xu J. The development of retinotectal maps: a review of models based on molecular gradients. Network. 2005;16(1):5-34.
• Graham BP, van Ooyen A. Mathematical modelling and numerical simulation of the morphological development of neurons.BMC Neurosci. 2006; 7 Suppl 1:S9
• Guillery RW, Herrup K. Quantification without pontification: choosing a method for counting objects in sectioned tissues. J Comp Neurol. 1997; 386(1):2-7.
• Gundersen HJG. Stereology of arbitrary particles. A review of unbiased number and size estimators and the presentation of some new ones, in memory of William R. Thompson. J Microsc. 1986; 143:3–45.
• Hilgetag CC, Grant S. Uniformity, specificity and variability of corticocortical connectivity. Philos Trans R Soc Lond B Biol Sci. 2000; 355(1393):7-20.
• Hillman DE. Parameters of dendritic shape and substructure: intrinsic and extrinsic determination? in Lasek RJ, Black MM, ed. Intrinsic determinants of neuronal form and function. New York, Alan R. Liss, 1988, p.83-113.
• Janulevicius A, van Pelt J, van Ooyen A. Compartment volume influences microtubule dynamic instability: a model study. Biophys J. 2006;90(3):788-98.
• Krichmar JL, Nasuto SJ, Scorcioni R, Washington SD, Ascoli GA. Effects of dendritic morphology on CA3 pyramidal cell electrophysiology: a simulation study. Brain Res. 2002; 941(1-2):11-28.
• Lee L, Friston K, Horwitz B. Large-scale neural models and dynamic causal modelling. Neuroimage. 2006; 30(4):1243-54.
• Mainen ZF, Sejnowski TJ. Influence of dendritic structure on firing pattern in model neocortical neurons. Nature. 1996; 382(6589):363-6.
• Majewska AK, Newton JR, Sur M. Remodeling of synaptic structure in sensory cortical areas in vivo. J Neurosci. 2006; 26(11):3021-9.
• McCormick BH. Design of a brain tissue scanner. Neurocomputing. 1999; 26-27:1025–1032.
• Migliore M, Ferrante M, Ascoli GA. Signal propagation in oblique dendrites of CA1 pyramidal cells. J Neurophysiol. 2005; 94(6):4145-55.
• Miller K, Keller J and Stryker M. Ocular dominance column development: analysis and simulation. Science. 1989; 245(4918): 605 - 615.
• Mori S, Zhang J. Principles of diffusion tensor imaging and its applications to basic neuroscience research. Neuron. 2006; 51(5):527-39.
• Morozov Y, Khalilov I, Ben-Ari Y, Represa A. Correlative fluorescence and electron microscopy of biocytin-filled neurons with a preservation of the postsynaptic ultrastructure. J Neurosci Methods. 2002; 117(1):81-5.
• Mumm JS, Williams PR, Godinho L, Koerber A, Pittman AJ, Roeser T, Chien CB, Baier H, Wong RO. In vivo imaging reveals dendritic targeting of laminated afferents by zebrafish retinal ganglion cells. Neuron. 2006; 52(4):609-21.
• Nakagama H, Tani T, Tanaka S. Theoretical and experimental studies of relationship between pinwheel centers and ocular dominance columns in the visual cortex. Neurosci Res. 2006; 55(4):370-82.
• Netoff TI, Clewley R, Arno S, Keck T, White JA. Epilepsy in small-world networks. J Neurosci. 2004; 24(37):8075-83.
• Oliva AA Jr, Lam TT, Swann JW. Distally directed dendrotoxicity induced by kainic acid in hippocampal interneurons of green fluorescent protein-expressing transgenic mice. J Neurosci. 2002; 22(18):8052-62.
• Olshausen BA and Field DJ. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 1996; 381(6583):607-9.
• Overdijk J, Uylings HBM, Kuypers K, Kamstra AW. An economical, semi-automatic system for measuring cellular tree structures in three dimensions, with special emphasis on Golgi-impregnated neurons. J Microsc. 1978; 114:271–284.
• Pakkenberg B. Pronounced reduction of total neuron number in mediodorsal thalamic nucleus and nucleus accumbens in schizophrenics. Arch Gen Psychiatry. 1990; 47(11):1023-8
• Pakkenberg B. Total nerve cell number in neocortex in chronic schizophrenics and controls estimated using optical disectors. Biol Psychiatry. 1993; 34(11):768-72.
• Pakkenberg B, Gundersen HJ. Neocortical neuron number in humans: effect of sex and age. J Comp Neurol. 1997; 384(2):312-20.
• Peters A, Palay SL and Webster DeF. The Fine Structure of the Nervous System 3rd ed 1991 (Oxford: Oxford University Press)
• Poirazi P, Mel BW. Impact of active dendrites and structural plasticity on the memory capacity of neural tissue. Neuron. 2001; 29(3):779-96.
• Rabinowitch I, Segev I. The endurance and selectivity of spatial patterns of long-term potentiation/depression in dendrites under homeostatic synaptic plasticity. J Neurosci. 2006; 26(52):13474-84.
• Samsonovich AV, Ascoli GA. Statistical determinants of dendritic morphology in hippocampal pyramidal neurons: A hidden Markov model. Hippocampus. 2005; 15(2):166-83.
• Schumann CM, Amaral DG. Stereological analysis of amygdala neuron number in autism. J Neurosci. 2006;26(29):7674-9.
• Senft SL. Axonal navigation through voxel substrates: a strategy for reconstructing a brain circuitry. in Ascoli GA, ed, Computational Neuroanatomy, 2002, Humana Press, pp 245-70.
• Shepherd GM, Stepanyants A, Bureau I, Chklovskii D, Svoboda K. Geometric and functional organization of cortical circuits. Nat Neurosci. 2005; 8(6):782-90
• Shimony JS, Snyder AZ, Conturo TE, Corbetta M. The study of neural connectivity using diffusion tensor tracking. Cortex. 2004; 40(1):213-5.
• Slaughterbeck C, Datta S, Smith S, Ritchie D, Wohnoutka P, High Throughput Automated Microscopy Platform for the Allen Brain Atlas, JALA 2007; 12 (6)
• Snow PJ, Rose PK, Brown AG. Tracing axons and axon collaterals of spinal neurons using intracellular injection of horseradish peroxidase. Science. 1976;191(4224):312-3
• Somogyi P, Klausberger T. Defined types of cortical interneurone structure space and spike timing in the hippocampus. J Physiol. 2005; 562(Pt 1):9-26.
• Sporns O, Tononi G, Edelman GM. Theoretical neuroanatomy: relating anatomical and functional connectivity in graphs and cortical connection matrices. Cereb Cortex. 2000; 10(2):127-41.
• Stark AK, Pelvig DP, Jørgensen AM, Andersen BB, Pakkenberg B. Measuring morphological and cellular changes in Alzheimer's dementia: a review emphasizing stereology. Curr Alzheimer Res. 2005; 2(4):449-81.
• Stark AK, Uylings HB, Sanz-Arigita E, Pakkenberg B. Glial cell loss in the anterior cingulate cortex, a subregion of the prefrontal cortex, in subjects with schizophrenia. Am J Psychiatry. 2004; 161(5):882-8.
• Swindale NV. The development of topography in the visual cortex: a review of models. Network Comput Neural Syst. 1996; 7:161–247.
• Thompson P., Woods RP, Mega MS, Toga AW. Mathematical/computational challenges in creating deformable and probabilistic atlases of the human brain. Hum Brain Mapp. 2000; 9(2):81-92.
• Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR, Welker E, Svoboda K. Long-term in vivo imaging of experience-dependent synaptic plasticity in adult cortex. Nature. 2002; 420(6917):788-94.
• Uylings HBM, Malofeeva LI, Bogolepova IN, Jacobsen AM, Amunts K and Zilles K. No postnatal doubling of number of neurons in human Broca's area (BA 44 and 45)?: a stereological study. Neuroscience. 2005; 136:715- 728.
• Uylings HBM, Ruiz-Marcos A, van Pelt J. The metric analysis of three-dimensional dendritic tree patterns: a methodological review. J Neurosci Meth. 1986; 18:127-151.
• Uylings HBM, van Pelt J. Measures for quantifying dendritic arborizations. Network: Comput Neural Syst. 2002; 13:397-414.
• Van Essen DC. Functional organization of primate visual cortex. In: Cerebral Cortex, Vol.3. (Peters A and Rockland K editors ) 1985; Plenum Press, New York.
• Van Ooyen A, Duijnhouwer J, Remme MWH, van Pelt J. The effect of dendritic topology on firing patterns in model neurons. Network: Comput Neural Syst. 2002; 13: 311-325.
• Van Pelt J, Schierwagen A. Morphological analysis and modeling of neuronal dendrites. Math Biosc. 2004; 188:147-155.
• Van Pelt, J and Uylings HBM. Natural variability in the geometry of dendritic branching patterns. In: Modeling in the Neurosciences: From Ionic Channels to Neural Networks (Poznanski RP, Ed.) 1999; Harwood Academic Publishers, Amsterdam, pp.79-108.
• Van Pelt J, Uylings HBM. Modeling neuronal growth and shape. In Modeling Biology - Structures, Behaviors, Evolution; Manfred D. Laubichler and Gerd B. Müller (eds). 2007; The MIT Press, Cambridge, pp. 195-215.
• Watts DJ, Strogatz SH. Collective dynamics of 'small-world' networks. Nature. 1998; 393(6684):440-2.
• West MJ, Kawas CH, Stewart WF, Rudow GL, Troncoso JC. Hippocampal neurons in pre-clinical Alzheimer's disease. Neurobiol Aging. 2004;25(9):1205-12.
• West MJ, Slomianka L, Gundersen HJ. Unbiased stereological estimation of the total number of neurons in thesubdivisions of the rat hippocampus using the optical fractionator. Anat Rec. 1991; 231(4):482-97.
• Wittner L, Henze DA, Záborszky L, Buzsáki G. Three-dimensional reconstruction of the axon arbor of a CA3 pyramidal cell recorded and filled in vivo. Brain Struct Funct. 2007; 212(1):75-83.

Internal references

• Joseph E. LeDoux (2008) Amygdala. Scholarpedia, 3(4):2698.

• Valentino Braitenberg (2007) Brain. Scholarpedia, 2(11):2918.

• Paul L. Nunez and Ramesh Srinivasan (2007) Electroencephalogram. Scholarpedia, 2(2):1348.

• Seiji Ogawa and Yul-Wan Sung (2007) Functional magnetic resonance imaging. Scholarpedia, 2(10):3105.

• Almut Schüz (2008) Neuroanatomy. Scholarpedia, 3(3):3158.

• Marco M Picchioni and Robin Murray (2008) Schizophrenia. Scholarpedia, 3(4):4132.

Recommended reading

• Ascoli, GA. (Editor) Computational Neuroanatomy: Principles and Methods. Springer. 2002.
• Bjaaliehttp, JG. (Editor) Computational neuroanatomy: Tools for single neuron, system, and brain region analysis. Special Issue, Anatomy and Embryology. 2001; 204, 4.
• Singer E. A Wiring Diagram of the Brain. Technology Review. 2007.

External links

• Julia Berzhanskaya's website
• Giorgio Ascoli's website

See also

Brain, Connectome, Neuroanatomy, Neuroimaging

Sponsored by: Eugene M. Izhikevich, Editor-in-Chief of Scholarpedia, the peer-reviewed open-access encyclopedia

Reviewed by: Anonymous

Reviewed by: Dr. Jaap van Pelt, CNCR,VU University Amsterdam, The Netherland

Accepted on: 2008-03-09 19:40:43 GMT