Electron microscopy; Light microscopy; Computer vision; Histology; Image processing; Connectome; Brain wiring; Synapse; Circuit; Neural network; Brain; Network; segmentation; algorithm
Bopp R Macarico Da Costa N Kampa BM Martin KAC Roth MM (2014), Pyramidal cells make specific connections onto smooth (GABAergic) cells in mouse visual cortex., in PLoS Biology
, 12(8), 1.
Dmitry Laptev Alexander Vezhnevets Sarvesh Dwivedi Joachim M. Buhmann (2012), Anisotropic ssTEM Image Segmentation Using Dense Correspondence across Sections, in Medical Image Computing and Computer-Assisted Intervention – MICCAI 2012
, Nice, FranceSpringer Publishing, New York, US.
Funke J. Andres B. Hamprecht F. Cardona A. and Cook M. (2012), Efficient Automatic 3D-Reconstruction of Branching Neurons from EM Data, in Computer Vision and Pattern Recognition
, Piscataway, New Jersey, USInstitute of Electrical and Electronics Engineers, New Jersey, US.
Saalfeld S. Fetter R. Cardona A. and Tomancak P. (2012), Elastic volume reconstruction from series of ultra-thin microscopy sections, in Nature Methods
, 9, 717-720.
E. Serradell F. Moreno J. Kybic and P. Fua (2012), Robust Non-Rigid Registration of 2D and 3D Graphs, in Computer Vision and Pattern Recognition
, Providence, RI, USAInstitute of Electrical and Electronics Engineers, New Jersey, US.
Mostany R Anstey JE Crump KL Maco B Knott G and Portera-Cailliau C, Altered synaptic dynamics during normal brain aging, in Journal of Neuroscience
Maco B Holtmaat A Cantoni M Kreshuk A Straehle CN Hamprecht F Knott GW, Correlative in vivo 2 Photon and Focused Ion Beam Scanning Electron Microscopy of Cortical Neurons, in PLoSONE
N. M. Costa, Diversity of thalamorecipient spine morphology in cat visual cortex and its implication for synaptic plasticity, in Journal of Comparative Neurology
Miyasaka N Wanner AA Li J Mack-Bucher J Genoud C Yoshihara Y Friedrich RW, Functional development of the olfactory system in zebrafish, in Mech Dev
Laptev Dmitry, Vezhnevets Alexander, Buhmann Joachim, SuperSlicing frame restoration for anisotropic ssTEM, in IEEE 11th International Symposium on Biomedical Imaging ISBI 2014
, BeijingMicrotome Publishing, -.
The history of natural sciences has taught us that a new discipline emerges whenever the length scale of investigation is reduced by about three orders of magnitude. The recent development of 3-D electron microscopy (3DEM) techniques induces such a transition by allowing the ultrastructural analysis of biological tissues throughout large volumes (up to a few hundred micrometers in each dimension). Although 3DEM has an enormous potential for advancing our understanding of brain function in health and disease, its biomedical application is currently impeded by the lack of accessory technologies, particularly informatics tools for the automated analysis of large and complex sets of image data. The scope of our Sinergia Proposal is to develop technologies for using 3DEM and reconstruct neuronal circuits in the vertebrate brain.High-throughput 3DEM hardware for large-scale brain imaging is currently developed at a rapid pace. However, there are many experimental hurdles in the way of achieving a synaptic resolution level of analysis of dense neuropil. Most importantly, the tools to extract relevant information from the large and extremely complex image stacks are virtually non-existent. Today, the only available method for 3DEM image processing is manual image segmentation. However, methods with strong reliance on human input are not feasible for the analysis of large tissue volumes. Our goal is therefore to develop automated analysis procedures for 3DEM data based on computer vision approaches. We propose to develop state-of-the-art 3DEM technologies, in which we integrate lower-resolution light microscopy (LM) and multicolor staining techniques. Our technological goal is to design a workflow for the acquisition of high-quality 3D image stacks of brain tissue and to develop computational methods for the automated extraction of all connectivity information contained in 3DEM image stacks. In particular, we will design computational tools for the registration of EM and LM imagery, for annotating features in EM data to identify synapses and other subcellular structures, for classifying image voxels to aid volumetric image segmentation, and for tracing of neuron morphologies based on local and global information. We will apply 3DEM techniques to determine connectivity patterns in three vertebrate model circuits: the olfactory system of zebrafish, the song control system of zebra finches, and the neocortex. Since each of these systems is also intensively studied at the physiological level, the connectivity information we provide through our efforts can be used directly to construct quantitative circuit models and derive structural principles governing circuit function. Our project is interdisciplinary and we are composed of neurobiologists and computer vision scientists. We are divided into groups that work on different animal models and pursue different 3DEM approaches. The groups will benefit from each other’s work on virtually all levels because we develop methods that are compatible among the groups and applicable to essentially all animal species. We anticipate that our methods and developments for ultrastructural imaging and analysis will contribute significantly to a new field of research (sometimes referred to as ‘connectomics’ or ‘projectomics’) and make critical progress towards one of the major goals in systems neurobiology, which is the explanation of how brain circuits function.