Hardware Accelerated Segmentation of Complex Volumetric Filament Networks

Mayerich D., Keyser J.

IEEE Transactions on Visualization and Computer Graphics, 2009 (preprint PDF)

ACM Symposium on Solid and Physical Modeling, 2008 [BibTex] (short paper)

Filaments are common structures in microscopy data sets. Common filament structures include microvascular networks and neuronal processes (axons and dendrites). Although these structures are highly complex, they generally take up a very small fraction of the overall volume of the microscopy data sets. We present a method for efficiently tracking and storing filament networks by taking advantage of their limited overall volume.

Visualization of Cellular and Microvascular Relationships

Mayerich D., Abbott L.C., Keyser J.

IEEE Transactions on Visualization and Computer Graphics, 2008 [BibTex] (preprint PDF)

Proceedings of IEEE Visualization, 2008 (video)

Understanding the structure of microvessels and their relationship to cells in biological tissue is an important and complex problem. In this paper, we describe methods for encoding the unique structure of microvascular networks, allowing researchers to selectively explore microvascular anatomy. By associating cellular structures with our microvascular framework, we allow researchers to explore interesting anatomical relationships in dense and complex data sets.

Constructing High-Resolution Microvascular Models

Mayerich D., Kwon J., Choe Y., Abbott L.C., Keyser J.

3^rd Microscopic Image Analysis with Applications in Biology Workshop, 2008 [BibTex] (full PDF)

The anatomical structure of the brain microvascular system plays an important role in understanding the function of chemical transport within the brain. We describe the imaging and segmentation methods used to construct a structural model of complex microvascular networks. This model is then used to extract high-resolution anatomical statistics while providing a framework for further study.

Knife-Edge Scanning Microscopy for Imaging and Reconstruction of Three-Dimensional Anatomical Structures of the Mouse Brain

Mayerich D., Abbott L.C., McCormick B.H.

Journal of Microscopy, 2008 [BibTex] (cover image) (preprint PDF)

Anatomical information at the cellular level is important in many fields including organ systems development, computational biology, and informatics. We describe a new microscopy technique known as Knife-Edge Scanning Microscopy (KESM), which allows us to image large volumes of tissue at microscopic resolution. We do this by performing automated serial sectioning of the specimen, allowing us to overcome the depth and resolution constraints that limit current optical sectioning methods. 

Automated Lateral Sectioning for Knife-Edge Scanning Microscopy

Kwon J., Mayerich D., Choe Y., McCormick B.H.

IEEE International Symposium on Biomedical Imaging, 2008 [BibTex] (preprint PDF)

Knife-Edge Scanning Microscopy (KESM) overcomes several depth-related constraints on optical imaging by using serial physical sectioning. However, since the imaged tissue is physically ablated, KESM is limited to imaging specimens smaller than the field of view (FOV) of the objective. We discuss sectioning and alignment algorithms used to overcome these problems, allowing us to automatically image large tissue specimens such as entire organs.

Noise and Artifact Removal in Knife-Edge Scanning Microscopy

Mayerich D., McCormick B., Keyser J.

IEEE International Symposium on Biomedical Imaging, 2007 [BibTex] (preprint PDF)

Knife-Edge Scanning Microscopy (KESM) is a high-throughput imaging technique for large specimens. Due to the high data rate, very little time is available for image processing to remove noise and imaging artifacts. In this paper, we discuss fast and memory-efficient methods for removing noise and imaging artifacts from KESM data sets.

Visualization of Fibrous and Thread-like Data

Melek Z, Mayerich D., Yuksel C., Keyser J.

Transactions on Visualization and Computer Graphics, 2006  [BibTex]

Proceedings of IEEE Visualization, 2006 (preprint PDF)

Thread-like structures are commonly found in biomedical images and volumetric data. Unlike fibers found in DT-MRI images, filaments found in microscopy samples are less directed and contain complex interconnections. In this paper, we discuss methods for interactively visualizing these complex filament networks using oriented imposters. We also describe methods for selective visualization of connected components (such as individual neurons) and shading techniques for understanding large-scale structure.

Construction of Anatomically Correct Models of Mouse Brain Networks

McCormick B., Koh W., Choe Y., Abbott L., Keyser J., Mayerich D., Melek Z., Doddapaneni P.

Neurocomputing, 2004 [BibTex] (preprint PDF)

In this paper, we describe how to construct a database, The Mouse Brain Web, based on large-scale images of mouse brain tissue at the microscopic level. Each web page in this database provides the position, orientation, morphology, and putative synapses for each biologically observed neuron. The Mouse Brain Web has been designed to support (1) mapping of the spatial distribution and morphology of neurons by type; (2) wiring of the network and synaptic assembly; (3) projection of neuron morphology and synapses to geometric multi-compartmental models; (4) search for motifs and canonical circuits in the brain networks using customized web-crawlers; and (5) the mapping of anatomically correct networks to physiologically correct network simulations.