Arunachalam Narayanaswamy

A Broadly Applicable 3-D Neuron Tracing Method Based on Open-Curve Snake

This paper presents a broadly applicable algorithm and a comprehensive open-source software implementation for automated tracing of neuronal structures in 3-D microscopy images. The core 3-D neuron tracing algorithm is based on three-dimensional (3-D) open-curve active Contour (Snake). It is initiated from a set of automatically detected seed points. Its evolution is driven by a combination of deforming forces based on the Gradient Vector Flow (GVF), stretching forces based on estimation of the fiber orientations, and a set of control rules. In this tracing model, bifurcation points are detected implicitly as points where multiple snakes collide. A boundariness measure is employed to allow local radius estimation. A suite of pre-processing algorithms enable the system to accommodate diverse neuronal image datasets by reducing them to a common image format. The above algorithms form the basis for a comprehensive, scalable, and efficient software system developed for confocal or brightfield images. It provides multiple automated tracing modes. The user can optionally interact with the tracing system using multiple view visualization, and exercise full control to ensure a high quality reconstruction. We illustrate the utility of this tracing system by presenting results from a synthetic dataset, a brightfield dataset and two confocal datasets from the DIADEM challenge.

The FARSIGHT Trace Editor: An Open Source Tool for 3-D Inspection and Efficient Pattern Analysis Aided Editing of Automated Neuronal Reconstructions

Despite considerable progress in the field of automated neurite arbor reconstruction from two-dimensional (2-D) or three-dimensional (3-D) images acquired using optical microscopes, even the best available automated systems of today have a non-zero error rate, implying the continued need for visual proofreading and corrective editing systems. Currently available systems require the user to visually detect potential errors, and perform corrective edits serially, one at a time. Consequently, it is not uncommon for the task of proofreading an automated reconstruction to take longer than de novo manual reconstruction in the hands of a skilled operator, appearing to defeat the utility of automated reconstruction systems. There is a compelling need in the neuroscience research community for smarter and more scalable proofreading tools that can significantly accelerate the process, and reduce the tedium and manpower cost. This need is especially compelling in large-scale and highthroughput studies that require large numbers of neurite arbor measurements. Meeting the above-mentioned need calls for a well-integrated combination of methods for (i) detailed visualization of reconstructions overlaid on the source images (usually large 2-D/3-D, and multi-channel image data); (ii) rapid identification of tracing errors; and (iii) rapid and minimal-effort correction of errors using interactive graphical tools. There is a parallel need among computer science researchers who are developing new and improved algorithms for automated reconstruction. Each newly developed algorithm must be validated, and its performance quantified in order to determine the extent to which it improves upon previously developed systems. Detailed performance evaluation data can enable algorithm developers to focus efforts on the most important algorithm deficiencies. Traditionally, this goal has been met by comparing automatically generated reconstructions against pre-established “gold standard” reconstructions. Currently available gold standards are generated by manual reconstruction, and sometimes proofread by multiple human observers to generate a consensus of reconstructions. This process is expensive, slow, and not scalable. Manual reconstructions are simply unavailable in an operational setting such as a highthroughput study. These considerations suggest the need for alternative performance evaluation and validation methodologies. In this paper, we advocate edit-based methodologies for validation and performance assessment. These methods become practically appropriate as and when the performance of automated algorithms is close to being optimal, i.e., the differences between automated reconstructions and human reconstructions are less than 10%. Under these 0 Peng H., Long F., Zhao T., Myers E.W., (2010) Proof-editing is the bottleneck of 3D neuron reconstruction: The problem and solutions, Neuroinformatics, published online: 17 Dec. 2010, in press. 0 Meijering E. (2010). Neuron tracing in perspective, Cytometry A., 77 (7):693–704. 1 ij ri . ( ). r tr i i rs ti , t tr ., ( ): . 2

Robust Adaptive 3-D Segmentation of Vessel Laminae From Fluorescence Confocal Microscope Images and Parallel GPU Implementation

This paper presents robust 3-D algorithms to segment vasculature that is imaged by labeling laminae, rather than the lumenal volume. The signal is weak, sparse, noisy, nonuniform, low-contrast, and exhibits gaps and spectral artifacts, so adaptive thresholding and Hessian filtering based methods are not effective. The structure deviates from a tubular geometry, so tracing algorithms are not effective. We propose a four step approach. The first step detects candidate voxels using a robust hypothesis test based on a model that assumes Poisson noise and locally planar geometry. The second step performs an adaptive region growth to extract weakly labeled and fine vessels while rejecting spectral artifacts. To enable interactive visualization and estimation of features such as statistical confidence, local curvature, local thickness, and local normal, we perform the third step. In the third step, we construct an accurate mesh representation using marching tetrahedra, volume-preserving smoothing, and adaptive decimation algorithms. To enable topological analysis and efficient validation, we describe a method to estimate vessel centerlines using a ray casting and vote accumulation algorithm which forms the final step of our algorithm. Our algorithm lends itself to parallel processing, and yielded an 8× speedup on a graphics processor (GPU). On synthetic data, our meshes had average error per face (EPF) values of (0.1-1.6) voxels per mesh face for peak signal-to-noise ratios from (110-28 dB). Separately, the error from decimating the mesh to less than 1% of its original size, the EPF was less than 1 voxel/face. When validated on real datasets, the average recall and precision values were found to be 94.66% and 94.84%, respectively.

3-D Image Pre-processing Algorithms for Improved Automated Tracing of Neuronal Arbors

The accuracy and reliability of automated neurite tracing systems is ultimately limited by image quality as reflected in the signal-to-noise ratio, contrast, and image variability. This paper describes a novel combination of image processing methods that operate on images of neurites captured by confocal and widefield microscopy, and produce synthetic images that are better suited to automated tracing. The algorithms are based on the curvelet transform (for denoising curvilinear structures and local orientation estimation), perceptual grouping by scalar voting (for elimination of non-tubular structures and improvement of neurite continuity while preserving branch points), adaptive focus detection, and depth estimation (for handling widefield images without deconvolution). The proposed methods are fast, and capable of handling large images. Their ability to handle images of unlimited size derives from automated tiling of large images along the lateral dimension, and processing of 3-D images one optical slice at a time. Their speed derives in part from the fact that the core computations are formulated in terms of the Fast Fourier Transform (FFT), and in part from parallel computation on multi-core computers. The methods are simple to apply to new images since they require very few adjustable parameters, all of which are intuitive. Examples of pre-processing DIADEM Challenge images are used to illustrate improved automated tracing resulting from our pre-processing methods.

Self-aligned Schwann cell monolayers demonstrate an inherent ability to direct neurite outgrowth

In vivo nerve guidance channel studies have identified Schwann cell (SC) presence as an integral factor in axonal number and extension in an injury site, and in vitro studies have provided evidence that oriented SCs can direct neurite outgrowth. However, traditional methods used to create oriented SC monolayers (e.g. micropatterns/microtopography) potentially introduce secondary guidance cues to the neurons that are difficult to de-couple. Although SCs expanded on uniform laminin-coated coverslips lack a global orientation, the monolayers contain naturally formed regions of locally oriented cells that can be used to investigate SC-mediated neurite guidance. In this work, novel image analysis techniques have been developed to quantitatively assess local neurite orientation with respect to the underlying regional orientation of the Schwann cell monolayer. Results confirm that, in the absence of any secondary guidance cues, a positive correlation exists between neurite outgrowth and regional orientation of the SC monolayer. Thus, SCs alone possess an inherent ability to direct neurite outgrowth, and expansion of the co-culture-based quantitative method described can be used to further deconstruct specific biomolecular mechanisms of neurite guidance.

Novel 4-D Open-Curve Active Contour and curve completion approach for automated tree structure extraction

We present novel approaches for fully automated extraction of tree-like tubular structures from 3-D image stacks. A 4-D Open-Curve Active Contour (Snake) model is proposed for simultaneous 3-D centerline tracing and local radius estimation. An image energy term, stretching term, and a novel region-based radial energy term constitute the energy to be minimized. This combination of energy terms allows the 4-D open-curve snake model, starting from an automatically detected seed point, to stretch along and fit the tubular structures like neurites and blood vessels. A graph-based curve completion approach is proposed to merge possible fragments caused by discontinuities in the tree structures. After tree structure extraction, the centerlines serve as the starting points for a Fast Marching segmentation for which the stopping time is automatically chosen. We illustrate the performance of our method with various datasets.

5-D imaging and parallel automated analysis of cellular events in living immune tissue microenvironments

The mammalian immune system consists of vital tissue microenvironments that exhibit remarkable structural complexity and dynamic cellular behaviors. It is now possible to acquire time-lapse series of multi-channel three dimensional images of multiple cell types and vasculature simultaneously, revealing dynamic events in their living tissue context. This talk will describe automated image analysis algorithms, parallel computation methods, and large-scale edit-based validation methods to detect and quantify key events such as homogeneous and heterogeneous cell-cell interactions, and methods to map events to their tissue context.