DIADEM Challenge Research Articles

Automated 3-D Neuron Tracing With Precise Branch Erasing and Confidence Controlled Back Tracking.

The automatic reconstruction of single neurons from microscopic images is essential to enable large-scale data-driven investigations in neuron morphology research. However, few previous methods were able to generate satisfactory results automatically from 3-D microscopic images without human intervention. In this paper, we developed a new algorithm for automatic 3-D neuron reconstruction. The main idea of the proposed algorithm is to iteratively track backward from the potential neuronal termini to the soma centre. An online confidence score is computed to decide if a tracing iteration should be stopped and discarded from the final reconstruction. The performance improvements comparing with the previous methods are mainly introduced by a more accurate estimation of the traced area and the confidence controlled back-tracking algorithm. The proposed algorithm supports large-scale batch-processing by requiring only one user specified parameter for background segmentation. We bench tested the proposed algorithm on the images obtained from both the DIADEM challenge and the BigNeuron challenge. Our proposed algorithm achieved the state-of-the-art results.

A Manual Segmentation Tool for Three-Dimensional Neuron Datasets.

To date, automated or semi-automated software and algorithms for segmentation of neurons from three-dimensional imaging datasets have had limited success. The gold standard for neural segmentation is considered to be the manual isolation performed by an expert. To facilitate the manual isolation of complex objects from image stacks, such as neurons in their native arrangement within the brain, a new Manual Segmentation Tool (ManSegTool) has been developed. ManSegTool allows user to load an image stack, scroll down the images and to manually draw the structures of interest stack-by-stack. Users can eliminate unwanted regions or split structures (i.e., branches from different neurons that are too close each other, but, to the experienced eye, clearly belong to a unique cell), to view the object in 3D and save the results obtained. The tool can be used for testing the performance of a single-neuron segmentation algorithm or to extract complex objects, where the available automated methods still fail. Here we describe the software's main features and then show an example of how ManSegTool can be used to segment neuron images acquired using a confocal microscope. In particular, expert neuroscientists were asked to segment different neurons from which morphometric variables were subsequently extracted as a benchmark for precision. In addition, a literature-defined index for evaluating the goodness of segmentation was used as a benchmark for accuracy. Neocortical layer axons from a DIADEM challenge dataset were also segmented with ManSegTool and compared with the manual “gold-standard” generated for the competition.

SparseTracer: the Reconstruction of Discontinuous Neuronal Morphology in Noisy Images.

Digital reconstruction of a single neuron occupies an important position in computational neuroscience. Although many novel methods have been proposed, recent advances in molecular labeling and imaging systems allow for the production of large and complicated neuronal datasets, which pose many challenges for neuron reconstruction, especially when discontinuous neuronal morphology appears in a strong noise environment. Here, we develop a new pipeline to address this challenge. Our pipeline is based on two methods, one is the region-to-region connection (RRC) method for detecting the initial part of a neurite, which can effectively gather local cues, i.e., avoid the whole image analysis, and thus boosts the efficacy of computation; the other is constrained principal curves method for completing the neurite reconstruction, which uses the past reconstruction information of a neurite for current reconstruction and thus can be suitable for tracing discontinuous neurites. We investigate the reconstruction performances of our pipeline and some of the best state-of-the-art algorithms on the experimental datasets, indicating the superiority of our method in reconstructing sparsely distributed neurons with discontinuous neuronal morphologies in noisy environment. We show the strong ability of our pipeline in dealing with the large-scale image dataset. We validate the effectiveness in dealing with various kinds of image stacks including those from the DIADEM challenge and BigNeuron project.

Rivulet: 3D Neuron Morphology Tracing with Iterative Back-Tracking.

The digital reconstruction of single neurons from 3D confocal microscopic images is an important tool for understanding the neuron morphology and function. However the accurate automatic neuron reconstruction remains a challenging task due to the varying image quality and the complexity in the neuronal arborisation. Targeting the common challenges of neuron tracing, we propose a novel automatic 3D neuron reconstruction algorithm, named Rivulet, which is based on the multi-stencils fast-marching and iterative back-tracking. The proposed Rivulet algorithm is capable of tracing discontinuous areas without being interrupted by densely distributed noises. By evaluating the proposed pipeline with the data provided by the Diadem challenge and the recent BigNeuron project, Rivulet is shown to be robust to challenging microscopic imagestacks. We discussed the algorithm design in technical details regarding the relationships between the proposed algorithm and the other state-of-the-art neuron tracing algorithms.

Automated Reconstruction of Dendritic and Axonal Trees by Global Optimization with Geometric Priors

We present a novel probabilistic approach to fully automated delineation of tree structures in noisy 2D images and 3D image stacks. Unlike earlier methods that rely mostly on local evidence, ours builds a set of candidate trees over many different subsets of points likely to belong to the optimal tree and then chooses the best one according to a global objective function that combines image evidence with geometric priors. Since the best tree does not necessarily span all the points, the algorithm is able to eliminate false detections while retaining the correct tree topology. Manually annotated brightfield micrographs, retinal scans and the DIADEM challenge datasets are used to evaluate the performance of our method. We used the DIADEM metric to quantitatively evaluate the topological accuracy of the reconstructions and showed that the use of the geometric regularization yields a substantial improvement.

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.

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.

A Brief History of Neuronal Reconstruction

The routes taken by peripheral nerves and the intricacy present even in the external features of the CNS had been understood in considerable detail at the time of Andreas Vesalius (1543) and Bartholomeo Eustatius (1552). But the astonishing underlying and microscopically symplectic character of CNS brain tissue only became evident in fitful stages and over the following 300 and 50 years. Today’s opportunity to precisely reconstruct individual neurons and circuits, presently expressed in the DIADEM challenge, is based on a long sequence of conceptual and technological advances in our understanding of brain. A selection among these are highlighted in the following abbreviated historical account, which focuses first on our means of knowing that neurons exist, and then on how we learned to record and analyze their shapes.

Bright Field Neuronal Preparation Optimized for Automatic Computerized Reconstruction, a Case with Cerebellar Climbing Fibers

To make the best use of computerized automatic reconstruction it is crucial to supply optimized neuronal preparations for reconstruction. For bright field preparations a major technique for labeling a specific population of neurons is the injection of tracers, such as dextran conjugated with biotin (biotinylated dextran amine, BDA). BDA-labeled neurons and axons can then be visualized in brain sections using the diaminobenzidine (DAB) reaction, which causes BDA to form a dark precipitate. It is important for such brain sections to be oriented in an appropriate direction so that labeled neurons and axons are localized within individual sections to the greatest extent possible, because automatic reconstruction can be applied only within single brain sections (thickness of 50–100 μm) at the moment. In this sense, cerebellar climbing fibers, which are organized within a flat longitudinal plane, are among the best suited neuronal elements for automatic reconstruction. For automatic reconstruction, it is also important to separate individual, labeled neurons (or axons) from each other by controlling the amount of tracer. By solving these points the histological preparation can be optimized for automatic computerized reconstruction. Developers and users of automatic reconstruction software are usually different individuals. Mutual understanding of the technical difficulties and scientific needs on either side is indispensable to optimize collaboration between the two groups. We provided bright-field preparation of climbing fibers labeled with BDA to the DIADEM challenge. By looking at how the automatic reconstruction software works in the DIADEM conference, I realized that whether the computerized

Diadem X: Automated 4 Dimensional Analysis of Morphological Data

The development of multi-photon imaging technique has greatly facilitated in vivo time-lapse imaging and enables comparison of the fine morphological structures of individual neurons over time. Despite the fact that 4D data acquisition has become easier and can be applied to a variety of brain tissues, both in vivo and in tissue slices, the analysis of these 4D data remains extremely laborious and painstaking. Manual analysis greatly limits the pace of research and introduces errors. Recent work suggests that an automated dynamic analysis tool can be successfully applied to time-lapse images of cultured neurons,1 which are flat and relatively simple. So far no such automated analysis program has yet been reported for in vivo 4D image analysis, which poses several technical challenges, including alignment of complex 3 dimensional structures across time points, and identification of persistent and dynamic structures. The goals of the Diadem Challenge were to generate algorithms for automated reconstruction of light microscope images of 3D neural structures. Quantitative characterization of neuronal structure will be extremely valuable for the identification of cell types, to establish and interpret neuronal connectivity maps, for comparative analysis across individuals, brain regions or experimental conditions, and to identify plasticity events and mechanisms that regulate plasticity, for instance by analysis of the structure of neurons from animals of different genotypes or with different experience or training. These goals are based on the premise that neuronal morphology is stereotypic, so that data from single reconstructions are meaningful within a dataset. Participants of the Diadem Challenge explored different model-based algorithms based on either local voxel information or global objective function of the image to automatically reconstruct the neuronal structure.2 The ultimate goal is to achieve completely automated reconstruction with high precision and efficiency. Progress demonstrated by participants of the Diadem Challenge indicates that the development of algorithms for automated reconstruction of neuronal structures is very promising and will be able to significantly facilitate supervised reconstruction of neurons in complex data sets. This progress is very exciting and will have a huge impact on current research in basic neuroscience and as it pertains to human disease. One important advance from this point will be to automate analysis of time-lapse imaging data. Applications of automated analysis of time-lapse imaging data include analysis of the development of neuronal structure and acquisition of cell type specific morphological features, analysis of mechanisms of structural plasticity, and events underlying degenerative and regenerative events.

The DIADEM data sets: representative light microscopy images of neuronal morphology to advance automation of digital reconstructions.

The comprehensive characterization of neuronal morphology requires tracing extensive axonal and dendritic arbors imaged with light microscopy into digital reconstructions. Considerable effort is ongoing to automate this greatly labor-intensive and currently rate-determining process. Experimental data in the form of manually traced digital reconstructions and corresponding image stacks play a vital role in developing increasingly more powerful reconstruction algorithms. The DIADEM challenge (short for DIgital reconstruction of Axonal and DEndritic Morphology) successfully stimulated progress in this area by utilizing six data set collections from different animal species, brain regions, neuron types, and visualization methods. The original research projects that provided these data are representative of the diverse scientific questions addressed in this field. At the same time, these data provide a benchmark for the types of demands automated software must meet to achieve the quality of manual reconstructions while minimizing human involvement. The DIADEM data underwent extensive curation, including quality control, metadata annotation, and format standardization, to focus the challenge on the most substantial technical obstacles. This data set package is now freely released ( http://diademchallenge.org ) to train, test, and aid development of automated reconstruction algorithms.