Large-Scale Automatic Reconstruction of Neuronal Processes from Electron Microscopy Images

@article{Kaynig2015LargeScaleAR,
  title={Large-Scale Automatic Reconstruction of Neuronal Processes from Electron Microscopy Images},
  author={Verena Kaynig and Amelio V{\'a}zquez Reina and Seymour Knowles-Barley and Mike Roberts and Thouis Raymond Jones and Narayanan Kasthuri and Eric L. Miller and Jeff William Lichtman and Hanspeter Pfister},
  journal={Medical image analysis},
  year={2015},
  volume={22 1},
  pages={
          77-88
        }
}
Automated sample preparation and electron microscopy enables acquisition of very large image data sets. These technical advances are of special importance to the field of neuroanatomy, as 3D reconstructions of neuronal processes at the nm scale can provide new insight into the fine grained structure of the brain. Segmentation of large-scale electron microscopy data is the main bottleneck in the analysis of these data sets. In this paper we present a pipeline that provides state-of-the art… Expand
Large-Scale Electron Microscopy Image Segmentation in Spark
TLDR
A novel strategy to apply image segmentation algorithms to apply on very large datasets that exceed the capacity of a single machine is proposed and is robust to potential segmentation errors which could otherwise severely compromise the quality of the overall segmentation. Expand
Residual Deconvolutional Networks for Brain Electron Microscopy Image Segmentation
TLDR
The proposed residual deconvolutional network consists of two information pathways that capture full-resolution features and contextual information, respectively, and it was shown that the proposed model is very effective in achieving the conflicting goals in dense output prediction; namely preserving full- resolution predictions and including sufficient contextual information. Expand
Scalable algorithms for semi-automatic segmentation of electron microscopy images of the brain tissue
TLDR
A set of fast and scalable algorithms for segmenting very large 3D images of brain tissue and a set of modification to the standard algorithm and a quasi–linear agglomerative clustering algorithm that can greatly reduce the over–segmentation produced by the standard watershed algorithm are proposed. Expand
A modular hierarchical approach to 3D electron microscopy image segmentation
TLDR
A fully automatic approach for intra-section segmentation and inter-section reconstruction of neurons using EM images is proposed and a semi-automatic method that utilizes the intermediate outputs of the automatic algorithm and achieves intra-segmentation with minimal user intervention is developed. Expand
Automatically segmenting and reconstructing neurons in SEM images
TLDR
A method that Deep Convolutional Neural Network is used for neuronal boundary detection; and then, with the membrane detection probability map (MDPM) generated by DCNN, a marker-controlled watershed method is applied to segment neurons in the EM images over the MDPM. Expand
Flexible Learning-Free Segmentation and Reconstruction of Neural Volumes
TLDR
Compared to existing supervised learning methods, FLoRIN is one to two orders of magnitude faster and produces high-quality reconstructions that are tolerant to noise and artifacts, as is shown qualitatively and quantitatively. Expand
Segmentation in Stacks of Electron Microscopy Images Using Deep Learning
TLDR
Experiments showed that DCN accuracy can be improved by changing the kernel size of the deconvolution layer by the proposed hyperparameter tuning method based on visualization of feature maps. Expand
An automated images-to-graphs framework for high resolution connectomics
TLDR
This manuscript presents the first fully-automated images-to-graphs pipeline (i.e., a pipeline that begins with an imaged volume of neural tissue and produces a brain graph without any human interaction), and develops a metric to assess the quality of the output graphs. Expand
SegEM: Efficient Image Analysis for High-Resolution Connectomics
TLDR
SegEM resolves the tradeoff between synapse detection and semi-automated reconstruction performance in high-resolution connectomics and makes efficient circuit reconstruction in fully-stained EM datasets a ready-to-use technique for neuroscience. Expand
An Automated Images-to-Graphs Pipeline for High Resolution Connectomics
TLDR
This work has created the first fully automated images-tographs pipeline (i.e., a pipeline that begins with an imaged volume of neural tissue and produces a brain graph without any human interaction), and develops a metric to assess the quality of the output graphs. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 57 REFERENCES
Automation of 3D reconstruction of neural tissue from large volume of conventional serial section transmission electron micrographs
  • Y. Mishchenko
  • Computer Science, Medicine
  • Journal of Neuroscience Methods
  • 2009
TLDR
This work describes an approach for automation of the process of reconstruction of neural tissue from serial section transmission electron micrographs, and uses these to evaluate reconstruction speed, quality, error rates, and presence of ambiguous locations in neuropil ssTEM imaging data. Expand
Detection of neuron membranes in electron microscopy images using a serial neural network architecture
TLDR
The method of serial ANNs is developed and it is shown that the learned context does improve detection over traditional ANNs and also demonstrates advantages over previous membrane detection methods. Expand
Efficient automatic 3D-reconstruction of branching neurons from EM data
TLDR
The core of the system is a set of possible assignments, each of which proposes with some cost a link between neuron regions in consecutive sections, which can model the continuation, branching, and end of neurons. Expand
Supervoxel-Based Segmentation of Mitochondria in EM Image Stacks With Learned Shape Features
TLDR
This work proposes an automated graph partitioning scheme that is able to segment mitochondria at a performance level close to that of a human annotator, and outperforms a state-of-the-art 3-D segmentation technique. Expand
Biologically inspired EM image alignment and neural reconstruction
TLDR
An alternative computational approach for ssTEM image analysis using biologically inspired receptive fields as a basis for a ridge detection algorithm to identify cell membranes, synaptic contacts and mitochondria and joined small segments of membrane into cell surfaces using a dynamic programming algorithm similar to the Needleman-Wunsch and Smith-Waterman DNA sequence alignment procedures. Expand
Increasing depth resolution of electron microscopy of neural circuits using sparse tomographic reconstruction
TLDR
First, it is shown that the brain tissue can be represented as sparse linear combination of local basis functions that are thin membrane-like structures oriented in various directions that enables tracing of neuronal connections across layers and, hence, high throughput reconstruction of neural circuits to the level of individual synapses. Expand
Neural Process Reconstruction from Sparse User Scribbles
We present a novel semi-automatic method for segmenting neural processes in large, highly anisotropic EM (electron microscopy) image stacks. Our method takes advantage of sparse scribble annotationsExpand
Computational methods and challenges for large-scale circuit mapping
TLDR
This field of neuronal data analysis with emphasis on reconstructing neurons from EM data cubes is reviewed, with goals that include reconstructions of the morphology of individual neurons as well as entire circuits. Expand
Neuron geometry extraction by perceptual grouping in ssTEM images
TLDR
This work proposes a novel framework for the segmentation of thin elongated structures like membranes in a neuroanatomy setting using the probability output of a random forest classifier in a regular cost function, which enforces gap completion via perceptual grouping constraints. Expand
An optimal-path approach for neural circuit reconstruction
TLDR
An automatic method for finding neurons in sequences of 2D sections by formulating the problem of finding paths through sets of sections as an optimal path computation and solving this optimization problem using Dijkstra's algorithm. Expand
...
1
2
3
4
5
...