Fast Mitochondria Segmentation for Connectomics

In connectomics, scientists create the wiring diagram of a mammalian brain by identifying synaptic connections between neurons in nano-scale electron microscopy images. This allows for the identification of dysfunctional mitochondria which are linked to a variety of diseases such as autism or bipolar. However, manual analysis is not feasible since connectomics datasets can be petabytes in size. To process such large data, we present a fully automatic mitochondria detector based on a modified U-Net architecture that yields high accuracy and fast processing times. We evaluate our method on multiple real-world connectomics datasets, including an improved version of the EPFL Hippocampus mitochondria detection benchmark. Our results show a Jaccard index of up to 0.90 with inference speeds lower than 16ms for a 512x512 image tile. This speed is faster than the acquisition time of modern electron microscopes, allowing mitochondria detection in real-time. Compared to previous work, our detector ranks first among real-time methods and third overall. Our data, results, and code are freely available.

[1]  Eric L. Miller,et al.  Segmentation fusion for connectomics , 2011, 2011 International Conference on Computer Vision.

[2]  Alexander G. Gray,et al.  Automatic joint classification and segmentation of whole cell 3D images , 2009, Pattern Recognit..

[3]  Mark H. Ellisman,et al.  Segmentation of mitochondria in electron microscopy images using algebraic curves , 2013, 2013 IEEE 10th International Symposium on Biomedical Imaging.

[4]  Amelio Vázquez Reina,et al.  Radon-Like features and their application to connectomics , 2010, CVPR Workshops.

[5]  Ryan Newton,et al.  A Multiscale Parallel Computing Architecture for Automated Segmentation of the Brain Connectome , 2012, IEEE Transactions on Biomedical Engineering.

[6]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[7]  Nir Shavit,et al.  The big data challenges of connectomics , 2014, Nature Neuroscience.

[8]  José Miguel Buenaposada,et al.  Multi-class segmentation of neuronal structures in electron microscopy images , 2018, BMC Bioinformatics.

[9]  Won-Ki Jeong,et al.  FusionNet: A Deep Fully Residual Convolutional Neural Network for Image Segmentation in Connectomics , 2016, Frontiers in Computer Science.

[10]  Markus Hadwiger,et al.  Scalable Interactive Visualization for Connectomics , 2017, Informatics.

[11]  W. Denk,et al.  The Big and the Small: Challenges of Imaging the Brain’s Circuits , 2011, Science.

[12]  Luca Maria Gambardella,et al.  Deep Neural Networks Segment Neuronal Membranes in Electron Microscopy Images , 2012, NIPS.

[13]  William R. Gray Roncal,et al.  Saturated Reconstruction of a Volume of Neocortex , 2015, Cell.

[14]  Anirban Chakraborty,et al.  Graph-based active learning of agglomeration (GALA): a Python library to segment 2D and 3D neuroimages , 2014, Front. Neuroinform..

[15]  J. Lichtman,et al.  Imaging a 1 mm 3 Volume of Rat Cortex Using a MultiBeam SEM , 2016, Microscopy and Microanalysis.

[16]  Filiz Bunyak,et al.  Mitochondria segmentation in electron microscopy volumes using deep convolutional neural network , 2017, 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM).

[17]  Pascal Fua,et al.  Supervoxel-Based Segmentation of Mitochondria in EM Image Stacks With Learned Shape Features , 2012, IEEE Transactions on Medical Imaging.

[18]  Vincent Lepetit,et al.  Fast Ray features for learning irregular shapes , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[19]  Hanspeter Pfister,et al.  Automatic Neural Reconstruction from Petavoxel of Electron Microscopy Data , 2016 .