Super-resolution using Sparse Representations over Learned Dictionaries:\n Reconstruction of Brain Structure using Electron Microscopy

2012/10/01 by Tao Hu, Juan Nunez-Iglesias, Hu, Tao +15 · 1 citation
Biochemistry, Genetics and Molecular Biology · Engineering · Materials Science · #Advanced Electron Microscopy Techniques and Applications #Advanced Fluorescence Microscopy Techniques #Computer Vision and Pattern Recognition (cs.CV) #Electron and X-Ray Spectroscopy Techniques #FOS: Biological sciences #FOS: Computer and information sciences #Integrated Circuits and Semiconductor Failure Analysis #Machine Learning (stat.ML) #Neurons and Cognition (q-bio.NC)

paper · pdf · doi:10.48550/arxiv.1210.0564

openalex publication_date 2012/10/01 · openalex created_date 2022/10/01 · openalex updated_date 2026/07/28

Abstract

A central problem in neuroscience is reconstructing neuronal circuits on the\nsynapse level. Due to a wide range of scales in brain architecture such\nreconstruction requires imaging that is both high-resolution and\nhigh-throughput. Existing electron microscopy (EM) techniques possess required\nresolution in the lateral plane and either high-throughput or high depth\nresolution but not both. Here, we exploit recent advances in unsupervised\nlearning and signal processing to obtain high depth-resolution EM images\ncomputationally without sacrificing throughput. First, we show that the brain\ntissue can be represented as a sparse linear combination of localized basis\nfunctions that are learned using high-resolution datasets. We then develop\ncompressive sensing-inspired techniques that can reconstruct the brain tissue\nfrom very few (typically 5) tomographic views of each section. This enables\ntracing of neuronal processes and, hence, high throughput reconstruction of\nneural circuits on the level of individual synapses.\n

Citations

• Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries
• The structure of the nervous system of the nematode Caenorhabditis elegans
• Stable signal recovery from incomplete and inaccurate measurements

Cited by

• Deep Learning for Isotropic Super-Resolution from Non-Isotropic 3D\n Electron Microscopy

Related

• No-Clean-Reference Image Super-Resolution: Application to Electron Microscopy
• Convolutional nets for reconstructing neural circuits from brain images\n acquired by serial section electron microscopy
• Large-Scale Automatic Reconstruction of Neuronal Processes from Electron Microscopy Images
• Learning Multimodal Volumetric Features for Large-Scale Neuron Tracing
• Learning and Segmenting Dense Voxel Embeddings for 3D Neuron Reconstruction