The FlyEM team strives to unravel the neural connectivity of the fly brain using Electron Microscopic (EM) imaging. Reconstructing a connectome is a time-consuming process, often requiring laborious manual tracing (proofreading) on massive image volumes. The FlyEM reconstruction process handles large-scale volumes through a technique called focused proofreading, leveraging synaptic connectivity as a prior for semi-automated tracing while embracing crowd-sourcing. Further processing and analysis uses cutting-edge connectome analysis algorithms and software that we publish in public software repositories under open-source licenses. FlyEM embraces collaboration and provides avenues for data and algorithm sharing and co-development.
Publicly Available Software Packages
DVID: Distributed, Versioned, Image-Oriented Datastore
NeuroProof: Segmentation engine and focused proofreading algorithms
NeuTu: cross-platform software tool for neuron reconstruction and visualization
Raveler: desktop proofreading tool for annotating and refining segmentation in large datasets
DAWMR: Deep And Wide Multiscale Recursive Networks