View article

Reconstructing neuronal circuits at such high resolutions that synaptic connections are clearly visible can currently only be done from image data acquired via electron microscopy (EM). These stacks of images enable precise 3D reconstructions of neuronal morphology. While automatic methods for segmenting such images are certainly improving, much annotation and segmentation still needs to be done by human operators carefully examining the images. In addition, the EM data sets that must be dealt with are often many terabytes in size. The requirement for hundreds of annotators to each have a local copy would be prohibitively expensive. To address these requirements, we have extended CATMAID, the Collaborative Annotation Toolkit for Massive Amounts of Image Data, ¹ to allow many researchers to trace neurons collaboratively in the same data set. CATMAID is a web-based system, so each annotator …