Neuroanatomic analysis depends on the reconstruction of complete cell shapes. High-throughput reconstruction of neural circuits, or connectomics, using volume electron microscopy requires dense staining of all cells, which leads even experts to make annotation errors. Currently, reconstruction speed rather than acquisition speed limits the determination of neural wiring diagrams. We developed a method for fast and reliable reconstruction of densely labeled data sets. Our approach, based on manually skeletonizing each neurite redundantly (multiple times) with a visualization-annotation software tool called KNOSSOS, is ∼50-fold faster than volume labeling. Errors are detected and eliminated by a redundant-skeleton consensus procedure (RESCOP), which uses a statistical model of how true neurite connectivity is transformed into annotation decisions. RESCOP also estimates the reliability of consensus skeletons. Focused reannotation of difficult locations promises a rather steep increase of reliability as a function of the average skeleton redundancy and thus the nearly error-free analysis of large neuroanatomical datasets.
Access optionsAccess options
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Subscribe to Journal
Get full journal access for 1 year
only $18.75 per issue
All prices are NET prices.
VAT will be added later in the checkout.
We thank B. Andres, F. Hamprecht, U. Köthe, V. Jain, S. Seung and S. Turaga for many fruitful discussions and comments on the manuscript, J. Bollmann and A. Schaefer for helpful comments on the manuscript, C. Roome for information technology support, J. Kornfeld and F. Svara for programming KNOSSOS, and J. Hanne, H. Jakobi and H. Wissler for help with annotator training. We thank N. Abazova, E. Abs, A. Antunes, P. Bastians, J. Bauer, M. Beez, M. Beining, S. Bender, S. Best, L. Brosi, M. Bucher, E. Buckler, J. Buhmann, C. Burkhardt, F. Drawitsch, L. Ehm, S. Ehm, C. Fianke, R. Foltin, S. Freiss, M. Funk, A. Gebhardt, M. Gruen, K. Haase, J. Hammerich, J. Hanne, B. Hauber, M. Hensen, L. Hofmann, P. Hofmann, M. Hülser, F. Isensee, H. Jakobi, M. Jonczyk, A. Joschko, A. Juenger, S. Kaspar, K. Kessler, A. Khan, M. Kiapes, A. Klein, C. Klein, S. Laiouar, T. Lang, L. Lebelt, H. Lesch, C. Lieven, D. Luft, E. Moeller, A. Muellner, M. Mueller, D. Ollech, A. Oppold, T. Otolski, S. Oumohand, S. Pfarr, M. Pohrath, A. Poos, S. Putzke, J. Reinhardt, A. Rommerskirchen, M. Roth, J. Sambel, K. Schramm, C. Sellmann, J. Sieber, I. Sonntag, M. Stahlberg, T. Stratmann, J. Trendel, F. Trogisch, M. Uhrig, A. Vogel, J. Volz, C. Weber, P. Weber, K. Weiss, L. Weisshaar, E. Wiegand, T. Wiegand, M. Wiese, R. Wiggers, C. Willburger and A. Zegarra for neurite skeletonizations. This work was funded by the Max Planck Society.
Skeleton tracing with KNOSSOS. 3 orthogonal views are displayed (xy, top left, yz, top right, xz, bottom left) and a 3-dimensional view of the dataset bounding box and the skeleton being created (bottom right). The 19MB version requires a video player compatible with DivX-encoded files. The 68MB version can be viewed with a generic player but is shorter.