Launch Viewer
What is CATMAID?
CATMAID (Collaborative Annotation Toolkit for Massive Amounts of Image Data) is an open-source, browser-based platform for navigating and annotating large-scale 3D biological image datasets — particularly serial section electron microscopy (ssEM) stacks used in connectomics.
Inspired by Google Maps, CATMAID renders image tiles on demand, allowing seamless pan-and-zoom navigation through datasets that are terabytes in size — without downloading the data first. Multiple users can simultaneously trace neurons, mark synapses, and annotate structures, with all annotations stored in a shared database.
The OCP hosts a CATMAID instance pre-loaded with the Kasthuri et al. 2015 mouse neocortex dataset and related annotation layers (segments, synapses, vesicles, mitochondria) at full 6×6×30 nm resolution.
Google Maps for the Brain
Tile-based rendering at multiple zoom levels. Navigate from a millimeter-scale overview down to individual synaptic vesicles at nanometer resolution without leaving your browser.
Collaborative Annotation
Multiple users can simultaneously trace neuron skeletons and place synapse annotations. All work is stored in a shared database with full provenance and version tracking.
Deep Linking
Any view — specific coordinates, zoom level, active stack — is captured in the URL. Share a bookmark and collaborators instantly jump to the exact same location in the dataset.
SVG + WebGL Viewer
Traced neuron skeletons rendered in SVG overlay on EM tiles. Full 3D morphology viewer using WebGL for reconstructed neurons, with NBLAST similarity analysis support.
Available in This Viewer
The OCP CATMAID instance (Project 4) hosts the following stacks from the Kasthuri et al. 2015 reconstruction. Use the stack selector in the CATMAID toolbar to switch between layers.
kasthuri11cc
EM Image
Color-corrected electron microscopy image stack. The recommended layer for navigation and visual inspection. Full 6×6×30 nm resolution.
kat11segments
Segmentation
Automated + proofread neuron segment labels. Each color represents a unique neural process (axon or dendrite). Overlay on the EM image.
kat11synapses
Annotation
~1,700 annotated synaptic sites. Displays pre- and post-synaptic markers and PSD locations in the EM volume.
kat11vesicles
Annotation
Individual synaptic vesicle cluster annotations within presynaptic boutons.
kat11mito
Annotation
Mitochondria segmentation and annotation throughout the reconstruction volume.
ac3 / ac4
ROI
Fully proofread high-resolution cylinder regions of interest — the gold standard annotation areas used for algorithm benchmarking.
CATMAID Features
Terabyte Image Browsing
Tile-pyramid rendering handles datasets of arbitrary size. Only visible tiles are fetched — no waiting for bulk downloads.
Neuron Tracing
Place skeleton nodes along neuron processes. CATMAID automatically builds the tree topology and lets you tag nodes (TODO, soma, uncertain continuation, etc.).
Synapse Annotation
Mark pre- and post-synaptic sites by clicking on connector nodes. Connectivity is computed automatically from the annotation graph.
Connectivity Widgets
Built-in widgets show adjacency matrices, partner neurons, and N-hop connectivity graphs — all computed live from the annotation database.
Semantic Annotations
Apply hierarchical labels (e.g. "excitatory", "layer IV", "myelinated") to any neuron or node. Filter and group neurons by annotation.
REST API
All annotation data is accessible via CATMAID's HTTP API. Use catpy (Python) or rcatmaid (R) for programmatic skeleton export and analysis.
Citation
Saalfeld S, Cardona A, Hartenstein V, Tomančák P. CATMAID: collaborative annotation toolkit for massive amounts of image data. Bioinformatics. 2009 Aug 1;25(15):1984–6.
Bioinformatics · 2009 · DOI: 10.1093/bioinformatics/btp266 · PMID: 19494038