News |  Sitemap |  Contact
PDF Export  | 

Register, edit, analyze and 3D model terabytes of images with TrakEM2


Our current work in neuroscience requires detailed anatomical descriptions of dendrite
s and axons (the wires), and synapses (the contact points). While we can inspect most wiring details at light microscopy with specific color labels, synapses and very fine processes so far lay under the sensitivity and resolution thresholds. The electron microscope, on the other hand, enables high-resolution imaging of such structures, but with severe costs: sections are not registered to each other; imaging itself induces non-uniform, non-linear deformations; and the sheer amount of image data obtained becomes unhandable for a human operator. To address these and related issues, we have develop
ed TrakEM2: an ImageJ plugin that virtualizes access to very large datasets, and makes its analysis feasible. In particular, TrakEM2 provides the means to create montages from a collection of tiles, to recompose an ultrathin section; to register montages to each other; to precisely segment structures such as neurites across multiple registered sections, and to simple sketch them with simplified tubes and spheres; to perform measurements such as double dissectors for synapse density estimation, and for areas, lengths and volumes; the 3D visualization of segmented elements via Bene Schmid's 3D Viewer plugin; to anotate (and search) with floating text labels; and the comparison of tubular structures (such as neurites or blood vessels) for identification purposes relative to reference data. TrakEM2 works also with confocal stacks and any sort of images supported by ImageJ.

ustitching, image registration, segmentation, 3D visualization, measurement, annotation


Albert Cardona

Institute for Neuroinformatics, Uni/ETH Zurich


Short Biography 

Albert Cardona completed a PhD in Biology (2005) at the University of Barcelona, Spain. After a 4-month internship at the Institute for Neuroinformatics, Zurich, with Rodney Douglas, Albert moved to the University of California in Los Angeles for a postdoc in Drosophila neuroscience, involving from DiI microinjections to light and electron microscopical imgaging, and to programming for large image data set analysis. Albert has developed several plugins for ImageJ and has acquired experience in numerous computer languages over the last 7 years.

© Luxembourg Institute of Science and Technology | Legal Notice