Neuronal morphology plays a crucial role in the information processing capabilities of neurons. Neuronal function closely depends on properties of neuronal morphology such as lengths and diameters. A typical neuron consists of soma (central part of the neuron), dendrites (cellular extensions with extensive branching), axon (a long, slender projection of a nerve cell) and axon terminal (contains synapses). Morphologies of dendrites and axons are key parameters in descriptions of neuronal morphology and hence cell types. The neuronal morphology-function relation is the major focus of current neuroscience research.

To understand the neuronal behavior it is important to reconstruct 3D depiction from sets of 2D images acquired at different depths using brightfield or fluorescence microscopy. A major obstacle to this research is the lack of automated methods to produce morphological reconstructions of neurons that would lead to the creation of libraries of neuronal morphologies. Further,

This process transforms the voxel-based content of the raw pictures into a set of interconnected vectors or cylinders representing the spatial coordinates and orientation of each branch. The resulting “digital reconstructions” provide a computationally compact but complete description of dendrites and axons, suitable for quantitative morphometric and stereological analysis, compartmental biophysical modeling, and electrophysiological simulations.

Geometric reconstructions are commonly done manually with programs such as Neurolucida or Neurozoom and the whole process is labor intensive and time consuming. The accuracy of manual reconstruction however is strongly dependent on individual data interpretation to estimate midlines and diameters of dendrites. A major obstacle to this research is the lack of automated methods to produce morphological reconstructions of neurons and in principle it could be accomplished by a variety of methods. However, accuracy of automated methods is low and they failed to provide generalized solution for the problem. Semi-automatic methods for neuron morphology reconstruction have been also proposed such as skeletonization-based cylindrical extraction-based approaches although they still require manual pre- or post- processing.
Digital Reconstruction of Axonal and Dendritic Morphology (DIADEM) competition is an attempt to solve this problem by challenging together computational and experimental scientists to test the most promising image analysis approaches against the latest data in a real-world environment. DIADEM will give a year to design algorithm to automatically reconstruct neuronal arbors and test it against the manual gold standard. The DIADEM Challenge is organized by the Allen Institute for Brain Science, the Howard Hughes Medical Institute and the Krasnow Institute for Advanced Study at George Mason University.

Organizers of a new competition hope to provide incentives for the development of new computer algorithms to advance the field – including a cash prize of up to $75,000 for the qualifying winner.

Learn more about DIADEM