Design of the first neuronal connectomics challenge: From imaging to connectivity

Abstract

We are organizing a challenge to reverse engineer the structure of neuronal networks from patterns of activity recorded with calcium fluorescence imaging. Unraveling the brain structure at the neuronal level at a large scale is an important step in brain science, with many ramifications in the comprehension of animal and human intelligence and learning capabilities, as well as understanding and curing neuronal diseases and injuries. However, uncovering the anatomy of the brain by disentangling the neural wiring with its very fine and intertwined dendrites and axons, making both local and far reaching synapses, is a very arduous task: traditional methods of axonal tracing are tedious, difficult, and time consuming. This challenge proposes to approach the problem from a different angle, by reconstructing the effective connectivity of a neuronal network from observations of neuronal activity of thousands of neurons, which can be obtained with state-of-the-art fluorescence calcium imaging. To evaluate the effectiveness of proposed algorithms, we will use data obtained with a realistic simulator of real neurons for which we have ground truth of the neuronal connections. We produced simulated calcium imaging data, taking into account a model of fluorescence and light scattering. The task of the participants is to reconstruct a network of 1000 neurons from time series of neuronal activities obtained with this model. This challenge is part of the official selection of the WCCI 2014 competition program.

Connectivity inference from neural recording data: Challenges, mathematical bases and research directions

“…Although the concept of an end-to-end artificial neural network is appealing, the performance comparison in Veeriah et al (2015) deserves further consideration. Their reconstructed performances for generalized transfer entropy and partial correlation methods are lower than the performance documented during the First Neural Connectomics Challenge using the same data sets (Guyon et al 2014;Orlandi et al 2014a;Sutera et al 2014).…”

“…If the objective is to infer the graphical structure of the network, what matters is the binary existence/nonexistence of connections. The Area Under the Curve of the Receiver-Operator Characteristic (AUROC) is a popular performance metric used in such a case (Guyon et al 2014;Garofalo et al 2009;Stetter et al 2012). The ROC Curve describes the relationship between the False Positive (FP) Rate ( F P F P +T N ) or Fall-out and the True Positive (TP) Rate ( T P T P +F N ) or Recall at different thresholds.…”

“…The First Neural Connectomics Challenge (Guyon et al 2014) encouraged the development of very diverse solutions. Remarkably, the top three solutions were all model-free methods.…”

“…Some reviews focus on macroscopic connectivity analysis using MRI (Friston 2011;Lang et al 2012;Sporns 2012;. The reports from the First Neural Connectomics Challenge (http://connectomics.chalearn.org) reviews the top-ranked methods for connectivity inference from calcium imaging data (Orlandi et al 2014a;Guyon et al 2014).…”

First Connectomics Challenge: From Imaging to Connectivity

“…In that region all neurons are able to physically reach any other neuron and the network can be considered as a random graph. For the small training datasets we used N = 100 neurons with an average connectivity of k = 12 and varying levels of clustering 1 , from 0.1 to 0.6, and the neurons were placed randomly in a 1 × 1 mm square area (Guyon et al, 2014). For the larger datasets however, we used a different network structure that was never revealed to the participants.…”