Tutorials & E-Library

This tutorial will guide you through the simulation of one neuron in NEST.

This section covers how to use NEST to construct structured networks.

In this section we look at using synapse models to connect neurons.

In this section we look at creating and parameterising batches of neurons, and connecting them.

In this section we cover the first steps in using PyNEST to simulate neuronal networks.

This lecture focuses on higher-level simulation scenarios using stimulation protocols. We demonstrate how to build stimulation patterns in TVB, and use them in a simulation to induced activity dissipating into experimentally known resting-state networks in human and mouse brain, a well as to obtain EEG recordings reproducing empirical findings of other researchers.

This workflow allows you to segment images based on boundary information. Given a boundary probability map, it breaks the image up into superpixels and then merges them to recover segments limited by closed surfaces (no dangling edges). The main algorithm, known as multicut or correlation clustering, was presented in a paper by B. Andres. Its applications to biological image analysis can be found in, for example, connectomics data, bright field and phase contrast images, light microscopy tissue images or any other kind of imaging which uses membrane staining.

Brain Imaging Data Structure (BIDS) is a standardized way to organize and describe neuroimaging, electrophysiological and behavioral data. This organization has been adopted by a multitude of neuroscience labs around the world to facilitate sharing and analysis
(https://www.nature.com/articles/sdata201644).
BIDS Manager software described in this manual is a tool that allows various users to easily import and explore databases in BIDS format.
This document will guide you to import your data in BIDS format and explore your BIDS Dataset.
Manually driven processes for data storing can lead to human errors, which cannot be tolerated in the context of a research/clinical datasets. BIDS manager offers a secure system to import and structure
subject and patient datasets.

elephant.parallel module provides a simple interface to parallelize multiple calls to any user-specified function. The typical use case is calling a function many times with different parameters.

The Granger causality is a method to determine functional connectivity between time-series using autoregressive modelling. In the simpliest pairwise Granger causality case for signals X and Y the data are modelled as autoregressive processes. Each of these processes has two representations. The first representation contains the history of the signal X itself and a prediction error (or noise a.k.a. residual), whereas the second also incorporates the history of the other signal.

If inclusion of the history of Y next to the history of X into X model reduces the prediction error compared to just the history of X alone, Y is said to Granger cause X. The same can be done by interchanging the signals to determine if X Granger causes Y.

Conditional Granger causality can be used to further investigate this functional connectivity. Given signals X, Y and Z, we find that Y Granger causes X, but we want to test if this causality is mediated through Z. We can use Z as a condition for the aforementioned Granger causality.

In order to illustrate the function of time-domain Granger causality we will be using examples from Ding et al. (2006) chapter. Specifically, we will have two cases of three signals. In the first case we will have indirect connectivity only, whereas in the second case both direct and indirect connectivities will be present.

The tutorial demonstrates a method of finding patterns of synchronous spike times (synfire chains) which cannot be revealed by measuring neuronal firing rates only.

In this tutorial, we use 50 neurons per group in 10 successive groups of a synfire chain embedded in a balanced network simulation.

SPADE is a method to detect repeated spatio-temporal activity patterns in parallel spike train data that occur in excess to chance expectation. In this tutorial, we will use SPADE to detect the simplest type of such patterns, synchronous events that are found across a subset of the neurons considered (i.e., patterns that do not exhibit a temporal extent). We will demonstrate the method on stochastic data in which we control the patterns statistics.