We describe an analysis pipeline that classifies cognitive states using a 2-camera set-up for video-based estimation of attentiveness and screen engagement in nonhuman primates performing cognitive tasks. The procedure reconstructs 3D poses from 2D labeled DeepLabCut videos, reconstructs the head/yaw orientation relative to a task screen, and arm/hand/wrist engagements with task objects, to segment behavior into an attentiveness and engagement score.

Set of analysis scripts to reproduce the main finding of the paper Voloh, Oemisch, Womelsdorf (2020). Phase of Firing Coding of Learning Variables in the Fronto-Striatal Network during Feature Learning. Nature Communications. The analysis pipeline applies glm analysis to spike counts at different phases of the oscillation cycle and quantifies the phase modulation depth of the glm-beta coefficients. Statistical approaches are included and described in full in the paper.

Spikes introduce artificial phase lag to the LFP when they are recorded from the same channel. The scripts implement the adaptive spike removal method (ASR) that succeeds removing spike current leakage to the LFP across a broad range of frequencies. The details are described in: Banaie Boroujeni K, Tiesinga P, Womelsdorf T (2019) Adaptive spike-artifact removal from local field potentials uncovers prominent beta and gamma band neuronal synchronization.

Matlab analysis toolbox for detecting and characterizing transient neural burst events. This burst detection toolbox allows using real or surrogate data for burst detection and offers a variety of different burst detection algorithms (magnitude- based, frequency- based, wavelet- based, …). The framework is actively developed and will be updated regularly to allow evaluating the sensitivity and specificity of burst detections.

A Step-by-step How-To describing how to classify different experimental conditions from the dynamic changes in population firing of neurons using time-resolved principal component analysis (‘State-Space Analysis)’.

A step-by-step How-To for applying Tort’s Modulation Index and Eric Maris’ weighted phase locking factor method to identify correlations of phases and amplitudes between LFPs recorded from pairs of electrodes.

A Step-by-step How-To describing how to identify classes (modules) of similar behaving elements (nodes)

This How-to shows how to find out how many reliably distinct clusters of cells can be identified based on a multidimensional matrix of parameters about the cells, including their firing rate, firing variability, bursting propensity, action potential waveform shape, etc.

This How-to shows how to extract parameters from action potential waveforms and how to test for bimodality of the underlying distribution with various measures.

This How-to describes the sequence of analysis required to classify activations across multiple recorded neuronal activities according to internal states or task conditions.