Object-oriented analysis walkthroughs

This series of walkthroughs is designed to illustrate the CANlab interactive analysis tools, and some analysis principles as well. Code to run each walkthrough is included in the CANlab core toolbox or on Neurovault. See the main CANlab intro page for the philosophy behind the interactive analysis approach.

These walkthroughs focus on using CANlab code and are not didactic. For more extensive explanation of fMRI concepts, see the didactic tutorials using Matlab live scripts.

Tip: Use the search page to find a walkthrough by topic.

1. Getting started

Install the toolboxes, load a sample dataset, and learn the basic visualization and reporting commands.

#WalkthroughLive script
1.1Installing tools 
1.2Getting help.mlx
1.3Load a sample dataset 
1.4Basic image visualization 
1.5Loading datasets used in tutorials 
1.6Publishing results reports 
1.7Basic Matlab debugging.mlx
1.8Voxel and image spaces in fmri_data objects.mlx

2. Basic analyses

Group statistics, ROI analysis, masking, surface rendering, and design matrices.

#WalkthroughLive script
2.1Group t-test 
2.2Atlases and ROI analysis 
2.3Masking and writing NIfTI image files 
2.43-D visualization on brain surfaces 
2.5Create and explore 1st-level design matrices 
2.6Regression — interactive analysis 
2.7Interpreting maps with Neurosynth topic similarity and wedge plots 
2.8Annotating a results map (gradients, neurochemical maps, networks, Neurosynth topics) 

3. Multivariate predictive models

Cross-validated multivariate prediction, applying signature patterns, and Bayesian results.

#WalkthroughLive script
3.1Multivariate prediction with continuous outcomes 
3.2Effect sizes with SVMs (in progress) 
3.3Apply a multivariate pattern of interest 
3.4Bayes Factor maps 
3.5Interpreting maps with riverplots 

4. Mediation analysis

Single- and multi-level mediation effect parametric mapping (the M3 toolbox).

#WalkthroughLive script
4.1Mediation and M3 Toolbox basics.mlx
4.2Single-level Mediation Effect Parametric Mapping (fMRI).mlx
4.2bSample MEPM report from publish_mediation_report.m 
4.3Multilevel Mediation Effect Mapping (fMRI).mlx

5. Coordinate-based meta-analysis

MKDA-style meta-analysis of published activation coordinates.

#WalkthroughLive script
5.1MKDA meta-analysis 1 

6. Multivariate Pathway Identification (MPathI) and connectivity

Prepare time-series data and run multivariate pathway identification.

#WalkthroughLive script
6.1Prep time-series data for connectivity.mlx
6.2MPathI pathway dissociation.mlx

7. Visualization extras

Additional brain-rendering and surface-projection examples.

#WalkthroughLive script
7.1More visualization 
7.2Surface rendering 
7.3CANlab single-trials demo 

8. Behavioral data and plots

Working with the canlab_dataset object and generating publication plots.

#WalkthroughLive script
8.1The canlab_dataset object 
8.2Time series and bar plots 
8.3Mixed-effects models with glmfit_multilevel 

9. In progress (advanced)

These tutorials are works in progress and of primary interest to advanced users.

#WalkthroughLive script
9.1Within- and between-subject components of PCR models 
9.2Building ML pipelines with CANlab tools 
9.3Understanding the CANlab atlases object