Home

What is aa, and why use it?

Automatic analysis (aa) is a pipeline system for neuroimaging, written in Matlab. It provides scriptable access to many popular imaging packages such as SPM, FSL, FreeSurfer, and CONN, as well as a variety of third party SPM toolboxes. The primary features include:

• Automatic. Automatic detection of functional, structural, and fieldmap data. Automatic data transfer between processing stages.
• Flexible control. Essential analysis settings taking sensible defaults to assist the new user. Experienced users can change a range of settings. Advanced users can modify almost any processing behaviour or add new capabilities.
• Restartable. If aa halts processing for any reason, it will continue at the stage where it left off when restarted.
• Parallel processing. If multiple cores or a computing cluster is available, aa jobs can easily be distributed across them.
• Analysis paper trail. Unlike a click-through analysis, aa records all parameters used, and allows easy recreation of a dataset from the raw data at a later date.
• Multi-modal. aa supports EEG & MEG as well as fMRI analysis.
• Tool agnostic. aa supports analysis tools equally. Use functionality implemented by SPM, FSL, Freesurfer, or any of the other supported software using the same scripting syntax. Operations from multiple packages can be combined in a single script.
• Extendable. Matlab programmers can write new modules and incorporate them into the processing stream.
• Open source. Code is maintained in a Github repository. Fixes and new features are easily added.

Organization of the Documentation

The aa documentation is organized into three parts. The first is a tutorial-style introduction that covers installation, parameter file setup, and scripting your first pipeline. The tutorial uses the auditory example that is included in the SPM manual (Chapter 30, as of this writing). The second part a collection of reference pages which provide a concise summary of major analysis topics. Finally, a number of appendices are provided, including an FAQ and more.

If you are new to Automatic Analysis, we recommend you work through the tutorial to get you started, and at least skim the FAQ so that you know what topics you may need to revisit as you begin to use aa for your own work. The reference pages provide details you might not require immediately, but can refer to as-needed for information on a specific aa feature.

Links to the documentation sections are provided in the sidebar.