Skip to main content
Automatic Analysis for more efficient and reproducible M/EEG pipelines
Tibor Auer
Presenting author:
Tibor Auer
Automatic Analysis (aa) is a comprehensive MATLAB-based framework for designing, executing, parallelizing, and sharing neuroimaging workflows. aa tightly integrates toolboxes for neuroimaging, and the unique stream concept for data flow in aa allows automatic linkage between processing steps, ‘modules’ in aa term. Finally, most aa modules offer in-depth diagnostics ensuring quality control.
aa has been developed with an initial focus on MRI, including integration with the SPM toolbox. The recent integration of the highly popular EEGLAB and FieldTrip toolboxes further extends aa’s functionalities to the M/EEG community. EEGLAB offers an exceptional array of tools for pre-processing M/EEG data and FieldTrip excels in analysing various features of the data at both sensor and source levels. An example workflow is also provided based on the openly available MPI Leipzig Mind-Brain-Body dataset, which can cover many common electrophysiological use cases.
The integration of larger packages can result in namespace and dependency ambiguities. It is resolved by means of the recently developed tool management framework, which can be also used to facilitate the integration of tools from other languages, including Python, into aa.
These new developments improve the efficiency and versatility of aa providing human-readable building blocks for processing and analysis batteries for a growing range of multimodal neuroimaging workflows without compromising transparency and flexibility.