OpenDvM: Decoding the Visual Mind — A Unified Toolbox for EEG Attention Research
Year: 2026
Type: preprint
Abstract: Multivariate pattern analysis has become a standard complement to univariate event-related potentials (ERPs) in EEG research, yet integrated toolboxes that combine ERP, time-frequency, decoding, and forward-encoding analyses within a single, consistent framework remain largely confined to MATLAB. We introduce OpenDvM (Decoding the Visual Mind), an open-source Python toolbox built on MNE-Python that unifies these four complementary analysis types behind a shared two-step, initialize-then-execute workflow. Developed with the needs of attention researchers specifically in mind, but readily applicable to other domains, OpenDvM computes stimulus- and response-locked ERPs and time-frequency representations, characterizes when and how strongly experimental conditions are represented in multivariate EEG activity via backward decoding models (BDM), and reconstructs the continuous, graded representational geometry underlying spatial attention via forward encoding models using channel tuning functions (CTF). The toolbox standardizes data organization, batch processing, and automated quality-control reporting across all four modules, supporting reproducible analysis pipelines from raw data through group-level statistics. We demonstrate OpenDvM's functionality on an openly shared EEG dataset probing multiple facets of attentional selection, including feature-based selection, spatial attention, and distractor suppression, and provide ten accompanying Jupyter notebook tutorials covering both core and advanced functionality. By lowering the barrier to combining these complementary analytical perspectives, OpenDvM aims to make integrated univariate and multivariate EEG analysis more accessible to researchers without extensive programming backgrounds.
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7212275855