The Decision Decoding ToolBOX (DDTBOX) – A Multivariate Pattern Analysis Toolbox for Event-Related Potentials

Abstract

In recent years, neuroimaging research in cognitive neuroscience has increasingly used multivariate pattern analysis (MVPA) to investigate higher cognitive functions. Here we present DDTBOX, an open-source MVPA toolbox for electroencephalography (EEG) data. DDTBOX runs under MATLAB and is well integrated with the EEGLAB/ERPLAB and Fieldtrip toolboxes (Delorme and Makeig 2004; Lopez-Calderon and Luck 2014; Oostenveld et al. 2011). It trains support vector machines (SVMs) on patterns of event-related potential (ERP) amplitude data, following or preceding an event of interest, for classification or regression of experimental variables. These amplitude patterns can be extracted across space/electrodes (spatial decoding), time (temporal decoding), or both (spatiotemporal decoding). DDTBOX can also extract SVM feature weights, generate empirical chance distributions based on shuffled-labels decoding for group-level statistical testing, provide estimates of the prevalence of decodable information in the population, and perform a variety of corrections for multiple comparisons. It also includes plotting functions for single subject and group results. DDTBOX complements conventional analyses of ERP components, as subtle multivariate patterns can be detected that would be overlooked in standard analyses. It further allows for a more explorative search for information when no ERP component is known to be specifically linked to a cognitive process of interest. In summary, DDTBOX is an easy-to-use and open-source toolbox that allows for characterising the time-course of information related to various perceptual and cognitive processes. It can be applied to data from a large number of experimental paradigms and could therefore be a valuable tool for the neuroimaging community.

Spatio-temporal dynamics of face perception

“…We used searchlight-based (Kriegeskorte, Goebel, & Bandettini, 2006) decoding with a radius of 10 mm (isotropic), and with default settings of TDT; L2-norm SVM with regularizing parameter C = 1 running in LIBSVM (Chang & Lin, 2011). For EEG data, we used DDTBOX (Bode, Feuerriegel, Bennett, & Alday, 2018) to decode ERPs (event related potentials) averaged over 6 repeated trials (Grootswagers, Wardle, & Carlson, 2017;Isik, Meyers, Leibo, & Poggio, 2014). We used spatiotemporal decoding with 2 consecutive time points, resulting in non-overlapping windows of 10 ms and all 64 channels, resulting in 2*64 = 128 features.…”

Category-selective human brain processes elicited in fast periodic visual stimulation streams are immune to temporal predictability

“…Unlike the permutation testing described above, MVPA does not rely on single channel waveform data (i.e., a single averaged channel for each ROI), but considers fine differences in the pattern of activity elicited by periodic and nonperiodic faces across all channels simultaneously. We used an MPVA procedure for spatiotemporal decoding in EEG introduced by Bode and colleagues (The Decision Decoding Toolbox, available at http://ddtbox.github.io/DDTBOX/; Bode and Stahl 2014;Bode et al 2016). Using the lowpass filtered data in which the image presentation frequency of 12 Hz was preserved (i.e., not yet filtered out), we extracted epochs corresponding to periodic faces and nonperiodic faces (-166.66ms to 500ms) for each participant.…”

“…Collated across participants, classification accuracies for the permutated data form a reference distribution at each time point, the mean of which we compared to the real classification accuracy using a paired t-test (34 time points, significance criterion = p < .05, Bonferroni corrected). This method is considered stricter than testing against theoretical chance accuracy of 50% (Bode and Stahl 2014;Bode et al 2016).…”

From ERPs to MVPA Using the Amsterdam Decoding and Modeling Toolbox (ADAM)

“…Thus although it aligns with the growing movement to promote open source in cognitive neuroscience (Gleeson et al, 2017 ), it does not necessarily provide the latest and greatest in multivariate analysis. For those already comfortable with programming and/or multivariate analysis, a number of more versatile alternatives for time-series based MVPA exist which have larger development teams, notably CoSMoMVPA ( http://www.cosmomvpa.org , MATLAB) (Oosterhof et al, 2016 ), the Neural Decoding Toolbox ( http://www.readout.info , MATLAB) (Meyers, 2013 ), the Decision Decoding Toolbox ( http://ddtbox.github.io/DDTBOX , MATLAB) (Bode et al, 2018 ), MNE ( http://www.martinos.org/mne/stable/manual/decoding.html , Python) (Gramfort et al, 2014 ) and the PyMVPA toolbox ( http://www.pymvpa.org , Python) (Hanke et al, 2009 ). Yet, for those wanting to dip their toes into multivariate waters for the first time, ADAM could be a great start.…”