Data Editor

Data Editor

The EMSE Suite DATA EDITOR module may be used as a standalone program to view, analyze, filter, and transform EEG and MEG time series data, and to select intervals or events for further analyses, including event-related potentials, power spectral densities, and time-frequency analysis.  DATA EDITOR includes tools for automation, signal processing, spatial component analysis, topographic mapping, and a powerful framework for statistical nonparametric mapping.  In addition, the data are directly accessible to the SOURCE ESTIMATOR module for source analysis, and to the VISUALIZER module for 3D rendering and animation.  DATA EDITOR takes input in ASCII format or in native modes from many hardware vendors.


Time Domain Analysis

  • Time Series Display.  Four modes are available to view time series data: chart recorder, superimposed, sensor position, and channel array.   Multichannel overlays are available.   The mouse cursor may be used to measure latencies and amplitudes or to select intervals for further processing.
  • Averaging.  Data may be averaged on trigger events selectively and conditionally, with baseline correction and artifact rejection.  Prestimulus and poststimulus intervals are user selectable.  You can automate averaging and grand averaging in batch mode.
  • Event Review Mode.  For datasets with events, such as those obtained from ERP studies, you can rapidly review your data from event to event.  In synchrony with the events, you can inspect component time series obtained from spatial PCA or ICA.
  • Basic Statistics include the standard error of the mean, which you can visualize around an average waveform.
  • Advanced Statistics include permution tests and the generation of p-value maps exportable to the VISUALIZER module.
  • Robust Statistics include trimmed-mean ERPs, median ERPs, and ERPs for any quantile or quantile range.
  • Baseline correction/artifact rejection. Any user-selected interval may be used for baseline correction and/or linear detrending.  Artifacts are rejected when a user-selected amplitude range is exceeded.
  • Filters. Time series filters include zero-phase-shift IIR (low pass, high pass, band pass, notch), polynomial detrend (for DC recordings), median, Savitsky-Golay, and Hilbert envelope rectifier.
  • Filter Pipeline:  Displays the sequence of all spatial and temporal filters you have applied to the time series data.  You can review the properties of all history filters.  New filters are applied on the fly.  Their properties can be changed immediately; they can be enabled/disabled; or they can be deleted.
  • Ocular Artifact Correction applies a spatial filter that removes blink and/or eye movement artifacts while preserving frontal-generated EEG.
  • Waveform operations. Sum, difference, and linear merge.
  • Peak and area measures.  Automate peak detection based on user-selected component identification criteria.  Amplitude, latency, area,  and other measures are obtained across participants, conditions, channels, and components.  Results may be reformatted for input to other statistical packages.
  • Peak detection results overlay. Selected peak results for each component may be visualized as dots on waveforms. Start and stop latencies displayed for each component.
  • Principal components analysis (PCA).  Spatial principal components are calculated by singular value decomposition (SVD) of the cross-channel covariance matrix.  Data may be filtered by component inclusion or exclusion.  Source space dimensionality may be estimated either by inspection of the SVD spectrum, or automatically via a minimum descriptor length statistic.
  • PCA Rotations. Varimax and Promax rotations are available for principal components.
  • Independent Components Analysis (ICA).  InfoMax and SOBI ICA algorithms are provided.  Reconstruction of the data is possible with arbitrary user-selected components.
  • Montages.  Arbitrary EEG reference montages may be created, including: implicit reference, physical reference, digital reference, common average reference, and any bipolar montage.   Scalp current density (SCD) is implemented via the spherical spline surface Laplacian.
  • Simulation. EEG and MEG time series data may be simulated using up to 32 dipoles. Each dipole may be given an independent position and orientation. Time series types include exponential on/off (“PSP”), square, and sine.  Gaussian noise may be added independently to sensors and/or to source signals.

Frequency Domain Analysis

  • Fourier transform.  The Fast Fourier Transform (FFT) implementation of the Discrete Fourier Transform (DFT) provides amplitude and phase spectra for a user-selected time series interval or a single epoch.
  • Power Spectral Density (PSD).  Average power spectra may be obtained for event-related epochs or for consecutive (possibly overlapping) epochs in one or more intervals of data.
  • Robust PSD: Trimmed mean power spectra, median power spectra, and power spectra for arbitrary quantiles are available to provide PSD estimates that are less sensitive to outliers.
  • Cross-Spectral Density and Coherence.  Inter-channel PSD and coherence may be obtained concurrently with intra-channel PSD.
  • Steady state ERPs.  Harmonic analysis of steady state ERPs may be used to extract signal components at the stimulation frequency and its harmonics.

Time-Frequency Analysis

  • Wavelet decomposition.  Combined time-frequency analysis of event-related data may be accomplished using the Morlet wavelet transform.  Results are displayed in a time-frequency-amplitude format, making it easy to distinguish power variations in different frequency domains over a given temporal interval, with optional baseline normalizations.
  • Phase-locked and induced signals are distinguished.
  • Time-frequency functional connectivity is available via interchannel coherence and phase-synchronization.
  • Statistical nonparametric tests may be performed with corrections for multiple comparisons across space, time, and frequency.

Topographic Mapping

  • Pseudocolor and contour topographic maps are available for electric or magnetic field data for time domain, frequency domain, and time-frequency analyses, and for surface Laplacian maps in each case.
  • Maps may be viewed in 2 modes.  ‘Instance’ mode animates the display in step with the cursor location in the time series display window.  ‘Sequence’ mode generates a set of maps corresponding to a selected time series interval.

Filter Pipeline

  • The Filter Pipeline includes both temporal (e.g., IIR) and spatial (e.g., surface Laplacian or Ocular Artifact Correction) filters for use with time series data. The pipeline allows you to turn filters on and off, change their properties, and save or restore the filter history, which is stored automatically along with the data.

Event Pipeline

  • The Event Pipeline permits you to set up all event categories (bins) in advance for automatic ERP processing of new data sets, with updated grand averages.  Results are named automatically and consistently.
  • Arbitrary complexity of conditional events is supported.
  • Event-related frequency domain and time-frequency analyses may be performed concurrently and automatically with ERP analyses.


  • User Interface. A conventional Microsoft Windows style graphical user interface is used throughout. Dynamic linking of graphical display windows permits selections in one window to update others automatically.
  • Number of channels.  256+ channels (EEG and MEG) are supported.

Intermodule Communication

  • Input. LOCATOR may be used for sensor locations.
  • Output. VISUALIZER may be used for topographic mapping on a realistic head surface.

File I/O

  • Input Data time series: ASCII, headerless binary, BioSemi, EGI, BVA, Neuroscan, Sensorium, EDF+, Nihon Kohden, MEG (Elekta Neuromag, 4D Neuroimaging, VSM MedTech), and more (including EMSE native formats).
  • Input Sensor location: LOCATOR, Neuroscan ASCII export, or ASCII.
  • Output.  ASCII files may be input to spreadsheet or other analysis programs.  Time series data with events may be written to the universal EDF+ format or to the legacy Neuroscan CNT format.  EMSE native formats are openly documented.
  • MATLAB code is provided for reading and writing EMSE native time series.

Other I/O

  • Graphics may be transferred to clipboard for use by other graphics programs or for insertion into word processors.


  • Printed. User manual, including file formats.
  • On-line. User manual and tutorial available as context sensitive hypertext.
Sample functionality in Data Editor.
Sample functionality in Data Editor.
Filter pipeline and workspace.
Filter pipeline and workspace.
Time-frequency analysis (inset: Visualizer).
Time-frequency analysis (inset: Visualizer).
Event / segment marking,
Event / segment marking,
Peak / area measures.
Peak / area measures.
Filter pipeline functions.
Filter pipeline functions.
Principal components analysis.
Principal components analysis.
Event Pipeline setup.
Event Pipeline setup.
Event Pipeline Output Selection.
Event Pipeline Output Selection.
Data Editor Module Requirements minimum required  recommended
CPU Intel or AMD, 2-core, 2.0 GHz 4-core or more, 3.0 GHz or faster
RAM 2 GByte 4+ GByte recommended for >32 channels
Hard drive 100 Mbyte for executable 512 Mbyte swap space + additional data storage requirements
Operating System Win XP/Vista/7/8 Win 7
Display SVGA or better
Installation medium CD-ROM
Input modules none