econnectome Tutorial Version 2.0, May 20, 2012 Bin He et al. Biomedical Functional Imaging and Neuroengineering Laboratory Minneapolis, Minnesota, USA
Navigation About econnectome Install econnectome EEG Sample ECoG Sample MEG Sample ERP Sample ADTF Sample Acknowledgement
About econnectome econnectome (Electrophysiological Connectome) is a free and open-source software platform for imaging brain functional connectivity from electrophysiological signals including EEG, ECoG and MEG. The software is developed d by the Biomedical Functional Imaging and Neuroengineering Laboratory at the Univ. of Minnesota, directed by Dr. Bin He, under the support of NIH/NIBIB (RO1 EB006433 and RO1EB007920 to B.H.). The visualization module is jointly developed with Drs. Fabio Babiloni and Laura Astolfi at the Univ. of Rome "La Sapienza".
About econnectome Reference Citations: [1] He B, Dai Y, Astolfi L, Babiloni F, Yuan H, Yang L. econnectome: A MATLAB Toolbox for Mapping and Imaging of Brain Functional Connectivity. Journal of Neuroscience Methods, 195: 261-269, 2011. [2] Dai Y, He B. MEG-based Brain Functional Connectivity ii Analysis Using econnectome. Proc. of 8th International Symposium on Noninvasive Functional Source Imaging of the Brain and Heart and the 8th International Conference on Bioelectromagnetism. 9-11, 2011. [3] Dai Y, Zhang W, Dickens DL, He B. Source Connectivity Analysis from MEG and its Application to Epilepsy Patients. Brain Topography. 25(2):157-166 166, 2012.
Install econnectome MATLAB environment MATLAB Toolboxes: Signal Processing Toolbox Spline Toolbox
Install econnectome Download the econnectome package and unzip the econnectome.zip. Download the ARfit package developed by Drs. Tapio Schneider and Arnold Neumaier, unzip the arfit.zip i and put the package in \econnectome \tools\ Download the Regularization Tools developed by Dr. Per Christian Hansen, unzip the Software.zip and put the package in \econnectome\tools\
Review the econnectome folder ARfit package Regularization package
Add \econnectome to MATLAB search path 1 2 3 4
Start the econnectome software Type econnectome and press the Enter key 1 2 Main Window
Software documentations
Start EEG/ECoG/MEG modules
EEG Sample About the EEG sample data: Simulated scalp EEG data from two sources Simulated two-node connectivity pattern (1 primary source and 1 propagated source) [Wilke et al., 2010] Real ictal data as the primary source waveform Number of EEG channels: 62 Sampling rate: 400 Hz
Import sample data 1 In \econnectome\sample data\ EEG\Simulation\Standard\TXT File folder 2
Display EEG waveforms Change the page EEG Waveforms Change the channel displayed Updated by cursor motion
Potential mapping 1 Make sure all tools are disabled, otherwise other functions may be blocked (similarly in other econnectome modules) Potential map 2 Mouse Left Click
Rotate and zoom tools Note: the tools will block mouse click functions in the 1 waveforms window (similarly in other econnectome modules) Select Tool 2 Rotate, Zoom
Display options 1 Make sure all tools are disabled (similarly in other econnectome modules) 2 Right Click Window Background
Start source imaging gmodule 1 Click Source Imaging to start the module 2 Cortical source imaging window
Display scalp and source maps C. click the buttons to display maps at certain time point Source Map A. Mouse Left Click B. Input time or data point Scalp Map
Select inverse algorithms 1 all tools are disabled 2 right click in the background and select the method in the popup menu
Compute source images 3 click the Imaging button Note: right click in the waveforms window to display the popup menu 4 click the Yes button to compute 1 set start t 2 set end 5 progress will be displayed in computing
Display source image movie 2 click the play button to see imaging g movie in the time interval Interval 1 computation is done
Source display options right click in the background and select options in the popup p p menu
Create Region of Interest (ROI) 1 select the Data Cursor tool 2 locate ROI center on the cortex model 4 set name and radius 3 click the Build New ROI Note: for 3, firstly disable the Data Cursor tool, then right click in the window background to display the popup menu
Compute ROI Time Series 2 the popup ROI Time Series window 1 click the Compute ROI Time Series menu
Compute ROI Time Series 1 right click in the background to display the popup menu 3 ROI waveforms 2 click the Compute ROI Time Series in the popup p p menu
Source Functional Connectivity 1 click the Image ROI Connectivity to popup the DTF/ADTF computation window 4 modify order 3 plot to see automatic order 2 set frequency 5 8 6 7
Connectivity visualization Color and size of the arrow from j-th ROI to i-th ROI code the information flow from j-th ROI to i-th ROI A. Click in the Information Flow Image to update B. Display options C. Connectivity patterns D. Thresholding E. Select frequency enc component/band
Connectivity visualization Click pixel in the Information Flow Image to display the information flow from j-th channel to i-th channel Display options o
Connectivity visualization Color and size of the i-th ball code the outflow from the i-th channel Connectivity patterns
Connectivity patterns Arrow Pattern Outflow Pattern inflow Pattern The source images and their connectivity patterns are achieved. The results agree with the simulated patterns.
ECoG Sample About the ECoG sample data: Simulated nine-node connectivity pattern (1 primary node and 8 propagated nodes) [Wilke et al., 2010] Real ictal data as the primary source waveform Number of ECoG electrodes: 9 Sampling rate: 400 Hz Number of data points: 1200
Import ECoG sample data - electrode positions 1 select the Data Cursor tool 3 click the Generate ECoG Locations (after Data Cursor is disabled) d) 4 set grid size (M*N electrodes, M,N>=2) 2 edit four counterclockwise corners (1 -> 2 -> 3 -> 4) In 2, use Alt + left click to add a new corner illustrated by a datatip, select and move the datatip to change the corner location.
Import ECoG sample data 1 select the Data Cursor tool 3 click the Use Edited d Locations (after Data Cursor is disabled) d) 4 set grid size (M*N electrodes, M,N>=2) In 2, use Alt + left click to add a new 2 edit all electrode locations location illustrated by a datatip, select and (1 -> 2 -> 3 -> 4 -> 5 ) move the datatip to change the location.
Import ECoG sample data 1 click the ECoG Data In \econnectome\sample data\ ECoG\Simulation folder 2 ECoG electrode locations and a map surface through the electrodes over the cortex model are created.
ECoG Potential Mapping B. set display options A. mouse left click
ECoG Time-frequency analysis 2 right click on a channel 1 set time interval ( Epoch Start / Epoch End in the popup menu) 4 3 Note: right click in the waveforms window to display the popup menu
ECoG Functional Connectivity 2 1 set time interval 3 popup the DTF/ADTF computation window
ECoG Functional Connectivity 1 set frequency 3 modify order 2 plot to see automatic order 4 click OK to compute DTF values and see connectivity
ECoG connectivity visualization Operations are similar il to the ones in EEG source connectivity it visualization
ECoG connectivity patterns Arrow Pattern Outflow Pattern Inflow Pattern Connectivity patterns among the 9 channels are achieved. The results agree with the simulated patterns.
MEG Sample About the MEG sample data: Simulated MEG data from two sources Simulated two-node connectivity pattern (1 primary source and 1 propagated source) [Wilke et al., 2010] Real interictal ECoG data as the primary source waveform Number of MEG channels: 148 Sampling rate: 400 Hz
Import sample data 1 In \econnectome\sample data\ MEG\Simulation\TXT File folder 2
MEG Preprocessing Change the page Field map Adjust waveforms MEG Waveforms Operations are similar to the ones in EEG analysis
Butterfly/Global Field Power 1 set time interval ( Epoch Start / Epoch End in the popup menu) 3 2 GFP Butterfly
MEG Source Imaging Module 1 Click Source Imaging to start the module 2 MEG cortical source imaging window
Forward Modeling 1 Click Compute Transfer Matrix 2 Select options 3 4 Calculating Note: the transfer matrix must be computed or imported before source imaging
MEG Source Imaging Operations are similar to the ones in EEG source imaging Source Map Mouse Left Click Field University Map of Minnesota
Compute sources in a time interval 3 click the Imaging button Note: right click in the waveforms window to display the popup menu 4 click the Yes button to compute 1 set start 2 set end 5 progress will be displayed in computing
Create Regions of Interest (ROI) 1 select the Data Cursor tool 2 locate ROI center on the cortex model 4 set name and radius 3 click the Build New ROI Note: for 3, firstly disable the Data Cursor tool, then right click in the window background to display the popup menu
Compute ROI Time Series 2 the popup ROI Time Series window 1 click the Compute ROI Time Series menu
Compute ROI Time Series 1 right click in the background to display the popup menu 3 ROI waveforms 2 click the Compute ROI Time Series in the popup p p menu
ROI Connectivity Estimation 1 click the Image ROI Connectivity to popup the DTF/ADTF computation window 4 modify order 3 plot to see automatic order 2 set frequency 5 6 7 8
Connectivity visualization Operations are similar to the ones in EEG source connectivity visualization
Connectivity patterns Arrow Pattern Outflow Pattern inflow Pattern The source images and connectivity patterns are achieved. The results agree with the simulated patterns.
ERP Sample About the ERP sample data: EEG recordings in a visual task The subject passively viewed a flashing check board at the lower right quadrant Number of EEG electrodes: 61 Sampling rate: 250 Hz Number of data points: 15402 Stimulus marker: S2 Number of stimuli: 122
Import sample data 1 In \econnectome\sample data\ ERP\EEG folder 3 2
ERP analysis Change the page Updated by cursor motion Change displayed channels ERP Waveforms
Re-referencing ② Select channels ③ ①
Filtering ② Set filtering parameters ③ ①
Mask Bad Segments For each segment, firstly left mouse button down, then move, finally left mouse button up
Epoch Extraction ② Set epoch ① Select event ③
Epoch Analysis Keep or reject the present epoch manually Epoch information Waveforms of the present epoch Updated by cursor motion Select epoch
Linear Detrending 2 Performed on all epochs 1 Set time interval ( Interval Start / Interval End in the popup menu)
Baseline Correction Right click to pop up the menu 2 Performed ed on all epochs 1 Set time interval ( Interval Start / Interval End in the popup menu)
ERP Averaging ② Averaged ERP data ①
Analysis of ERP Average ③ Analysis of ERP average ② Waveforms of the ERP average ① Early visual response located in the left posterior region.
ADTF Sample About the ECoG sample data: Simulated four-node connectivity ypattern (1 primary node to a secondary generator) [Wilke et al., 2008] Real ictal data as the primary and secondary source waveforms Number of ECoG electrodes: 4 Sampling rate: 400 Hz Number of data points: 1200
Import Electrode Locations 1 click the Electrode Locations In \econnectome\sample data\ ECoG\Simulation folder 2
Import ECoG data 1 click the ECoG Data In \econnectome\sample data\ ECoG\Simulation folder 2
ADTF Computation 1 2 popup the DTF/ADTF computation window
ADTF Computation 1 set frequency 3 modify order 2 plot to see automatic order 4 Select ADTF 5 click OK to compute ADTF values and see connectivity
Time-varying Connectivity Visualization Operations are similar to the ones in source connectivity visualization Manipulate slider to update connectivity visualization
Time-varying Connectivity Visualization 1 2
Time-varying Connectivity Visualization Primary Arrow Outflow Inflow Secondary Arrow Outflow Inflow Time-varying connectivity patterns among the 4 channels agree with the simulated patterns.
For more information Please refer to the econnectome manual at http://econnectome.umn.edu/
Ak Acknowledgement ld econnectome is developed d by the Biomedical Functional Imaging & Neuroengineering i Laboratory at the under the support from NIH/NIBIB (RO1EB006433 and RO1EB007920 to B.H.). Bin He initiated the project idea and directed the development of software. Yakang Dai implemented all functions and wrote the codes. Lin Yang and Han Yuan participated in daily discussions and tested the software including EEG and ECoG modules. Christopher Wilke participated in discussion on connectivity mapping. The visualization module of the econnectome software is jointly developed with Fabio Babiloni and Laura Astolfi at the University of Rome "La Sapienza". The cortical source imaging methods in the software are implemented based on the Regularization Tools developed by Per Christian Hansen. The ARfit package developed by Tapio Schneider and Arnold Neumaier is used in the DTF computation module of the software. The standard cortex model and BEM models used in the software are constructed based on the Montreal Neurological Institute (MNI) brain. The standard 10-5 system and electrode locations on spherical head model developed by Robert Oostenveld and Peter Praamstra are used in the econnectome.