Learning the CONN Toolbox for Functional Connectivity Analysis

 

Fall 2024: November 11-15, 2024. IN PERSON

Resting-state functional connectivity has taken the brain imaging community by storm. Five to ten minutes of MRI data collection from almost any subject or patient can reveal organized systems of activity in the brain that can be used in a wide variety of ways for basic and clinical research, and even to guide non-invasive brain stimulation. The primary tool for analyzing such data has been correlation-based analysis of the connections between seed regions in the brain. A premier software package for such analysis is the CONN toolbox, created by Susan Whitfield-Gabrieli and Alfonso Nieto Castañòn. CONN is an SPM-based toolbox for functional connectivity analyses in the resting state and during tasks. It is ranked as one of the top 10 most visited and downloaded neuroimaging tools and resources at NITRC, with over 40,000 downloads and a very active user community. Also, it has been cited and used in over 400 papers (the second-most cited reference in the Brain Connectivity journal). The program consists of five intensive days of training in the use of this software. Participants will bring their own laptop computers; classroom presentations will be followed by substantial time using the packages with expert guidance available.

 

Software description

CONN is an open-source Matlab/SPM-based cross-platform software for the computation, display, and analysis of functional connectivity Magnetic Resonance Imaging (fcMRI). CONN is used to analyze resting state data (rsfMRI) as well as task-related designs. Processing and analysis steps in CONN include:

  •  
  • Importing DICOM, ANALYZE, and NIfTI functional and anatomical files, either raw or partially/fully preprocessed volumes. Automatic import tools forBIDSdatasets and fMRIPrep outputs
  • Standardized preprocessing pipelines of functional and anatomical volumes powered by SPM12 (including susceptibility distortion correction, motion correction / realignment, slice-timing correction, outlier identification, coregistration, tissue-class segmentation, MNI-normalization, and smoothing)
  • Control of residual physiological and motion artifacts (e.g. scrubbing, aCompCor, ICA-based denoising, Global Regression, band-pass filtering)
  • Integrated quality control procedures and measures (e.g. FC histogram plots, BOLD signal carpetplots, Framewise Displacement, GCOR measures, FC-QC correlations)
  • Multipleconnectivity analyses and measures, including Seed-Based Correlations (SBC), ROI-to-ROI analyses, complex-network analyses, generalized Psycho-Physiological Interaction models (gPPI), Independent Component Analyses (group-ICA), masked ICA, Amplitude of Low-Frequency Fluctuations (ALFF & fALFF), Intrinsic Connectivity (ICC), Local Homogeneity (LCOR), Global Correlations (GCOR), Multivoxel Pattern Analyses (MVPA), and dynamic connectivity analyses (dyn-ICA, sliding-window correlations)
  • Group- and population-level inferencesand models, including ANOVA, regression, longitudinal, experimental, and mixed within- and between-subject designs. Univariate and Multivariate statistics. Control for multiple comparisons using parametric (e.g. Random Field Theory), and non-parametric permutation/randomization techniques (e.g. Threshold Free Cluster Enhancement)

 

Faculty

The faculty for the program includes the originators and developers of the CONN Toolbox and several expert users and teachers to help with the interactive portion of the training: Susan Whitfield-Gabrieli: Software developer and principal speaker at this program. She was a keynote speaker at the 2015 meeting of the Organization for Human Brain Mapping, speaking on the topic of Connectomic Insights into Psychiatric Disorders. Alfonso Nieto Castañòn: A prize-winning Matlab Programmer; software developer of CONN; an active computational neuroscientist researcher. Sheeba Arnold Anteraper: Senior Programmer at the McGovern Institute, and an expert user of CONN. Robert L. Savoy: Organizer of numerous brief, intense training programs at the Martinos Center for Biomedical Imaging of the Massachusetts General Hospital and elsewhere around the world.  

REGISTRATION

 

Sample Schedule

The details and exact order of the schedule below are likely to be modified as the program approaches.   Information from registrants about adding specific topics will be strongly considered and accommodated if possible.  The “Final Application Lecture” may be re-scheduled to a time earlier in the week; perhaps the end of one of the days (probably Wednesday), or after a group dinner on that evening.  

DAY 1

Morning Program Introduction Susan Whitfield-Gabrieli
  Approaches to MultiVariate Analysis and Functional Connectivity Robert Savoy
  MATLAB and its Relationship to The CONN Toolbox faculty
  Orientation to Relevant Software Packages Susan Whitfield-Gabrieli
Lunch    
 Afternoon Resting State Connectivity: History; Examples; Cautionary Stories Susan Whitfield-Gabrieli
  Orientation to the Specifics of the CONN Toolbox     faculty
  Preprocessing: Artifact Prevention, Detection, and Mitigation Susan Whitfield-Gabrieli
  Homework Assignment: Preprocessing  
     

DAY 2

Morning Homework Review: Preprocessing  
  Single Subject Processing in CONN Susan Whitfield-Gabrieli
  Homework Assignment: Single Subject Processing  
Lunch    
Afternoon Single Group Second-Level Design Susan Whitfield-Gabrieli
  Single Group Second-Level Estimation Susan Whitfield-Gabrieli
  Homework Assignment: Single Group  
     

DAY 3

Morning Homework Review: First Level Analysis  
  Multiple Group Second-Level Design Alfonso Nieto Castañón
  Multiple Group Second-Level Estimation Alfonso Nieto Castañón
  Homework: Second Level (NYU Group Data)  
Lunch    
Afternoon Review/Discussion of Basic CONN Toolbox faculty
  Overview of Advanced Features in the CONN Toolbox: Alfonso Nieto Castañón
       ROI-to-ROI Analyses  
       Graph Theory  
       Voxel-to-Voxel Analyses  
       Surface-Based Analysis  
       PPI  
       Dynamic Connectivity  
       Latest CONN Features:  
     

DAY 4

Morning Homework Review: Second Level Analysis  
  Connectome Analysis: ROI-to-ROI Alfonso Nieto Castañón
  Connectome Analysis: Graph Theory Alfonso Nieto Castañón
  Homework: ROI-to-ROI & Graph Theory  
Lunch    
Afternoon Connectome Analysis: Voxel-to-Voxel Alfonso Nieto Castañón
  Connectome Analysis: Surface-Based Alfonso Nieto Castañón
  Homework: Voxel-to-Voxel & Surface-Based  
     

DAY 5

Morning Homework Review: Voxel-to-Voxel & Surface  
  Connectome Analysis: PPI Alfonso Nieto Castañón
  Connectome Analysis: Dynamic Connectivity Alfonso Nieto Castañón
  Connectome Analysis: Latest Features Alfonso Nieto Castañón
  Homework-during-Lunch! PPI & Dynamic Connectivity  
Lunch    
Afternoon Homework Review: PPI & Dynamic Connectivity  
  Final Application Lecture (may be scheduled earlier in week) Susan Whitfield-Gabrieli
  Informal Discussion faculty
     

 

contact fmrivfp@mgh.harvard.edu with any questions