fMRI Design and Analysis - New Haven
fMRI Design and Analysis is a 5-day workshop designed for investigators having familiarity with the fundamental principles of fMRI measurement and design, with some experience in functional neuroimaging data acquisition and analysis.
The strongly interactive workshop will focus on using SPM12 and its extensions for preprocessing, statistical modeling and visualization of data associated with a range of basic and clinical fMRI experimental designs. The fundamentals of analyzing both task-related and resting state data will be covered. While the primary emphasis will involve using the core SPM12 programs for these purposes, there will also be extensive discussion of additional software tools, including the CONN Toolbox. Some of these tools facilitate fMRI quality assurance through artifact detection and mitigation at various analysis stages. Other tools support a variety of data visualization methods, including MRIcron, xjView, and CAT12. The workshop will cover the entire acquisition and processing pipeline from collecting raw fMRI data to group level inference and will address the modeling consequences of recent advances in acquisition and modeling techniques.
This course is appropriate for investigators who have had some experience with previous versions of SPM, FSL or AFNI.
As mastery of the topics presented early in the week is needed to maximally benefit from the material presented later in the week, we encourage all participants to arrange their schedules to make it possible to attend all the classes. In addition, please reserve 1-2 hours each evening to complete the homework assignments.
As the presentations will include demonstrations and tutorials utilizing SPM12, participants are expected to bring a laptop with MATLAB, SPM12 and MRIcron already installed. Detailed setup instructions can be found here.
The course will be sponsored by the Garrison Lab and held at Yale Medical School, in the Jane Elliot Hope Building, Room H103 in New Haven, CT from February 6-10, 2017. The 315 Cedar Street entrance is accessible only to those who have medical school key cards. If you are planning a meeting in this building with non-medical school participants, please instruct them to enter at 333 Cedar Street and proceed along the C-Wing to the Hope Building.
The fee for the program is $1500. Reduced rates of $1250 for post-doctoral fellows and $1000 for undergraduate, graduate or medical students are available.
Accommodations: There are rooms available nearby:
There is a free Yale shuttle that is great for getting around town:
Download the app here: http://yale.transloc.com/info/mobile
Air Rights Garage at 66 York Street approx. $15/day
Howard Avenue Garage at 790 Howard Avenue: open 6-9 pm, approx. $15/day
Registration information for this program may be found here.
Payment can be made using PAYPAL here.
The course is organized by Tom Zeffiro and Kathleen Garrison. To maintain a high degree of student/faculty interaction, the enrollment will be capped at 30 participants.
Questions about the course content should be directed to
Process automation using batch processing in SPM
fMRI data acquisition optimization
Special considerations when collecting multiband EPI data
Geometric distortion correction using field maps
Physiological denoising techniques
-Control of physiological/movement confounds using CompCor
-Artifact detection and elimination using the ART Toolbox
First level modeling for task designs
First level modeling for resting state MRI Using the CONN Toolbox
-Independent component analysis
First level modeling of psychophysiological interactions
Second level modeling from basic to complex
-Repeated measures models
-Mixed within and between factor designs
-Hierarchical linear models
Inference and critical threshold determination - recent controversies and their resolution
Classical and Bayesian inference
Anatomical labeling using atlases
Sample size estimation for neuroimaging studies
Guidelines for reporting fMRI results
Addressing rigor and reproducibility in grant applications
CONN: The Connectivity Toolbox is an SPM toolbox designed for the computation, display, and analysis of functional connectivity measures in fMRI data. Connectivity measures available in CONN include seed-to-voxel connectivity maps, ROI-to-ROI connectivity matrices, graph properties of connectivity networks, and voxel-to-voxel measures, including intrinsic connectivity, local correlation maps, and others. CONN can be used to model data from both resting state fMRI and task-related designs.
- Understand the basic organization of the SPM and FSL GUIs
- Understand the organization of the SPM "toolbox"
- Be able to construct batch processing scripts for preprocessing and statistical modeling
- Understand the basic fMRI data preprocessing steps
- Be able to construct a preprocessing sequence including slice time correction, realignment, and spatial filtering using both SPM and FSL
- Understand the origins of the artifacts most commonly encountered in fMRI datasets
- Be able to utilize explore an fMRI dataset for artifacts and effect repairs as needed
- Understand the basic fMRI single subject experimental design types
- Be able to implement statistical analysis procedures for the basic single subject fMRI designs in SPM and FSL
- Understand the basic fMRI single group experimental design types
- Be able to implement statistical analysis procedures for the basic single group fMRI designs in SPM and FSL
- Understand the basic fMRI multiple group experimental design types
- Be able to implement statistical analysis procedures for the basic multiple group fMRI designs in SPM and FSL
- Understand the process of incorporation of covariates in fMRI experimental designs
- Be able to construct and estimate statistical models involving covariates in SPM and FSL
- Be able to utilize the RIC Talairach Daemon for region labeling
- Be able to utilize the SPM Anatomy toolbox for region labeling
- Be able to utilize the FreeSurfer for region labeling
- Be able to export SPM statistical maps to FreeSurfer to visualize results on the cortical surface
- Be able to use MRIcron for visualization, including volume rendering
- Be able to use xjView for visualization and labeling
- SPM installation and setup
- MATLAB 7 Getting Started Guide
- MATLAB Tutorials
- Introduction to SPM
- SPM12 new features
- SPM Interface overview
- File format conversion
- Optimizing fMRI acquisition
- Artifact identification and mitigation
- The general linear model
- First level model specification
- Second level model specification
- Conjunction analysis
- Anatomical labeling
- Batch processing
- Further reading