Differences between revisions 92 and 93

Course Material for COGNESTIC 2024

The Cognitive Neuroimaging Skills Training In Cambridge (COGNESTIC) is a 2-week course that provides researchers with training in state-of-the-art methods for reproducible and open neuroimaging analysis and related methods. You can find more information on the COGNESTIC webpage.

Below you will find documents, videos and web links that will be used for the course or can be used for preparation.

Software Installation Instructions

Access to much of the COGNESTIC-23 materials is available via Virtual Machine (VM). You will need at least 70GB of free space on your local hard drive, and at least 4GB of RAM. For instructions on how to install and set up the Cognestic23 VM see section ‘1 COGNESTIC Virtual Machine (full hands-on)’ for instructions. Data for the Structural MRI and Diffusion MRI are located inside the VM.

Full installation instructions can be found here. The installation can take some time (potentially more than an hour, depending on your download speed), so please reserve some time for this ahead of the event.

Essential

You will find the course easier if you can study as much of the material below in advance (e.g, many of the videos below give the theory to the examples we will work through in the course).

(sessions below should be ordered as they will be in course, but just did my ones for an example)

Background to Open Science Rik Henson
Viewing	Open Cognitive Neuroscience

Network Analysis Rik Henson
Viewing	Introduction to Network Neuroscience

Structural MRI I and II Marta Correia
Viewing	Introduction to MRI Physics and image contrast Slides

Diffusion MRI I - Preprocessing, Model Fitting and Group Analysis Marta Correia
Viewing	Introduction to Diffusion MRI - Part I Slides

Diffusion MRI II - Tractography and the Anatomical Connectome Marta Correia
Viewing	Introduction to Diffusion MRI - Part II Slides

fMRI I - Data Management Dace Apšvalka
Viewing	fMRI Data Structure & Terminology (6:47) Brain imaging data structure (11:07)

fMRI II - Pre-processing Dace Apšvalka
Viewing	fMRI Artifacts and Noise (11:57) Pre-processing I (10:17) Pre-processing II (7:42)

fMRI III - Analysis Dace Apšvalka
Viewing	GLM applied to fMRI (11:21) Model Building – conditions and contrasts (11:48) Model Building - nuisance variables (13:58) Multiple Comparisons (9:03) Group-level Analysis I (7:05)

EEG/MEG I – Measurement and Pre-processing Olaf Hauk
Viewing	1. Overview of EEG/MEG data processing from raw data to source estimates Event-related paradigm, sample dataset, power spectrum, pre-processing, artefact correction, epoching and averaging, visualization, source estimation. 2. The generation of EEG/MEG signals Dipole sources, volume currents, sensor types (EEG, magnetometers, gradiometers) and their leadfields. 3. Frequency and temporal filtering of EEG/MEG data Frequency spectrum, temporal smoothing, relationship between frequency and time domain, filters (low-/high-/band-pass, Notch), aliasing, decibels. 4. Differential sensitivity of EEG and MEG Volume conduction, sensor types and their leadfields, sensitivity maps, dipoles vs spatially extended sources. 5. Event-related potentials and fields Averaging, evoked and induced activity, number of trials, artefact rejection, parametric designs, regression. Fore more on this topic see here.

EEG/MEG II – Head Modelling and Source Estimation Olaf Hauk
Viewing	1. The EEG/MEG forward model Basic formulation of the EEG/MEG forward problem, linear equation, basics of head modelling, examples of sensory evoked responses. 2. The EEG/MEG inverse problem Non-uniqueness, under-determinedness, examples of non-uniqueness, source estimates for sensorily evoked activity. 3. The spatial resolution of linear EEG/MEG source estimation Leakage and blurring, resolution matrix, point-spread functions (PSFs), cross-talk functions (CTFs), examples of PSFs and CTFs, regions-of-interest for source estimation. 4. Noise and regularisation in EEG/MEG source estimates Over- and under-fitting, smoothing, regularisation parameter, data whitening, noise covariance matrix. Fore more on this topic see here.

EEG/MEG III – Time-Frequency and Functional Connectivity Analysis Olaf Hauk
Viewing	1. Frequency spectra and the Fourier analysis Periodic basis functions, Fourier Decomposition, frequency spectrum, Nyquist Theorem, steady state response. 2. Time-frequency analysis and wavelets Fourier analysis, wavelets, trade-off between time and frequency resolution, wavelets, number of cycles, evoked and induced activity, beta bursts. 3. The basics of functional connectivity methods Types of connectivity, amplitude envelope correlation, resting state analysis, Hilbert envelope, phase-locking, coherence, SNR bias, time-resolved connectivity. Fore more on this topic see here.

EEG/MEG IV – Further Topics and BIDS Olaf Hauk & Máté Aller
Viewing	1. Primer on group statistics for EEG/MEG data Regions-of-interest (ROI) analysis, multiple comparison problem, cluster-based permutation tests, problems estimating cluster extent, MNE-Python tutorial. 2. Primer on decoding and RSA with EEG/MEG data Basics of linear decoding, temporal generalisation, interpreting decoding weights, back-projection, representational similarity analysis (RSA). 3. Primer on multimodal integration Types of neural “activity”, differential sensitivity of EEG/MEG vs fMRI, source weighting and priors, estimating deep sources with EEG/MEG. Fore more on this topic see here.

Additional Extra

If you want additional background, consider some of the below:

Background to Open Science Rik Henson
Websites	OSF UKRN BIDS
Reading	Munafo et al, 2017, problems in science Button et al, 2013, power in neuroscience Poldrack et al, 2017, reproducible neuroimaging Marek et al, 2022, power in neuroimaging association studies
Viewing	Statistical power in neuroimaging PayWall: open access Comedian's Perspective on science and media
Slides	Open Science Talk Slides

Brain Network Analysis Rik Henson
Software	Python 3.7+, nxviz, python-louvain
Datasets
Reading	- (Review article) Bullmore, E., Sporns, O. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci 10, 186–198 (2009). https://doi.org/10.1038/nrn2575 - (Textbook reference for more information) Alex Fornito, Andrew Zalesky, and Edward Bullmore. Fundamentals of brain network analysis. Academic press, 2016.
Viewing	Understanding your brain as a network and as art by Prof. Dani Bassett.
Slides	https://github.com/isebenius/COGNESTIC_network_analysis/ Slides

Structural MRI I - Voxel-based morphometry Marta Correia
Software	FSL
Suggested reading	Introduction to GLM for structural MRI analysis Good et al, 2001, A VBM study of ageing Smith et al, 2004, Structural MRI analysis in FSL

Structural MRI II - Surface-based analyses Marta Correia
Software	Freesurfer
Suggested reading	Dale et al, 1999, Cortical surface-based analysis I Fischl et al, 1999, Cortical surface-based analysis II
Suggested viewing	Using the command line

Diffusion MRI I - Preprocessing, Model Fitting and Group Analysis Marta Correia
Software	dipy, FSL
Suggested reading	FSL Diffusion Toolbox Wiki Le Bihan et al, 2015, What water tells us about biological tissues Soares et al, 2013, A short guide to Diffusion Tensor Imaging Smith et al, 2006, Tract-based spatial statistics (TBSS)

Diffusion MRI II - Tractography and the Anatomical Connectome Marta Correia
Software	dipy
Suggested reading	MR Diffusion Tractography

fMRI I - Data Management Dace Apšvalka
Software	HeudiConv, PyBIDS, NiBabel, Nilearn
Websites	Brain Imaging Data Structure BIDS Starter Kit BIDS Specification v1.9.0
Suggested reading	The brain imaging data structure (BIDS), Gorgolewski et al., 2016 The past, present, and future of the brain imaging data structure (BIDS), Poldrack et al., 2024

fMRI II - Pre-processing Dace Apšvalka
Software	MRIQC, fMRIprep, NiPype
Suggested reading	Functional Magnetic Resonance Imaging Methods, Chen & Glover, 2015 Quality control in functional MRI studies with MRIQC and fMRIPrep, Provins et al., 2023 fMRIPrep: a robust preprocessing pipeline for functional MRI, Esteban et al., 2018 Nipype: a flexible, lightweight and extensible neuroimaging data processing framework in Python, Gorgolewski et al., 2011

fMRI III - Analysis Dace Apšvalka
Software	Nilearn
Suggested reading	The Statistical Analysis of fMRI Data, Lindquist, 2008 Controlling the familywise error rate in functional neuroimaging: a comparative review, Nichols & Hayasaka, 2003 Analysis of task-based functional MRI data preprocessed with fMRIPrep, Esteban et al., 2020 Guidelines for reporting an fMRI study, Poldrack et al., 2008
Suggested viewing	Model Building - temporal basis sets (11:08) GLM Estimation (9:11) Noise Models- AR models (9:57) Inference- Contrasts and t-tests (11:05) Multiple Comparisons by Martin Lindquist and Tor Wager (9:03) FWER Correction (16:11) FDR Correction (5:25) More about multiple comparisons (14:39)

EEG/MEG I – Measurement and Pre-processing Olaf Hauk
Software and datasets	This will be part of a download that will become available later. MNE-Python software homepage MNE stand-alone installation instructions for COGNESTIC Jupyter script to download sample datasets in MNE-Python
Essential and suggested viewing	0. Overview of EEG/MEG data processing from raw data to source estimates Event-related paradigm, sample dataset, power spectrum, pre-processing, artefact correction, epoching and averaging, visualization, source estimation. 1. A brief history of timing A brief overview of the history of bioelectromagnetism, EEG and MEG. 2. The generation of EEG/MEG signals Dipole sources, volume currents, sensor types (EEG, magnetometers, gradiometers) and their leadfields. 3. Basics of EEG/MEG artefact correction Physiological and non-physiological artefacts, data decompositions, frequency/temporal/spatial filtering. 4. Frequency and temporal filtering of EEG/MEG data Frequency spectrum, temporal smoothing, relationship between frequency and time domain, filters (low-/high-/band-pass, Notch), aliasing, decibels. 5. Topographical artefact correction of EEG/MEG data Independent Component Analysis (ICA), Signal Space Projection (SSP), eye movement and heart beat artefacts. 6. Maxfiltering of MEG data Signal Space Separation, options of Maxfilter software (e.g. movement compensation). 7. Differential sensitivity of EEG and MEG Volume conduction, sensor types and their leadfields, sensitivity maps, dipoles vs spatially extended sources. 8. Event-related potentials and fields Averaging, evoked and induced activity, number of trials, artefact rejection, parametric designs, regression. + Origin, significance, and interpretation of EEG (Michael X Cohen) + Analysing MEG data with MNE-Python and its ecosystem (Alex Gramfort) + List of EEG/MEG lectures MNE-Python tutorials: Overview of MNE-Python processing pipeline from preprocessing to source estimation Preprocessing
Suggested reading	Digitial Filtering Filtering How To Maxwell Filtering General EEG/MEG Literature
Slides and scripts	TBA

EEG/MEG II – Head Modelling and Source Estimation Olaf Hauk
Software and datasets	This will be part of a download that will become available later. MNE-Python software homepage MNE stand-alone installation instructions for COGNESTIC Jupyter script to download sample datasets in MNE-Python
Essential and suggested viewing	0. Overview of EEG/MEG data processing from raw data to source estimates Event-related paradigm, sample dataset, power spectrum, pre-processing, artefact correction, epoching and averaging, visualization, source estimation. 1. The EEG/MEG forward model Basic formulation of the EEG/MEG forward problem, linear equation, basics of head modelling, examples of sensory evoked responses. 2. Source spaces for EEG/MEG source estimation Cortical surface, volumetric source space, spatial sampling, spatial normalisation, subcortical areas, source orientation. 3. Head models for EEG/MEG source estimation <<BR>>Volume conduction, Boundary Element Method (BEM), Finite Element Method (FEM), head model accuracy. 4. The EEG/MEG inverse problem Non-uniqueness, under-determinedness, examples of non-uniqueness, source estimates for sensorily evoked activity. 5. The spatial resolution of linear EEG/MEG source estimation Leakage and blurring, resolution matrix, point-spread functions (PSFs), cross-talk functions (CTFs), examples of PSFs and CTFs, regions-of-interest for source estimation. 6. Comparison of spatial resolution for linear EEG/MEG source estimation methods Point-spread functions (PSFs), cross-talk functions (CTFs), resolution metrics (localisation error, spatial deviation), combination of EEG and MEG, PSFs and CTFs for minimum-norm type methods and beamformers, comparison of resolution metrics for minimum-norm type methods and beamformers. 7. Noise and regularisation in EEG/MEG source estimates Over- and under-fitting, smoothing, regularisation parameter, data whitening, noise covariance matrix. + List of EEG/MEG lectures MNE-Python Tutorials: Forward Models and Source Spaces Source Estimation
Suggested reading	Linear source estimation and spatial resolution Comparison of common head models (e.g. BEM) Guidelines for head modelling (incl. FEM) General EEG/MEG Literature
Slides and scripts	TBA

EEG/MEG III – Time-Frequency and Functional Connectivity Analysis Olaf Hauk
Software and datasets	This will be part of a download that will become available later. MNE-Python software homepage MNE stand-alone installation instructions for COGNESTIC Jupyter script to download sample datasets in MNE-Python
Essential and suggested viewing	1. The basics of signals in the frequency domain Oscillations, periodic signals, sine and cosine, polar representation, complex numbers. 2. Frequency spectra and the Fourier analysis Periodic basis functions, Fourier Decomposition, frequency spectrum, Nyquist Theorem, steady state response. 3. Time-frequency analysis and wavelets Fourier analysis, wavelets, trade-off between time and frequency resolution, wavelets, number of cycles, evoked and induced activity, beta bursts. 4. The basics of functional connectivity methods Types of connectivity, amplitude envelope correlation, resting state analysis, Hilbert envelope, phase-locking, coherence, SNR bias, time-resolved connectivity. 5. Spatial resolution (leakage) and connectivity Connectivity in sensor and source space, point-spread and cross-talk, (non-)zero-lag signals, orthogonalisation, imaginary part of coherency, source space parcellations.
Suggested reading	Tutorial on Functional Connectivity Analyzing Neural Time Series Data General EEG/MEG Literature
Slides and scripts	TBA

EEG/MEG IV – Statistics and BIDS Olaf Hauk & Máté Aller
Software	MNE-Python MNE Installation for Cognestic
Datasets	Sample dataset in MNE-Python. Tutorials MNE Installation for Cognestic M/EEG combined dataset Download Datasets
Suggested reading	Estimating subcortical sources from EEG/MEG Tutorial on converting MEG data to BIDS format Example using MNE-BIDS-Pipeline for processing combined M/EEG data
Suggested viewing	Talk on Multimodal Integration
Slides and scripts	Notebooks Exercises Slides1 Slides2 Notebooks mne-bids-pipeline Slides mne-bids-pipeline

Primer on Python Kshipra Gurunandan
Software	Python, Pandas, NumPy, Matplotlib, Seaborn
Datasets	Wakeman Multimodal
Useful references	Python concepts with examples, Quick reference, Cheatsheets
Slides and scripts	To be added

MVPA/RSA I Daniel Mitchell
Software	Python 3.7+, including numpy, matplotlib, & scikit-learn. (To visualise MRI data, you can use your software of choice, although for nifti format data you might like to consider MRIcroN or MRIcroGL.)
Datasets
Reading	Mur et al. (2009) Revealing representational content with pattern-information fMRI--an introductory guide
Viewing	Excellent presentations from Martin Hebart's MVPA course, on: Introduction to MVPA Introduction to classification. I've suggested these two, but the others are worth a look too. If the links don't work, download from here and here.
Slides and scripts	To be added nearer the time.

MVPA/RSA II Daniel Mitchell & Máté Aller
Software	Python implementation of the RSA Toolbox: Version 3.0
Datasets	Example data included with RSA toolbox
Reading	Kriegeskorte et al. (2008) Representational similarity analysis - connecting the branches of systems neuroscience Kriegeskorte & Kievit (2013) Representational geometry: integrating cognition, computation, and the brain Nili et al. (2014) A toolbox for representational similarity analysis Schutt et al. (2023) Statistical inference on representational geometries EEG/MEG: Tutorial on EEG/MEG decoding Temporal Generalization Interpretation of Weight Vectors
Viewing	Martin Hebart's lecture on RSA. If the link fails, download from here.
Slides and scripts	We will demo the RSA toolbox using the jupyter notebooks in the "demos" folder of the toolbox. EEGMEG Notebooks EEG/MEG Slides

***Below not updated yet

fMRI Connectivity Petar Raykov
Software	Nilearn Python
Datasets	movie dataset
Suggested reading	Resting-state functional Connectivity Learning and comparing functional connectomes across subjects
Suggested viewing	fMRI Functional Connectivity in fMRI Overview of Effective Connectivity (not covered in person)
Tutorial slides and scripts	Functional Connectivity Nilearn Practical DCM tutorial in SPM (not covered in-person)

-  ⇤ ← Revision 92 as of 2024-07-15 16:32:39 → 
  Size: 37251
  Editor: DaceApsvalka
  Comment:
+   ← Revision 93 as of 2024-07-15 16:38:34 → ⇥
  Size: 35961
  Editor: DaceApsvalka
  Comment:
-Deletions are marked like this.
+Additions are marked like this.
 Line 118:
-||__Datasets__ ||             ||
+||__Datasets__ || ||
 Line 178:
-||<10%>__Software__ ||[[http://nilearn.github.io/stable/index.html|Nilearn]] ||
||__Suggested reading__ ||[[https://doi.org/10.1002/hbm.460020402|Friston et al. (1994), Statistical parametric maps in functional imaging: A general linear approach]]<<BR>>[[https://doi.org/10.1016/j.neuroimage.2012.01.133|Poline & Brett (2012), Poline, J. B., & Brett, M. (2012). The general linear model and fMRI: does love last forever?]]<<BR>>[[https://doi.org/10.3389/fnhum.2011.00028|Monti (2011), Statistical analysis of fMRI time-series: a critical review of the GLM approach]]<<BR>>[[https://doi.org/10.1191/0962280203sm341ra|Nichols & Hayasaka (2003), Controlling the familywise error rate in functional neuroimaging: a comparative review]]<<BR>>[[https://doi.org/10.1016/j.neuroimage.2008.05.021|Chumbley & Friston (2009), False discovery rate revisited: FDR and topological inference using Gaussian random fields]]<<BR>>[[https://doi.org/10.1016/j.neuroimage.2013.12.058|Woo et al. (2014), Cluster-extent based thresholding in fMRI analyses: Pitfalls and recommendations]]<<BR>>[[https://doi.org/10.1214/09-STS282|Lindquist (2008), The Statistical Analysis of fMRI Data]] ||
||__Suggested viewing__ ||[[https://youtu.be/GDkLQuV4he4|The General Linear Model]] by Martin Lindquist and Tor Wager (12:24)<<BR>>[[https://www.youtube.com/watch?v=OyLKMb9FNhg|GLM applied to fMRI]] by Martin Lindquist and Tor Wager (11:21)<<BR>>[[https://www.youtube.com/watch?v=7MibM1ATai4|Model Building I – conditions and contrasts]] by Martin Lindquist and Tor Wager (11:48)<<BR>>[[https://www.youtube.com/watch?v=YfeMIcDWwko|Model Building II – temporal basis sets]] by Martin Lindquist and Tor Wager (11:08)<<BR>>[[https://www.youtube.com/watch?v=DEtwsFdFwYc|Model Building III- nuisance variables]] by Martin Lindquist and Tor Wager (13:58)<<BR>>[[https://www.youtube.com/watch?v=Ab-5AbJ8gAs|GLM Estimation]] by Martin Lindquist and Tor Wager (9:11)<<BR>>[[https://youtu.be/Mb9LDzvhecY|Noise Models- AR models]] by Martin Lindquist and Tor Wager (9:57)<<BR>>[[https://youtu.be/NRunOo7EKD8|Inference- Contrasts and t-tests]] by Martin Lindquist and Tor Wager (11:05)<<BR>>[[https://youtu.be/AalIM9-5-Pk|Multiple Comparisons]] by Martin Lindquist and Tor Wager (9:03)<<BR>>[[https://youtu.be/MxQeEdVNihg|FWER Correction]] by Martin Lindquist and Tor Wager (16:11)<<BR>>[[https://youtu.be/W9ogBO4GEzA|FDR Correction]] by Martin Lindquist and Tor Wager (5:25)<<BR>>[[https://youtu.be/N7Iittt8HrU|More about multiple comparisons]] by Martin Lindquist and Tor Wager (14:39) ||
+||<10%>Software ||[[http://nilearn.github.io/stable/index.html|Nilearn]] ||
||Suggested reading ||[[https://doi.org/10.1214/09-STS282|The Statistical Analysis of fMRI Data]], Lindquist, 2008 <<BR>> [[https://doi.org/10.1191/0962280203sm341ra|Controlling the familywise error rate in functional neuroimaging: a comparative review]], Nichols & Hayasaka, 2003 <<BR>> [[https://www.nature.com/articles/s41596-020-0327-3|Analysis of task-based functional MRI data preprocessed with fMRIPrep]], Esteban et al., 2020 <<BR>> [[https://doi.org/10.1016/j.neuroimage.2007.11.048|Guidelines for reporting an fMRI study]], Poldrack et al., 2008 ||
||Suggested viewing ||[[https://www.youtube.com/watch?v=YfeMIcDWwko|Model Building - temporal basis sets]] (11:08)<<BR>>[[https://www.youtube.com/watch?v=Ab-5AbJ8gAs|GLM Estimation]] (9:11)<<BR>>[[https://youtu.be/Mb9LDzvhecY|Noise Models- AR models]] (9:57)<<BR>>[[https://youtu.be/NRunOo7EKD8|Inference- Contrasts and t-tests]] (11:05)<<BR>>[[https://youtu.be/AalIM9-5-Pk|Multiple Comparisons]] by Martin Lindquist and Tor Wager (9:03)<<BR>>[[https://youtu.be/MxQeEdVNihg|FWER Correction]] (16:11)<<BR>>[[https://youtu.be/W9ogBO4GEzA|FDR Correction]] (5:25)<<BR>>[[https://youtu.be/N7Iittt8HrU|More about multiple comparisons]] (14:39) <<BR>> ||
 Line 239:
-||__Datasets__ ||             ||
+||__Datasets__ || ||
 Line 248:
-||||||<tablewidth="100%" tablestyle="margin:0.5em 0px;border-collapse:collapse;border:1px dotted rgb(211, 211, 211);        "style="padding:0.25em;border:1px dotted rgb(211, 211, 211);text-align:center;">~+'''MVPA/RSA II'''+~''' '''<<BR>> Daniel Mitchell & Máté Aller ||
||<12%  style="padding:0.25em;border:1px dotted rgb(211, 211, 211);        ">Software ||<style="padding:0.25em;border:1px dotted rgb(211, 211, 211);">Python implementation of the RSA Toolbox: [[https://github.com/rsagroup/rsatoolbox|Version 3.0]] ||
+||||||<tablewidth="100%" tablestyle="margin:0.5em 0px;border-collapse:collapse;border:1px dotted rgb(211, 211, 211);         "style="padding:0.25em;border:1px dotted rgb(211, 211, 211);text-align:center;">~+'''MVPA/RSA II'''+~''' '''<<BR>> Daniel Mitchell & Máté Aller ||
||<12%  style="padding:0.25em;border:1px dotted rgb(211, 211, 211);         ">Software ||<style="padding:0.25em;border:1px dotted rgb(211, 211, 211);">Python implementation of the RSA Toolbox: [[https://github.com/rsagroup/rsatoolbox|Version 3.0]] ||

MRC CBU Wiki

Quick Links

Search Wiki

Page Tools

Course Material for COGNESTIC 2024

Software Installation Instructions

Essential

Additional Extra