SPM

Statistical Parametric Mapping

Documentation & support

Written by members & collaborators of the Wellcome Department of Cognitive Neurology

| intro | distribution | SPM'96 | SPM'99 | documentation | email list | course | notes | SnPM | SPMwin | data |

Academic papers

The primary reference for the theories of Statistical Parametric Mapping are the academic papers in the peer reviewed literature. References are given below.

Manuscripts of these will be made available here when they are accepted.

Core SPM papers

[SPM_1] Spatial Registration and Normalization of Images

Friston KJ, Ashburner J, Poline JB, Frith CD, Heather JD, Frackowiak RSJ (1995)
Human Brain Mapping 2:165-189

[SPM_2] Assessing the Significance of Focal Activations Using their Spatial Extent

Friston KJ, Worsley KJ, Frackowiak RSJ, Mazziotta JC, Evans AC (1994)
Human Brain Mapping 1:214-220

[SPM_3] Statistical Parametric Maps in Functional Imaging: A General Linear Approach

Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD, Frackowiak RSJ (1995)
Human Brain Mapping 2:189-210

[SPM_4] Estimating Smoothness in Statistical Parametric Maps: Confidence Intervals on p-Values

Poline J-B, Friston KJ, Worsley KJ & Frackowiak RSJ (1995)
Journal of Computer Assisted Tomography 19(5):788-796

[SPM_5] Detecting Activations in PET and fMRI: Levels of Inference and Power

Friston KJ, Holmes A, Poline J-B, Price CJ & Frith CD (1995)
Neuroimage 4:223-235

[SPM_6] Cognitive Conjunction: A New Approach to Brain Activation Experiments

Price CJ & Friston KJ (1996)
Neuroimage 5:261-270

[SPM_7] Multimodal Image Coregistration and Partitioning - a Unified Framework

Ashburner J, Friston KJ (1997)
NeuroImage 6(3):209-217

[SPM_8] Incorporating Prior Knowledge into Image Registration

Ashburner J, Neelin P, Collins DL, Evans AC, Friston KJ (1997)
NeuroImage 6:344-352

[SPM_9] Nonlinear Spatial Normalization using Basis Functions

Ashburner J, Friston KJ (1999)
Human Brain Mapping 7(4):254-266

[fMRI_1] Movement-Related Effects in fMRI Time-Series

Friston KJ, Williams SR, Howard R, Frackowiak RSJ & Turner R (1996)
Magnetic Resonance in Medicine 35:346-355

[fMRI_2] Analysis of Functional MRI Time-Series

Friston KJ, Jezzard P, Turner R (1994)
Human Brain Mapping 1:153-171

[fMRI_3] Analysis of fMRI Time Series Revisited

Friston KJ, Holmes AP, Poline JB, Grasby PJ, Williams SCR, Frackowiak RSJ, Turner R (1995)
Neuroimage 2:45-53

[fMRI_4] Analysis of fMRI Time-Series Revisited - Again

Worsley KJ, Friston KJ (1995)
Neuroimage 2:173-181

[fMRI_5] Characterizing Dynamic Brain Responses with fMRI: A Multivariate Approach

Friston KJ, Frith CD, Frackowiak RSJ, Turner R (1995)
Neuroimage 2:166-172

[fMRI_6] Characterizing Evoked Hemodynamics with fMRI

Friston KJ, Frith CD, Turner R, Frackowiak RSJ (1995)
Neuroimage 2:157-165

[PET_1] A Multivariate Analysis of PET Activation Studies

Friston KJ, Poline J-B, Holmes AP, Frith CD, Frackowiak RSJ (1996)
Human Brain Mapping 4:140-151

[PET_2] Functional Topography: Multidimensional Scaling and Functional Connectivity in the Brain

Friston KJ, Frith CD, Fletcher P, Liddle PF, Frackowiak RSJ (1996)
Cerebral Cortex In press

SPM related & overview papers

Spatial Normalisation: A New Approach

Friston KJ (1995)
In Proceedings of BrainMap'95 UTHSC San-Antonio, Texas

Statistical Methods in Neuroimaging, with Particular Application to Emission Tomography

McColl JH, Holmes AP, Ford I. (1994)
Statistical Methods in Medical Research 3(1):63-86

Statistical Issues in functional Brain Mapping

Holmes AP (1994)
Doctor of Philosophy Thesis, University of Glasgow, December 1994.

Non-Parametric Analysis of Statistic Images From Functional Mapping Experiments

Holmes AP, Blair RC, Watson JDG, Ford I (1996)
Journal of Cerebral Blood Flow and Metabolism 16:7-22

MRI and PET Coregistration - A Cross Validation of Statistical Parametric Mapping and Automated Image Regression

Kiebel SJ, Ashburner A, Poline J-B, Friston KJ (1997)
Neuroimage 5:271-279

The role of registration and spatial normalization in detecting activations in functional imaging

Ashburner J, Friston KJ (1997)
Clinical MRI/Developments in MR 7(1):26-28

Related theory papers (offsite)

Local maxima and the expected Euler characteristic of excursion sets of \chi^2, F and t fields

Worsley KJ (1994)
Advances in Applied Probability 26:13-42

Quadratic tests for local changes in random fields with applications to medical images

Worsley KJ (1994)
Technical Report, Department of Mathematics and Statistics, McGill University 94-08

Testing for a signal with unknown location and scale in a stationary Gaussian random field

Siegmund DO & Worsley KJ (1995)
Annals of Statistics 23:608-639

Estimating the number of peaks in a random field using the Hadwiger characteristic of excursion sets, with applications to medical images

Worsley KJ (1995)
Annals of Statistics 23:640-669

Boundary corrections for the expected Euler characteristic of excursion sets of random fields, with an application to astrophysics

Worsley KJ (1995)
Advances in Applied Probability 27 943-959

A unified statistical approach for determining significant signals in images of cerebral activation

Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ, & Evans AC (1995)
Human Brain Mapping accepted

Tests for distributed, non-focal brain activations

Worsley KJ, Poline J-B, Vandal AC, & Friston KJ (1995)
NeuroImage 2:183-194

Analysis of fMRI time-series revisited - again

Worsley KJ, & Friston KJ (1995)
NeuroImage 2:173-181

Searching scale space for activation in PET images

Worsley KJ, Marrett S, Neelin P, & Evans AC (1995)
Human Brain Mapping 4:74-90

A unified statistical approach for determining significant signals in location and scale space images of cerebral activation

Worsley KJ, Marrett S, Neelin P, & Evans AC (1995)
In Quantification of brain function using PET Eds. R.Myers, V.J.Cunningham, D.L.Bailey & T.Jones, Academic Press, San Diego.

The geometry of random images

Worsley KJ (1996)
Chance 9(1):27-40

An unbiased estimator for the roughness of a multivariate Gaussian random field

Worsley KJ (1996)
Technical Report, Department of Mathematics and Statistics, McGill University

Characterizing the response of PET and fMRI data using multivariate linear models

Worsley KJ, Poline JB, Friston KJ, & Evans AC (1998)
NeuroImage 6:305-319

Scale space searches for a periodic signal in fMRI data with spatially varying hemodynamic response

Worsley KJ, Wolforth M, & Evans AC (1997)
Proceedings of BrainMap'96 Conference submitted

Testing for a signal with unknown location and scale in a chi^2 random field, with an application to fMRI

Worsley KJ(1997)
Advances in Applied Probability submitted

An overview and some new developments in the statistical analysis of PET and fMRI data

Worsley KJ(1997)
Human Brain Mapping 5:254-258.

Third party papers, documentation & guides (offsite)

MEDx SPM documentation

As part of Sensor systems port of SPM to their MEDx medical neuroimaging package, they have produced a comprehensive user guide, which is freely available to the neuroimaging community. This guide describes what each of the inputs mean and how to fill them in. Although written around the MEDx user interface, the information that it contains will be useful for those using the original Matlab version of SPM.

See the MEDx-SPM documentation page at http://www.sensor.com/medx_info/spmdocs.html for details.

Beginners guides & overviews of SPM

Cambdge Imagers SPM notes (MRC Cognition and Brain Sciences Unit)

http://www.mrc-cbu.cam.ac.uk/Imaging/spm.html

Matthew Brett's basic overview of models, design matrices, and contrasts.

http://www.mrc-cbu.cam.ac.uk/Imaging/spmstats.html
email <matthew.brett@mrc-cbu.cam.ac.uk>

Matthew Brett's basic tutorial on random field theory.

http://www.mrc-cbu.cam.ac.uk/Imaging/randomfields.html
email <matthew.brett@mrc-cbu.cam.ac.uk>

SPM course notes

The annual SPMcourse is accompanied by a comprehensive set of course notes. These notes provide an accessible and readable exposition of the statisical, mathematical and image processing concepts of Statistical Parametric Mapping, together with practical notes on their application. All SPM users are encouraged to read the SPMcourse notes. Of particular relevance is Chapter 10: SPM'96 usage - an illustrated guide, by Chloe Hutton.

Printed copies of the SPMcourse notes are given to course attendees. Electronic (Adobe Acrobat) editions of the course notes are being made freely available to aid understanding of Statistical Parametric Mapping. See http://www.fil.ion.ucl.ac.uk/spm/course/notes.html for details.

Online documentation and help

Derived from spm_help.m

Using the GUI online help system

The SPM function spm_help.m sets up a GUI help system for the SPM package. This can be launched from the SPM splash screen "About SPM", from the SPM menu window, or with the various context-sensitive help buttons.

It can also be used directly: `spm_help('Topic')` or `spm_help Topic` displays the help for a particular topic in the SPM help window.

Help topics are displayed in a special help window. Initially, a representation of the SPM Menu window is drawn. Clicking buttons in this representation leads to the help pages for the appropriate topic. The SPM Help ToolBar contains controls for the help system.

• Routines and manual pages (generically referred to as help "topics") referenced by the currently displayed help file are listed in the "Referenced Topics..." pull-down menu on the help interface. Selecting a routine there displays it's help information.
• Aternatively, a specific function name (with the ".m" extension) can be entered into the lilac editable text widget in the help window toolbar. It's help will be displayed.
• As the tree of routines is explored, the "Previous Topics" pulldown menu keeps track of recently visited help topics, allowing easy backtracking and exploration of the "tree" of SPM functions and dependencies.
• Special topic buttons lead to "About SPM", "Menu", and "Help" topics. "About SPM" displays the introduction file for this version of SPM. "Menu" displays the help representation of the SPM Menu window. "Help" displays this file.
• Multi-page help files are displayed by the help facility with pagination controls at the bottom right of the Graphics window.
• The print button enables printing to the specified Print device. This is set in the Defaults area, initially to the PostScript file "spm.ps" in the current directory. Multi page topics are paged on screen, and printed page by page.

Levels of online documentation

The SPM package provides help at three levels, the first two being available via the SPM graphical help system:

1. Manual pages on specific topics

   These give an overview of specific components or topics its relation to other components, the inputs and outputs and references to further information.

   Many of the buttons in the help menu window lead to such "man" pages. These are contained in ASCII files named spm_*.man. These can be viewed on the MatLab command line with the `help` command, e.g. `help spm_help.m` prints out this manual file in the MatLab command window.

2. Routine help info

   Help information for each routine within SPM (E.g. This is the help information for spm_help.m - the help function.) This help information is the help header of the actual MatLab function, and can be displayed on the command line with the `help` command, e.g. `help spm_help`.

   Commented header text from that spm_*.m file is displayed in the following format:

   	     A one line description
   	     FORMAT [outputs] = spm_routine(inputs);
   	     inputs  -  the input arguments
   	     outputs -  the output arguments
   	     ____________________________________________________________
   
   	     Short paragraphs detailing what the routine does, and other
   	     pertinent information.
   
   	     Ref: citations
   	     ____________________________________________________________
                Version and author information
   

3. The MatLab SPM code itself

   SPM is (mainly) implemented as MatLab functions and scripts. These are ASCII files named spm_*.m, which can be viewed in the MatLab command window with the `type` command, e.g. `type spm_help`, or read in a text editor.

   MatLab syntax is very similar to standard matrix notation that would be found in much of the literature on matrices. In this sense the SPM routines can be used (with Matlab) for data analysis, or they can be regarded as the ultimate pseudocode specification of the underlying ideas.

   The coding is concise but clear, and annotated with comments where necessary.

In addition, the MatLab help system provides keyword searching through the H1 lines (the first comment line) of the help entries of *all* M-files found on MATLABPATH. This can be used to identify routines from keywords. Type `help lookfor` in the MatLab command window for further details.

Bugs & features

Bugs?

...or are they features?

Please report bugs to spm-bugs@fil.ion.ucl.ac.uk. Peculiarities may actually be features, and should be discussed on the SPM email discussion list, spm@mailbase.ac.uk.

Feature requests

All suggestions are welcome, particularly those substantiated with code. Contact us at spm-authors@fil.ion.ucl.ac.uk.

Potential collaborators or those wishing to visit should contact us individually.

eMail discussion list