Spm

Statistical Parametric Mapping, usually abbreviated SPM, is a software package for the analysis of brain imaging data, particularly functional MRI (fMRI) and positron emission tomography (PET). Developed at University College London and affiliated centers, it is implemented as a MATLAB toolbox and is one of the most widely used tools in cognitive neuroscience for statistical analysis of neuroimaging data.

SPM provides an integrated workflow that includes data preprocessing (slice timing, realignment, co-registration, normalization, segmentation, smoothing), followed by first- and second-level statistical analysis using the general linear model (GLM). Inference relies on random field theory to correct for multiple comparisons across voxel fields, yielding corrected statistical maps and significance thresholds. It also supports Bayesian approaches and various modeling options, including flexible factorial designs.

History and versions: The project began in the 1990s; major milestones include SPM2, SPM5, SPM8, and the current SPM12 series. SPM12 introduced unified segmentation, a DARTEL-based high-dimensional normalization, improved display, and integration with other neuroimaging tools. The software remains actively developed with frequent updates and extensive documentation.

Usage and ecosystem: SPM runs on MATLAB and requires a license; while MATLAB is proprietary, SPM itself is freely available for academic use. It has a large user community, extensive tutorials, and mailing lists. It is often compared with other neuroimaging suites such as FSL and AFNI, each with different strengths.

In research practice, SPM is favored for GLM-based analyses of task and resting-state data, and remains a core reference in neuroimaging methodology.