Welcome to the wiki pages of pTFCE, developed and maintained by the Predictive Neuroimagiong Laboratory (PNI-lab) of the University Hospital Essen, Germany.
New R-package release 0.2.2, with a small bugfix (fixing NaN values at low Z-scores on imgages with low-smoothness).
From 01.01.2020, the maintanence and developement of pTFCE will be continued in the recently established Predictive Neuroimagiong Laboratory (PNI-lab), lead by Tamas Spisak.
Check out this thread on applying pTFCE on both side of your Z-score map.
v0.2.0 available for both the R-package and the SPM Toolbox. It features a speed performance patch: pTFCE is now almost twice as fast!
Since the start, 700 unique visitors, >1000 visits and a lot of e-mails. Thanks for all your interest!
Smoothness estimation is now possible based on 4D residual data
pTFCE (probabilistic TFCE) is a cluster-enahncement method to improve detectability of neuroimaging signal. It performs topology-based belief boosting by integrating cluster information into voxel-wise statistical inference.
Figure 1. pTFCE achieves a significant increase in statistical power in most of the typical fMRI processing scenarios.
See the paper for details.
For a detailed description and theory, please refer to (and please cite):
Tamás Spisák, Zsófia Spisák, Matthias Zunhammer, Ulrike Bingel, Stephen Smith, Thomas Nichols, Tamás Kincses, Probabilistic TFCE: a generalised combination of cluster size and voxel intensity to increase statistical power. Neuroimage, 185:12-26.
SPM Matlab Toolbox & Installation
- Overview
- Relation to TFCE
- The R-package
3.1 Installation
3.2 Usage - The SPM Toolbox
4.1 Installation
4.2 Usage - The FSL extension
5.1 Installation
5.2 Usage - The Nipype interface
Citation and References
Figure 2. A graphical representation of pTFCE depicting the integaration of cluster probabilities at various cluster-forming threshold via Bayes' Theorem and our incremental probability aggregation technique.