PCA MNI 2009b NLIN template Andreas Horn 10.6084/m9.figshare.4644472.v2 https://figshare.com/articles/dataset/PCA_MNI_2009b_NLIN_template/4644472 This is the first principal component combining T1-, T2- and proton-density (PD)-weighted templates of the MNI 2009b high resolution nonlinear asymmetric template series.<div><br></div><div>This single template retains ≥ 87.34% of variance in T1, T2 and PD templates combined within the brain segment of the templates.</div><div><br></div><div>This additional PCA template may be used for multispectral normalizations (e.g. as implemented in Lead-DBS software1).</div><div><br></div><div><b>Example case:</b> You did acquire T1-, T2-, PD- acquisitions from a subject but also acquired an inversion recovery (IR) sequence and you wish to combine all anatomical information to warp into standard space.</div><div><br></div><div>In this case, you could set up a combined nonlinear warp matching:</div><div><br></div><div>T1 acquisition -> T1 template</div><div>T2 acquisition -> T2 template<br></div><div>PD acquisition -> PD template<br></div><div>IR acquisition -> PCA template* (since no IR template available).<br></div><div><br></div><div>* using e.g. normalized mutual information as a convergence metric.</div><div><br></div><div>There is no empirical data to support this makes sense, but intuitively, the PCA template should be the most "universal" and contain information from all three templates available from the MNI.<br></div><div><br></div><div>Code to (re-) generate this template is available within Lead-DBS ('ea_pca_templates.m').</div> 2017-02-16 17:33:58 Magnetic resonance imaging Neuroscience