GIC-experimental-databases/OASIS VBM feature vectors
CONTROLES is the raw data of the control subjects, they have label -1 PACIENTES is the raw data of the patient subjects, they have label 1
If you read the first lines of Diverse_AdaBoost_LVQ_MeanAndStdDev.m you will see that I call the function that I am attaching (extractMeanAndStdDevFromEachCluster.m).
These data are grey matter segmentations as you can see with, e.g.,:
imshow (reshape(PACIENTES(1,60,:,:),91, 91), [min(min(PACIENTES(1,60,:,:))) max(max(PACIENTES(1,60,:,:)))])
female_crystal_brain_cov0 are corrected p-values of a typical statistical test in with anatomical brain MRI in neuroscience called Voxel-based morphometry (VBM).
To use them the same way you used the previous dataset, please use one of the functions I attached in this way:
[C P] = extractVoxelIntensitiesWithinClusters (female_crystal_brain_cov0, CONTROLES, PACIENTES);
Where C are the controls voxels within the VBM clusters and P the patients ones.
Otherwise, if you don't want to use the VBM result you could perform another feature extraction of yourself directly on CONTROLES and PACIENTES. I imagine you would have to reshape it, flattening their 2nd, 3rd and 4th dimesions into one.
- Other files of interest
- A. Savio, M.T. García-Sebastián, D. Chyzyk, C. Hernandez, M. Graña, A. Sistiaga, A. López de Munain, J. Villanúa,
- Neurocognitive disorder detection based on feature vectors extracted from VBM analysis of structural MRI
- Computers in Biology and Medicine, Volume 41, Issue 8, August 2011, Pages 600-610, ISSN 0010-4825, http://dx.doi.org/10.1016/j.compbiomed.2011.05.010.
- Keywords: Alzheimer's disease; Classification; Feature extraction; Structural MRI; Myotonic distrophy of type 1
- Darya Chyzhyk, Alexandre Savio, Manuel Graña,
- Evolutionary ELM wrapper feature selection for Alzheimer's disease CAD on anatomical brain MRI pdf
- Neurocomputing, Volume 128, 27 March 2014, Pages 73-80,
- ISSN 0925-2312, http://dx.doi.org/10.1016/j.neucom.2013.01.065.