3D data matrices (1 of 2) for Kiemen et al "Quantitative 3D histology reveals localized immune remodeling during early pancreatic cancer progression"

This database consists of .mat MATLAB data files containing 3D volumetric matrices of human pancreas tissue analyzed using the CODA platform. Each .mat file contains the following variables describing pancreatic data from a different human pancreas sample: vol = 3D matrix of size M x N x Z. Each voxel contains a number between 1 and 10 defining the tissue type present in that location. volp = 3D matrix of size M x N x Z. Zero values indicate no PanIN (pancreatic intraepithelial neoplasia, precursor to pancreatic cancer) is present in that voxel, numbers 1 through n indicate the "nth" largest PanIN in that sample by volume. All PanINs labelled inside this matrix were manually confirmed via review of the corresponding histology. volcell = 3D matrix of size M x N x Z. Zero values indicate no nucleus is present in that voxel, positive numbers indicate the number of nuclei present in that voxel. volcelldensity = 3D matrix of size M x N x Z. Numbers indicate the local cell density of CD45+ leukocytes (predicted from H&E) in that voxel, in units of cells per mm3. totvol = total volume of sample in units of cm3. sx = resolution of volume matrices in the x and y dimensions, in units of micron / voxel. This number should be 12. sz = resolution of volume matrices in the z dimension, in units of micron / voxel. This number should be 12 (i.e. the volume of one voxel in the volume matrices = sx*sx*sz = 12*12*12 micron3 / voxel = 1728 micron3 / voxel. classNames = a string array containing the tissue type corresponding to numbers 1 through 10 in the matrix named 'vol'. For example, classNames 1 = "islets of Langerhans," indicating that all voxels == 1 in vol are pancreatic islets. cmap = the RGB colors used for each of the tissue types present in the matrix named 'vol'. For example, cmap(1,:) = [121, 248, 252], corresponding to a cyan color in 8-bit RGB space, and this color is used in our anatomical renderings for vol == 1, or pancreatic islets. For .mat files ending in the extension _IHC, the following variables are present. These variables are in addition to those present in the .mat files with the same S### filename: vollymph (present in S085_IHC.mat and S086_IHC.mat): 3D matrix of size M x N x Z. Zero values indicate no nucleus is present in that voxel, positive numbers indicate the number of CD45+ nuclei (calculated from IHC) present in that voxel. vollymphdensity (present in S085_IHC.mat and S086_IHC.mat): 3D matrix of size M x N x Z. Numbers indicate the local cell density of CD45+ leukocytes (calculated from IHC) in that voxel, in units of cells per mm3. volT (present in S085_IHC.mat): 3D matrix of size M x N x Z. Zero values indicate no nucleus is present in that voxel, positive numbers indicate the number of CD3+ T cells (calculated from IHC) present in that voxel. volTdensity (present in S085_IHC.mat): 3D matrix of size M x N x Z. Numbers indicate the local cell density of CD3+ T cells (calculated from IHC) in that voxel, in units of cells per mm3. volTreg (present in S085_IHC.mat): 3D matrix of size M x N x Z. Zero values indicate no nucleus is present in that voxel, positive numbers indicate the number of foxp3+ regulatory T cells (calculated from IHC) present in that voxel. volTregdensity (present in S085_IHC.mat): 3D matrix of size M x N x Z. Numbers indicate the local cell density of foxp3+ regulatory T cells (calculated from IHC) in that voxel, in units of cells per mm3.