request: whole brain measurements

[iamjoostjanssen]

Hi,

FreeSurfer stats output includes whole brain/hemispheric averages or total, depending the measurement. Could this be added to mindboggle output tables as well?

Thanks,
-joost

[binarybottle]

The reason Mindboggle breaks brain images down into structures such as labeled regions is to compute structure-based shape measures such as curvature, depth, and thickness, rather than across an entire hemisphere. Which measurements were you hoping to aggregate across labeled regions for each hemisphere?

[iamjoostjanssen]

Depth and sulcal/fundi surface area (of un- and labeled structures) and corrected as per compute.py.

[binarybottle]

@iamjoostjanssen --

Total surface area across all sulci
To compute the total surface area for all labels, sulci, or fundi for either hemisphere, all you need to do is compute the sum of the values in the "area" column of the corresponding [label, sulcus, fundus]_shapes.csv table.

Average the mean depth per sulcus across all sulci
Mindboggle generates statistics for each (vertex-wise) shape measure for each structure, such as mean travel depth per sulcus. You could easily compute the mean of the average depth per label, sulcus, or fundus across all labels, sulci, or fundi by computing the mean of the values in the "mean" column of the corresponding [label, sulcus, fundus]_shapes.csv table.

Mean depth across all labels
If you want to get the mean travel depth (or other shape measure) across all labels, all you have to do is compute the mean of the values in the appropriate column in the vertices.csv file.

Mean depth across all sulci
If you want to get the mean travel depth (or other shape measure) across all sulcus vertices, compute the mean of the values in the "travel depth" (or other shape) column in the vertices.csv file for rows where the "sulcus ID" column does not equal -1.

Does this help?

[binarybottle]

If you have processed many brains through mindboggle and want to do a computation on a shape measure for each of the brains, you can create a table concatenating the desired column across all the brains, using the select_column_from_mindboggle_tables() function in tables.py:

Select column from Mindboggle shape tables and make a new table.
For example, extract the median travel depth column for the label regions
across a set of subjects, and make a new table.

[binarybottle]

Addendum:

I neglected to mention above that for area correction when computing Mean depth across all labels or Mean depth across all sulci you can use the corresponding values in the area column of the same vertices.csv file.