Date written: 2020-07-10
Last ran: 2020-07-10

Description: This report summarizes Slicer segmentation success, for our new data analysis plan of SPINS CMH GE and PNS CMH GE. We capture data for n=172 out of an expected total of n=210 eligible participants. The discrepancy is due to the fact that some (roughly ~30) PNS participants have not been preprocessed (in progress!). Note: I have not removed participants who are will be excluded from the CCA analysis due to missing social cognition and/or neurocognition data.

Table: Fiber summary.

Below, all data is collapsed across the SPINS and PNS studies. The count variable indicates the number of participants with data for a given tract, and percent indicates corresponding percentage. Note that the table does not show data for tracts missing across all participants: all participants are missing the following tracts: CPC -- LR. In total, we have data for 11935 tracts out of a possible maximum of 12384, i.e., 96.374354%. Again, we see that the SLF I and II can not be reliably segmented. However, we have high success everywhere else.

Visualization: Complete missingness.

Below is a visual summary of missing data. Black indicates that the tract is missing entirely for the given participant. Again, tracts missing across all participants are not shown here: CPC -- LR.

Table: Missing tracts by participant. Missing tracts appear to be spread normally across the participant sample (i.e., not determined by diagnosis, scan date, etc):

Tracts missing 0 1 2 3 4 5 6 7
Percent missing 0.00 1.39 2.78 4.17 5.56 6.94 8.33 9.72
Participant count 8 27 53 45 22 10 6 1

Visualizations

The following plots show, per tract, (i) the number of fibers segmented, (ii) mean length, (iii) and mean FA, for each of the n=172 participants, separated by hemisphere (n=74). Again, tracts missing across all participants are not shown here: CPC -- LR. The green diamond shows the median. Outlier values are plotted in red.

Number of Fibers

Mean Length

Mean FA