Sorry for the indecipherable original posting. It made perfect sense to me because I knew exactly what I was talking about. Here is my second attempt.

Here is the 3dinfo output for a bucket dataset created from .BRIK files that were never in .img format.

Filename = Reversal.full4CStype.percent.Anova+tlrc

-- At sub-brick #0 'CS_Type:Inten' datum type is short: 0 to 32767 [internal]
[* 0.0318653] 0 to 1044.13 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[0] ; Reversal.full4CStype.percent.Anova+tlrc[0] ; CS_Type+tlrc[0]
-- At sub-brick #1 'CS_Type:F-stat' datum type is short: 0 to 10521 [internal]
[* 0.01] 0 to 105.21 [scaled]
statcode = fift; statpar = 1 22
keywords = Reversal.full4CStype.percent.Anova+tlrc[1] ; Reversal.full4CStype.percent.Anova+tlrc[1] ; CS_Type+tlrc[1]
-- At sub-brick #2 'Block:Inten' datum type is short: 0 to 32767 [internal]
[* 0.0468081] 0 to 1533.76 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[2] ; Reversal.full4CStype.percent.Anova+tlrc[2] ; Block+tlrc[0]
-- At sub-brick #3 'Block:F-stat' datum type is short: 0 to 618 [internal]
[* 0.01] 0 to 6.18 [scaled]
statcode = fift; statpar = 5 110
keywords = Reversal.full4CStype.percent.Anova+tlrc[3] ; Reversal.full4CStype.percent.Anova+tlrc[3] ; Block+tlrc[1]
-- At sub-brick #4 'CS_typeXBlock:Inten' datum type is short: 0 to 32767 [internal]
[* 0.0318296] 0 to 1042.96 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[4] ; Reversal.full4CStype.percent.Anova+tlrc[4] ; CS_typeXBlock+tlrc[0]
-- At sub-brick #5 'CS_typeXBlock:F-stat' datum type is short: 0 to 554 [internal]
[* 0.01] 0 to 5.54 [scaled]
statcode = fift; statpar = 5 110
keywords = Reversal.full4CStype.percent.Anova+tlrc[5] ; Reversal.full4CStype.percent.Anova+tlrc[5] ; CS_typeXBlock+tlrc[1]
-- At sub-brick #6 'CSType.clsto[0]' datum type is short: 0 to 11
keywords = Reversal.full4CStype.percent.Anova+tlrc[6] ; CSType.clstorder+tlrc[0]
-- At sub-brick #7 'Block.clstor[0]' datum type is short: 0 to 0 [internal]
[* 0.0468081] 0 to 0 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[7] ; Block.clstorder+tlrc[0]
-- At sub-brick #8 'InterX.clsto[0]' datum type is short: 0 to 0 [internal]
[* 0.0318296] 0 to 0 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[8] ; InterX.clstorder+tlrc[0]
-- At sub-brick #9 'CSType.clust[0]' datum type is short: 0 to 293 [internal]
[* 0.0318653] 0 to 9.33654 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[9] ; CSType.clust+tlrc[0]
-- At sub-brick #10 'Block.clust[0]' datum type is short: 0 to 0 [internal]
[* 0.0468081] 0 to 0 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[10] ; Block.clust+tlrc[0]
-- At sub-brick #11 'InterX.clust[0]' datum type is short: 0 to 0 [internal]
[* 0.0318296] 0 to 0 [scaled]
keywords = Reversal.full4CStype.percent.Anova+tlrc[11] ; InterX.clust+tlrc[0]



*****Here is the 3dinfo output for a bucket dataset created from .BRIK files generated from an .img file. I used ANALYZEtoAFNI to create a .HEAD file for 'PLSbsr_rev_task_lv4_lag1.img' first.

Filename = Task_LV1+tlrc
-- At sub-brick #0 'PLSbsr_rev_task_lv4_lag1.img' datum type is float: -4.29764 to 3.91479
keywords = rawLV1.1+tlrc[0]
-- At sub-brick #1 'PLSbsr_rev_task_lv4_lag1.img' datum type is float: 0 to 3.91479
keywords = pos_1.1+tlrc[0]
-- At sub-brick #2 'PLSbsr_rev_task_lv4_lag1.img' datum type is float: 0 to 1
keywords = pos_mask_1.1+tlrc[0]
-- At sub-brick #3 'PLSbsr_rev_task_lv4_lag1.img' datum type is float: -4.29764 to 0
keywords = neg_1.1+tlrc[0]
-- At sub-brick #4 'PLSbsr_rev_task_lv4_lag1.img' datum type is float: 0 to 1
keywords = neg_mask_1.1+tlrc[0]


***The sub-bricks in this file are just mask files of different kinds generated using 3dcalc. Notice that for 'Task_LV1+tlrc' the sub-brick label is the original .img file (PLSbsr_rev_task_lv1_lag1.img). The key word is the name of the file that was added to create the bucket dataset.

When opening the bucket just described in afni I tried to select the functional sub-brick that I wanted to view. The drop down list looked like this:

# 0 PLSbsr_rev_tas
# 1 PLSbsr_rev_tas
# 2 PLSbsr_rev_tas
# 3 PLSbsr_rev_tas
# 4 PLSbsr_rev_tas

When opening the bucket 'Reversal.full4CStype.percent.Anova+tlrc' it looks like this:

#0 'CS_Type:Inten'
#1 'CS_Type:F-stat'
#2 'Block:Inten'
etc

Therefore, I concluded that afni seems to be reading in the sub-brick label from the header in order to show the names of the sub-brick files when using the functional sub-brick selector.

Becuase the sub-brick labels in 'Task_LV1+tlrc' are all the same, I am unable to tell which sub-brick is which when using afni. The bucket dataset was created fine, but the "sub-brick label" and the "keyword" seem to be swapped in these two bucket datasets described above.

To fix this I thought of swapping the "sub-brick label" and the "keyword" for each sub-brick, but could not find the option to do this in 3drefit. Appending text to the existing sub-brick label would be futile because as you can see above the file name is alreay longer than the sub-brick selector allows.

I hope this is clearer,
Christine Smith