schaefer atlas network order

[jdkent]

Ref #222, #223
Describe the bug
In ThomasYeoLab/CBIG#2 I proposed a mapping between the network numbers and network names based on the RGB components provided and was confirmed by @rubykong.

{
    "17Networks_1": "VisCent",
    "17Networks_2": "VisPeri",
    "17Networks_3": "SomMotA",
    "17Networks_4": "SomMotB",
    "17Networks_5": "DorsAttnA",
    "17Networks_6": "DorsAttnB",
    "17Networks_7": "SalVentAttnA",
    "17Networks_8": "SalVentAttnB",
    "17Networks_9": "Limbic",
    "17Networks_10": "Limbic",
    "17Networks_11": "ContC",
    "17Networks_12": "ContA",
    "17Networks_13": "ContB",
    "17Networks_14": "TempPar",
    "17Networks_15": "DefaultC",
    "17Networks_16": "DefaultA",
    "17Networks_17": "DefaultB"

}

Looking at the community numbers and the community names, the A,B,C networks should be listed in the order C,A,B, as well as some other flips.

So under my proposed mapping, the order of communities names should be:

visualCentral -> visualCentral (unchanged)
visualPeripheral -> visualPeripheral (unchanged)
somatomotorA -> somatomotorA (unchanged)
somatomotorBAuditory -> somatomotorBAuditory (unchanged)
dorsalAttentionA -> dorsalAttentionA (unchanged)
dorsalAttentionB -> dorsalAttentionB (unchanged)
salienceVentralAttentionA -> salienceVentralAttentionA (unchanged)
salienceVentralAttentionB -> salienceVentralAttentionB (unchanged)
limbicOrbitofrontal ->  limbicTemporopolar
limbicTemporopolar -> limbicOrbitofrontal
frontoparietalControlA -> frontoparietalControlC
frontoparietalControlB -> frontoparietalControlA
frontoparietalControlC -> frontoparietalControlB
defaultA -> temporoparietal
defaultB -> defaultC
defaultC -> defaultA
temporoparietal -> defaultB

@rubykong - could you please check to see if me proposed mapping from network number to name is correct, and if it is the same for Buckner, Choi, and yeo?

This is causing me confusion as well, so it will be helpful to get to the bottom of this.
thanks all!

-cc @tientong98

[a3sha2]

@jdkent
Thank you very much. We copied directly atlas and index directly from https://github.com/ThomasYeoLab/CBIG/tree/master/stable_projects/brain_parcellation/Schaefer2018_LocalGlobal/Parcellations.

But we didn't change the community name. We will update it now

[a3sha2]

or can you make pull request pls? @jdkent

[jdkent]

Sure, pinging @rubykong one last time to confirm if this order is correct, otherwise I will email Alexander Schaefer (alexschaefer83@gmail.com) and Thomas Yeo (yeoyeo02@gmail.com) to confirm the order is correct.

[rubykong]

@jdkent @a3sha2
Sorry! Somehow I didn't get any notification from github when you pin me..

Each RGB color corresponds to one network. For the case of Schaefer parcellation, we have to tune the RGB slightly so that the ROIs within a network will have similar color but different RGB values. This means that gray color is always control C, orange color is always control A for both Schaefer2018 and Yeo2011.
However, the ordering of these RGB colors could be different between Yeo2011 and Schaefer2018. For example,
Schaefer2018: orange (control A) -> brown (control B) -> gray (control C)
Yeo2011: gray (control C) -> orange (control A) -> brown (control B)

If you want to follow the network number in Yeo2011. Then for controlA,B,C it should be network 12, 13, 11.

Let me know if it's still confusing!

[Lnan1615]

I read a few articles and referenced the Yeo template, but there were 18 networks. Do you know the names of these 18 networks?

[a3sha2]

Index	Network Name	R	G	B
0	NONE	0	0	0	0
1	17Networks_1	120	18	134	0
2	17Networks_2	255	0	0	0
3	17Networks_3	70	130	180	0
4	17Networks_4	42	204	164	0
5	17Networks_5	74	155	60	0
6	17Networks_6	0	118	14	0
7	17Networks_7	196	58	250	0
8	17Networks_8	255	152	213	0
9	17Networks_9	200	248	164	0
10	17Networks_10	122	135	50	0
11	17Networks_11	119	140	176	0
12	17Networks_12	230	148	34	0
13	17Networks_13	135	50	74	0
14	17Networks_14	12	48	255	0
15	17Networks_15	0	0	130	0
16	17Networks_16	255	255	0	0
17	17Networks_17	205	62	78	0

there are only 17 networks, the 18th network is background (None)