Display atlas choice

[dkp]

It might be useful to explicitly say which atlas is being displayed. If there are multiple atlases, then it might be nice to have a toggle or dropdown that allows you to select the atlas to fetch.

I could also imagine that there would simply be separate notebooks for each atlas choice, but if it is relatively easy to allow users to pick which JSON file to fetch and display the current selection's name (e.g., HO-CB), that would be really nice.

[DevinBayly]

Information about the atlas can certainly be included. The consideration that I have right now is about how/when will a user be switching between atlases.

• as a whole application choice? At the beginning or during their session work?
• for each pane?

It goes without saying then that the import options should translate over to the atlas choice also correct?

[dkp]

Wow, an interesting point. I imagined that atlas would be per-session. I had never considered doing it per-pane. Yes, import options and atlas choice go hand-in-hand. If we count on using roixtractor, then the atlas name is prepended to the name of the CSV file. Could that be useful? It also occurs to me that since you now display the CSV name, the atlas name is right there.

---

Enough blathering. Here's what I imagined: A dropdown at the top of the session where you choose either the HO-CB or HCP-MMP1 atlas. HO-CB would be the default, so mostly people would not do anything at all but if they wanted HCP-MMP1 it could be selected.

Is this practical and possible?

[DevinBayly]

I guess my question went a bit deeper. Import as is now is non-session interrupting. It simply creates a new pane and plops it into the collection. Atlas selection, and importing a session using a different atlas seem like they should be fresh start kinds of things. I worry that including this feature is actually tied to a deeper set of interaction options than it first appears. As long as you don't mind that it resets the whole session back to square one its possible?

[DevinBayly]

I'd leave practical as a question for you haha.

[dkp]

One atlas per session then. It might be important to make that clear, especially before wiping out a session (resetting) if someone goofs. So, sure, reset it, but ask them if they mean for that to happen and let them back out or save?

[DevinBayly]

Sounds good. I'll make a little alert pop up if they are interested in doing that.

[DevinBayly]

This probably wont be done for elena's first test tomorrow.

[dkp]

I doubt Elena will test tomorrow, if so, it’d all be on the Default HO-CB atlas...so, this will not be an issue. -D

Get Outlook for iOShttps://aka.ms/o0ukef

---

From: Devin Bayly notifications@github.com Sent: Monday, January 20, 2020 2:51:31 PM To: DevinBayly/neuro-choro neuro-choro@noreply.github.com Cc: Patterson, Dianne K - (dkp) dkp@email.arizona.edu; Author author@noreply.github.com Subject: Re: [DevinBayly/neuro-choro] Display atlas choice (#30)

This probably wont be done for elena's first test tomorrow.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/DevinBayly/neuro-choro/issues/30?email_source=notifications&email_token=AAHLUQEOVSSRPMVTTO65YNTQ6YMGHA5CNFSM4J4AMVC2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJN4R3Q#issuecomment-576440558, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAHLUQH3U2LCYMEPZ33Y2RTQ6YMGHANCNFSM4J4AMVCQ.

[DevinBayly]

Ran into issue with descrepancies between mappings of atlas json and csvs. Let me elaborate.

An for HCP atlas stores the names of regions as properties using names found in the regionNames column of HCP-MMP1_List csv

HO-CB atlas store the names of the regions as properties using names found in the region column HO-CB_run-01_IC-06_Russian_Unlearnable_1.5 csv

can you tell me definitively which csv file represents the output of the roixtractor? In the code for now I'll write in a little conditional that makes sure we use values found in the csv for the correct brain regions. But in the future we might get into trouble if other atlases are added, or if there's unreliable naming of csv region column row names meant to map to the json data.

[dkp]

Feb 11th email: HO-CB_run-01_IC-06_Russian_Unlearnable_1.5 csv is an example of a particular output from roixtractor. It is not guaranteed to contain all the possible region names.

The relevant atlas lists are in the ATLASES directory for roixtractor here: https://bitbucket.org/dpat/roixtractor/src/master/ATLASES/ One is HCP-MMP1_List.csv: https://bitbucket.org/dpat/roixtractor/src/master/ATLASES/HCP-MMP1_List.csv The other is HO-CB_List.csv: https://bitbucket.org/dpat/roixtractor/raw/d50da8facf924386df36ea97c7f30369aa68a37a/ATLASES/HO-CB_List.csv

When roixtractor runs, it gets three arguments (bolded below): docker run -it --rm -v ${PWD}:/scif/data/roixtractor diannepat/scif_fsl run roixtractor run-01_IC-06_Russian_Unlearnable.nii 1.3 HO-CB

roixtractor requires 3 input arguments: (1) a statistical image to run on (2) a statistical threshold and (3) an atlas name.

Given these 3 inputs, roixtractor produces a CSV file that encodes all the information about the arguments passed into it.

So, you have been working with a sample output file called HO-CB_run-01_IC-06_Russian_Unlearnable_1.5.csv This sample was produced by running roixtractor on run-01_IC-06_Russian_Unlearnable.nii, using the threshold 1.5 and using the HO-CB atlas.

Sorry for the confusion. Let me know if any of this doesn't make sense.

[dkp]

I think you did this, right? Did you get a chance to process my response above about the basis of each atlas?