Hi,
If you are just keeping the strongest X% of the connections, I would definitely normalise and average before thresholding. This ensures you have the best possible estimate of the tractography connectivity before applying the threshold.
Best wishes,
Michiel
> On 29 May 2020, at 13:09, Samantha Baldi <[log in to unmask]> wrote:
>
> Hi Michiel,
>
> I was recently re-discussing my processing steps for the connectivity matrix and got back to the issue of the order of these steps. I am thus hoping that you could provide a (hopefully) final clarification on the matter.
>
> In the end, I opted for a density-based thresholding procedure, with which I retain a certain percentage of the strongest edge weights to keep the connection density of the network fixed across subjects. So all edges are ranked from strongest to weakest, and then the top 23% (the threshold I chose) is retained for each subject.
> Up until now, I sticked to the advice you gave and followed the order 1) thresholding, 2) normalization and 3) averaging.
>
> However, I recognize now that the thresholding approach that I originally mentioned in this thread is quite different from the one that I ended up using. I was just wondering if, given this type of thresholding, you would still stand with your advice, or if you think the order might be better reversed (i.e. 1) normalization, 2) thresholding and 3) averaging)?
>
> I hope this is clear, and thanks already a lot for the help!
>
> ########################################################################
>
> To unsubscribe from the FSL list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
>
> This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the FSL list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=FSL&A=1
This message was issued to members of www.jiscmail.ac.uk/FSL, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|