Conn pipeline how to decide?

Community Chat
1

Hello All,

please suggest me what changes in conn pipeline(sequence to alter) I should to do to get better functional connectivity results.

2

Hi Deepa,

Unfortunately there are a lot of factors that go into this question. You might find this paper, and those cited within, helpful: https://www.nature.com/articles/s41596-018-0065-y

Here are some general suggestions:

  1. preprocess with fmriprep, and use outputs to do quality control
  2. the only additional preprocessing step conn should have to do is smooth
  3. pick your denoising confounds carefully after considering your data, hypotheses, and the paper linked above

The paper linked uses software called the XCP pipeline after minimal preprocessing, which is great because it will do smoothing and has denoising pipelines built in, leaving the guesswork out on your end.

Hope this helps,
Steven