Hello,

I was just wondering if the sliding windows employed in the latest versions of CONN follow,

a) a simple block nature, taking the entire chuck of the time series, OR

b) a similar type to the sliding windows with sigmoid edges (described by Dr. Calhoun), and used in dynamic functional connectivity analysis by Allen et al (2013)

I have tried to visualize the two types I have mentioned, and attached the image (Apologies for my drawing!).

I would just like to clarify this, to understand how the sliding window separation is done in the SETUP.

Thank you,
Aditya

Originally posted by Aditya Jayashankar:
Hi,

If anyone could please tell me the type of window function used in CONN 17f during the SETUP, when you run temporal decomposition into sliding windows, it would really helpful.

I am confused whether CONN uses a simple rectangular filter, or a convolved rectangular and Gaussian (Allen et al, 2014), or a special type of tapering window (Tukey or Planck).

Thank you,
Aditya