About ReHo

Submitted by cherrychingju on Sat, 10/04/2014 - 06:55

Dear all,

Is ReHo value equal to Kendall's W (KCC)?

And if we use a region to be a ROI and extract its value, then this value is ReHo value or KCC value? (Some values are more than 1, so KCC can be more than 1?) 

Your help will be greatly appreciated!

 

Best, 

Cherry

 

Forums:

Dina Dajani

Mon, 10/06/2014 - 15:02

ReHo values are KCC values (W), which range from 0 to 1.  I am wondering what the "m" prefix means for the Reho output, because these values range from 0 to above 1.  Also, the "s" prefixed images should also range from 0 to 1, these are just smoothed.  

ZangYF

Mon, 10/06/2014 - 15:30


mReHo is a standardized value by dividing the global mean ReHo, so, its value is no longer between 0 to 1.

 

Yufeng

 

From:

Dina Dajani

Mon, 10/06/2014 - 16:16

Thank you for this information.  So what is the difference between the m* maps and the z* maps?

 

 

ZangYF

Mon, 10/06/2014 - 16:40


m* is standardized by dividing the global mean ReHo, and z* is further standardized by z value (minus mean, then divided by SD). If do a two-sample t-test (between two groups) or paired t-test, the resultant t maps are almost identical for m* and z*.

 

Yufeng

 

From:

cherrychingju

Tue, 10/07/2014 - 00:04

If all the values we extract are ReHo values, where can we extract KCC values ?

Dina Dajani

Tue, 10/07/2014 - 16:56

Hello Dr. Zang,

 

What would the difference be between performing a two sample t-test on the standardized maps versus using the original KCC-value maps?

 

Thank you for the information,

Dina Dajani

Dina Dajani

Tue, 10/07/2014 - 17:09

also to confirm, the global mean KCC value that is divided for these m* maps are gloabl means for each individual participant, correct? (not a global mean for the entire group)

ZangYF

Sat, 10/11/2014 - 13:40


The global mean value is used for standardization for many functional imaging techniques, e.g., PET. We do know the results will be different for standardization or not. But we do not know which is right. Either might be reasonable.

 

The global mean value here is at individual level.

 

Yufeng

 

From:

cherrychingju

Tue, 10/07/2014 - 00:02

Accroding to wiki, KCC values range from 0 to 1. Thus, KCC values seem not to be ReHo values! 

ZangYF

Tue, 10/07/2014 - 00:50


ReHo uses KCC. So the raw ReHo value is from 0 – 1. But, of course, other algorithms can also be used to measure ReHo, e.g., coherence (Liu et al., 2009).

 

Ref: Liu D, Yan C, Ren J, Yao L, Kiviniemi VJ, Zang Y. Using coherence to measure regional homogeneity of resting-state fMRI signal. Front. Syst. Neurosci. 2010, 4:24.

 

From:

cherrychingju

Tue, 10/07/2014 - 03:37

So KCC values can be extracted by using ReHoMap.nii to conduct one-sample t test (Further, generate ROI to extract)?

Thank you so much for your reply!

Cherry

ZangYF

Sat, 10/11/2014 - 13:40


ReHo is just a general term for local synchronization here. KCC, coherence, or linear correlation are the algorithms.

 

The original value will never below 1. But after standardization, the value may be any big, depending on the standardization method.

 

Yufeng

 

From: