Mean FA of a Tract | TrackVis Usage

Mean FA of a Tract

Page: 1

User

Post

3:57 pm
December 30, 2011

cibuthomas

Member

posts 16

Print this Post

Hi Roupeng

I noticed that the mean FA value of a tract reported in the Track statistics window is different from the mean of the FA values computed from the data exported using the 'more stats' tool. Could you please explain how the mean FA is calculated? I can upload you the .csv files if you'd like.

Thank you

Cibu

11:38 am
January 4, 2012

Ruopeng

Admin

posts 406

Print this Post

How different? Can you post the two values?

Ruopeng

1:55 pm
January 5, 2012

cibuthomas

Member

posts 16

Print this Post

Hi Roupeng

the FA values that I am reporting are derived by extracting the uncinate fasciculus in an adult subject using a multiple ROI appraoch. Trackvis reports the mean FA = 0.46 +/- 0.14

Now if I export the data into a csv file. I have the following values

AP_23_Only_R1_FA	Track Count
0.166471	5
0.182136	2
0.197802	5
0.213467	1
0.229132	4
0.244798	0
0.260463	1
0.276128	4
0.291794	2
0.307459	4
0.323124	9
0.33879	8
0.354455	5
0.37012	8
0.385786	11
0.401451	19
0.417116	34
0.432782	52
0.448447	58
0.464112	69
0.479778	73
0.495443	34
0.511108	19
0.526774	4
0.542439	1

The mean calculated from these FA values is 0.354455 std. dev 0.12

Can you help me understand why there is a discrpancy in the two means calculated?. Also, notice that one of the tract count values is 0, but there is an FA associated with this value. What does this mean?

Many thanks as always

Cibu

2:36 pm
January 5, 2012

Ruopeng

Admin

posts 406

Print this Post

When calculating mean FA, track length (or voxel count of the track) is used as weighting factor. As to why there is track cout of 0, remember this is a histogram. Count of 0 simply means there is no track falling in that FA range.

Ruopeng

4:50 pm
January 6, 2012

cibuthomas

Member

posts 16

Print this Post

Thanks Roupeng. That explains the differences in FA.

Page: 1

	You must be logged in to post	Login Register