I would avoid using the term "beta" and instead just say "contrast" or
"BOLD difference".
In some cases, the individual beta's can be high, although this is
unusual. One cause of high betas is collinearity in the design
matrices.
Best Regards, Donald McLaren
=================
D.G. McLaren, Ph.D.
Postdoctoral Research Fellow, GRECC, Bedford VA
Research Fellow, Department of Neurology, Massachusetts General Hospital and
Harvard Medical School
Website: http://www.martinos.org/~mclaren
Office: (773) 406-2464
=====================
This e-mail contains CONFIDENTIAL INFORMATION which may contain PROTECTED
HEALTHCARE INFORMATION and may also be LEGALLY PRIVILEGED and which is
intended only for the use of the individual or entity named above. If the
reader of the e-mail is not the intended recipient or the employee or agent
responsible for delivering it to the intended recipient, you are hereby
notified that you are in possession of confidential and privileged
information. Any unauthorized use, disclosure, copying or the taking of any
action in reliance on the contents of this information is strictly
prohibited and may be unlawful. If you have received this e-mail
unintentionally, please immediately notify the sender via telephone at (773)
406-2464 or email.
On Mon, Apr 9, 2012 at 3:21 AM, xiuqin jia <[log in to unmask]> wrote:
> Dear Mclaren Donald,
>
> Thanks very much for your kind response.
>
>
> 2012/4/6 MCLAREN, Donald <[log in to unmask]>
>>
>> See inline responses below.
>>
>> On Mon, Apr 2, 2012 at 10:55 PM, xiuqin jia <[log in to unmask]> wrote:
>> > Dear experts,
>> >
>> > Here, I have two question about Beta value. I have searched it in the
>> > SPM
>> > archive, but it seems I cannot understand it completely. So I post it
>> > here.
>> >
>> >
>> > 1) Is there a plausible range of beta value?
>>
>> There are several issues with this question. (1) beta values refer to
>> the estimates of each column of your design matrix; (2) when people
>> report "beta" they are usually referring to their contrast; (3) it is
>> also dependent on the scaling of the column; (4) it is dependent on
>> the scaling of the data (e.g. in SPM the data are scaled to 1000, I
>> believe).
>>
>> Because every column has a beta estimate, there isn't a specific
>> range. The beta's for tasks are usually very low (in the range of a
>> few percent). The beta for the constant is dependent on the scanner,
>> patient position, blood flow, scaling, among other factors. In papers,
>> you have to consider how the contrast was designed:
>> (a) 1s over the columns -- sum of the betas
>> (b) 1/N over the columns -- mean of the betas
>>
>> >
>> > In my experiment, there are 4 conditions: SI, CI, SA, CA. I extracted
>> > the
>> > beta values of the 4 conditions in bilateral DLPFC (dorsolateral
>> > prefrontal
>> > cortex) and PPC (posterior parietal cortext). However, It seems a bit
>> > larger
>> > than ones in the published papers. I have attached the figure.
>> >
>> > I'm not sure, whether is it plausible?
>>
>> What was your contrast vector?
>
> In my design matrix, the regressors from left to right are SI, CI, SA, CA.
> My contrast vector is c=[-1 1 0 0]', i.e., CI-SI. However, the extracted
> beta values in my defined ROI don't vary with the contrast. For example in
> the ROI the left PPC, the beta values are 12.79 for SI, 14.34 for CI, 12.98
> for SA, 14.92 for CA. As you mentioned above "The beta's for tasks are
> usually very low (in the range of a few percent)", but the beta values for
> the four conditions are more than 10, are they plausible?
>
> Can I say the beta value of the contrast of (CI-SI) (i.e., [-1 1 0 0]) is
> (14.34-12.79=1.55)?
>
>
> Thank you again.
>
>
> xiuqin
>
>>
>> >
>> > 2) Does the negative beta value indicate deactivation?
>>
>> Yes. Less activity than baseline.
>>
>> >
>> > The positive beta value can be understood as more active, can the
>> > negative
>> > beta value be understood as negative activation? How to explain the
>> > negative
>> > beta value?
>> >
>> >
>> > Any comments are deeply appreciated. Thanks a million!
>> >
>> > Best Regards,
>> >
>> >
>> > xiuqin jia
>> >
>> >
>> >
>> > -----------------------------------------------
>> >
>> > Xiuqin Jia, Ph.D Candidate
>> > Department of Radiology,
>> > Xuanwu Hospital,
>> > Capital Medial University,
>> > 45 Changchun Street,
>> > Xuanwu District,
>> > Beijing, 100053 China
>> > Tel: (+086)-138-1012-5607
>> >
>> >
>
>
>
>
> --
>
> Xiuqin Jia, Ph.D Candidate
> Department of Radiology,
> Xuanwu Hospital,
> Capital Medial University,
> 45 Changchun Street,
> Xuanwu District,
> Beijing, 100053 China
> Tel: (+086)-138-1012-5607
>
>
|