Dear Steve,
Thanks
No - I was referring to what you called "diagnostic criteria" as being ordinal (taking on several indicator values).
Yes you are right I have ordinal score ranging from 1 to 4 from your previous response to this " however if you have several different diagnostic criteria, and don't want to assume a linear relationship between the different indicator values, you should have a separate EV for each different value, and then do contrasts between these."
so in this case is it correct to have the following design EV1 (score1) EV2(score2) EV3(score3) EV4 (score4) so in this case is there a way to separate the controls from my "GM_mod_merg_s3" file and retain the same order of my patients data. Also I am interested in knowing whether there is a correlation in diagnostic criteria with the GM volume changes in each patient group so is it correct to use the following design EV1 (group1) EV2(group2) EV3(group3) EV4 (group4) EV5 (diagnostic criteria)
You can't - you would have to use a different model, with a separate EV for each group for each clinical measure.
so in this case my design goes like this EV1 (group1) EV2(group2) EV3(group3) EV4 (group4) EV5 (diagnostic criteria) right?
Thanks very much
Venkat
________________________________
From: FSL - FMRIB's Software Library on behalf of Stephen Smith
Sent: Fri 4/29/2011 1:48 AM
To: [log in to unmask]
Subject: Re: [FSL] VBM -Contrast matrix design for clinical measures
Hi
On 28 Apr 2011, at 14:41, Rajagopalan, Venkateswaran wrote:
All the above looks fine I think - however if you have several different diagnostic criteria, and don't want to assume a linear relationship between the different indicator values, you should have a separate EV for each different value, and then do contrasts between these.
I don't have several different diagnostic criteria but several clinical measures namely duration of disease, then a diagnostic criteria to tell about the disease stage and disease progression rate to tell how fast disease progressess.
No - I was referring to what you called "diagnostic criteria" as being ordinal (taking on several indicator values).
So I chose a separate EV for the above 3 different clinical measures
My first question is are the changes seen in GM volume in patients ( all patient groups considered as single patient group here originally there are 4 different patient groups based on their clinical phenotype) correlate with the above clinical measures.
So for this is the following design correct
0 0 0 0 0 1 0 0 ----for correalting clinical measure of disease duartion with GM volume changes (Where Ev1 is control group, EV2-EV5 patient groups, EV6 -disease duration, EV7-diagnostic criteria and EV8 disease progression rate)
0 0 0 0 0 0 1 0 ----for correalting clinical measure of diagnostic criteria with GM volume changes
0 0 0 0 0 0 0 1----for correalting clinical measure of disease progression rate with GM volume changes
If I see a correlation between GM volume changes with any of the clinical measures (in my above step) then as a second step I want to do a correlation of the clinical measures with the 4 patient groups separately so in that case how can I set the contrast matrix.
You can't - you would have to use a different model, with a separate EV for each group for each clinical measure.
Cheers.
Thanks
Regards
Venkat
________________________________
From: FSL - FMRIB's Software Library on behalf of Stephen Smith
Sent: Thu 4/28/2011 12:59 AM
To: [log in to unmask]
Subject: Re: [FSL] VBM -Contrast matrix design for clinical measures
Hi,
On 27 Apr 2011, at 22:21, Rajagopalan, Venkateswaran wrote:
Hi All,
I did FSL-VBM analysis on my T1 w images. I have a control group and 4 patient groups. I also have clinical measures namely disease duration, diagnostic criteria and disease progression rate. Of these clinical measures diagnostic criteria is ordinal (for instance 1 for possible, 2 for probable etc). I read archives related to GLM design and understood how to prepare my design matrix etc. I just would like to make sure that my design matrix is correct for the following investigation. I am interested in looking for any GM difference between control and patients and also interested in correlation of clinical measures with changes in GM volume in patients.
My EVs are EV1 (control), EV2 (patient group 1),....EV5(patient group4), EV6(demeaned disease duration values in all patients and zero for controls), EV7( demeaned diagnostic criteria) and Ev8 (demeaned disease progression rate). So my contrast matrix I as follows
Control >patient group1 1 -1 0 0 0 0 0 0
Control > patient group2 1 0 -1 0 0 0 0 0
.
.
.
Similarly I designed it for control vs other patient groups but for clinical measures correlation with patient groups GM changes I am not clear how to design the contrast matrix is it like this
Clinical measure disease duration changes with all patients (in the sense considering the 4 different groups as a single group)
0 0 0 0 0 1 0 0
Then For Clinical measure of diagnostic criteria
0 0 0 0 0 0 1 0
All the above looks fine I think - however if you have several different diagnostic criteria, and don't want to assume a linear relationship between the different indicator values, you should have a separate EV for each different value, and then do contrasts between these.
And demeaned disease progression rate
0 0 0 0 0 0 0 1
If I want to compare each and every patient group with clinical measure then is the following a correct way to design the matrix
disease duration
0 1 1 1 1 -1 0 0
I don't think this (or the below) makes sense - you are contrasting the mean patient level with the duration correlation - what question did you intend to ask?
Cheers
diagnostic criteria
0 1 1 1 1 0 -1 0
demeaned disease progression rate
0 1 1 1 1 0 0 -1
Thanks
Venkat
===================================
Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals
in America by U.S.News & World Report (2010).
Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/> <http://www.clevelandclinic.org/> for
a complete listing of our services, staff and
locations.
Confidentiality Note: This message is intended for use
only by the individual or entity to which it is addressed
and may contain information that is privileged,
confidential, and exempt from disclosure under applicable
law. If the reader of this message is not the intended
recipient or the employee or agent responsible for
delivering the message to the intended recipient, you are
hereby notified that any dissemination, distribution or
copying of this communication is strictly prohibited. If
you have received this communication in error, please
contact the sender immediately and destroy the material in
its entirety, whether electronic or hard copy. Thank you.
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
===================================
Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals
in America by U.S.News & World Report (2010).
Visit us online at http://www.clevelandclinic.org <http://www.clevelandclinic.org/> for
a complete listing of our services, staff and
locations.
Confidentiality Note: This message is intended for use
only by the individual or entity to which it is addressed
and may contain information that is privileged,
confidential, and exempt from disclosure under applicable
law. If the reader of this message is not the intended
recipient or the employee or agent responsible for
delivering the message to the intended recipient, you are
hereby notified that any dissemination, distribution or
copying of this communication is strictly prohibited. If
you have received this communication in error, please
contact the sender immediately and destroy the material in
its entirety, whether electronic or hard copy. Thank you.
---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director, Oxford University FMRIB Centre
FMRIB, JR Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------
===================================
Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals
in America by U.S.News & World Report (2010).
Visit us online at http://www.clevelandclinic.org for
a complete listing of our services, staff and
locations.
Confidentiality Note: This message is intended for use
only by the individual or entity to which it is addressed
and may contain information that is privileged,
confidential, and exempt from disclosure under applicable
law. If the reader of this message is not the intended
recipient or the employee or agent responsible for
delivering the message to the intended recipient, you are
hereby notified that any dissemination, distribution or
copying of this communication is strictly prohibited. If
you have received this communication in error, please
contact the sender immediately and destroy the material in
its entirety, whether electronic or hard copy. Thank you.
|