External Email - Use Caution
Dear Freesurfer experts,
I would like to perform a correlation between the volume of subiculum and some behavioural measures. Before making that correlation, is it necessary to correct the subiculum volume with the eTIV? I have thought of doing it by scaling the subiculum volume as follows:
*subic_vol = mean_eTIV_group * subic_vol / eTIV_subject * Is correct??
mean_eTIV_group = eTIV average of the group to which the subject belongs.
Thanks in advance.
Best regards,
Steve.
Hi Steve,
It looks good except for using the mean eTIV of the subgroup. Please use the global average or don’t multiply by anything at all. Multiplying by the global average has the cosmetic effect of giving you a volume that is the range of subiculum volumes (rather than fractions <<< 1), but won’t change the results of your analysis. If you normalize each group separately, the question you’re answering changes (even though in practice, I suspect there won’t be large differences between the eTIV of the groups, particularly if you correct for gender).
You can also correct for eTIV with regression, by the way. In that case, I would suggest demeaning the eTIV first (for the same cosmetic purposes).
Cheers,
/Eugenio
Juan Eugenio Iglesias Senior research fellow CMIC (UCL), MGH (HMS) and CSAIL (MIT) http://www.jeiglesias.com
From: freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Steve Petersen marfneuro@gmail.com Reply-To: Freesurfer support list freesurfer@nmr.mgh.harvard.edu Date: Wednesday, April 22, 2020 at 11:36 To: "freesurfer@nmr.mgh.harvard.edu" freesurfer@nmr.mgh.harvard.edu Subject: [Freesurfer] Correlation between Hippocampal subfields - behavioural test
External Email - Use Caution Dear Freesurfer experts,
I would like to perform a correlation between the volume of subiculum and some behavioural measures. Before making that correlation, is it necessary to correct the subiculum volume with the eTIV? I have thought of doing it by scaling the subiculum volume as follows:
subic_vol = mean_eTIV_group * subic_vol / eTIV_subject Is correct??
mean_eTIV_group = eTIV average of the group to which the subject belongs.
Thanks in advance.
Best regards,
Steve.
External Email - Use Caution
Dear Eugenio,
Thank you for your quick response. Sorry but I am not completely sure if I understood correctly your response. What do you mean with "demeaning" the eTIV first?.
If I understood, a good choice would be to correct the subic_volume before perform the regression using the folliwng formula:
subic_vol = average_eTIV_allsample * subic_vol / eTIV_subject
Thank you.
Best,
Steve.
El mié., 22 abr. 2020 a las 17:35, Steve Petersen (marfneuro@gmail.com) escribió:
Dear Freesurfer experts,
I would like to perform a correlation between the volume of subiculum and some behavioural measures. Before making that correlation, is it necessary to correct the subiculum volume with the eTIV? I have thought of doing it by scaling the subiculum volume as follows:
*subic_vol = mean_eTIV_group * subic_vol / eTIV_subject * Is correct??
mean_eTIV_group = eTIV average of the group to which the subject belongs.
Thanks in advance.
Best regards,
Steve.
No no sorry for the misunderstanding. Either you correct by division with subic_vol = average_eTIV_allsample * subic_vol / eTIV_subject OR y0ou correct by regression. Only if you do regression, I’d use (eTIV_subject - average_eTIV_allsample) rather than eTIV_subject directly (and the same for all other variables: age, etc). Again: the result will be the same, but if you look at the residuals (plus the global intercept), demeaning will leave you with something that looks like the volume of a subiculum. Cheers, /Eugenio
Juan Eugenio Iglesias Senior research fellow CMIC (UCL), MGH (HMS) and CSAIL (MIT) http://www.jeiglesias.com
From: freesurfer-bounces@nmr.mgh.harvard.edu on behalf of Steve Petersen marfneuro@gmail.com Reply-To: Freesurfer support list freesurfer@nmr.mgh.harvard.edu Date: Wednesday, April 22, 2020 at 12:39 To: "freesurfer@nmr.mgh.harvard.edu" freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Correlation between Hippocampal subfields - behavioural test
External Email - Use Caution Dear Eugenio,
Thank you for your quick response. Sorry but I am not completely sure if I understood correctly your response. What do you mean with "demeaning" the eTIV first?.
If I understood, a good choice would be to correct the subic_volume before perform the regression using the folliwng formula:
subic_vol = average_eTIV_allsample * subic_vol / eTIV_subject
Thank you.
Best,
Steve.
El mié., 22 abr. 2020 a las 17:35, Steve Petersen (<marfneuro@gmail.commailto:marfneuro@gmail.com>) escribió: Dear Freesurfer experts,
I would like to perform a correlation between the volume of subiculum and some behavioural measures. Before making that correlation, is it necessary to correct the subiculum volume with the eTIV? I have thought of doing it by scaling the subiculum volume as follows:
subic_vol = mean_eTIV_group * subic_vol / eTIV_subject Is correct??
mean_eTIV_group = eTIV average of the group to which the subject belongs.
Thanks in advance.
Best regards,
Steve.
freesurfer@nmr.mgh.harvard.edu
-
Iglesias Gonzalez, Juan E. -
Steve Petersen