Hi Hussain,
our LME is specifically targeted at the mass-univariate setting where we have thousands of measurements on the cortical surface of the brain. From what you describe below, I think it will be sufficient to use any linear mixed model implementation (e.g. Matlab statistic toolbox, R, or whatever statistical package you choose) to run the analysis. It would be best to involve a local bio-statistician in your research for advise on that. They usually have their preferred tools.
Best, Martin
On 07/18/2013 02:26 PM, hussain fatakdawala wrote:
Hi,
I am a Ph.D student at UC Davis in Laura Marcu's Lab. We are imaging ex-vivo human coronary arteries using two simultaneous imaging techniques (fluorescence and ultrasound).
For each image we extract 58 parameters from regions of interest that have been identified with a known disease condition from pathology. Each coronary sample can have multiple disease conditions. This is because we have multiple image slices along the length of the coronary and each slice may present different disease conditions.
As of now, we have 14 disease conditions.
We want to be able to statistically determine how well the different 58 parameters distinguish between different disease conditions.
We do not know if the data distribution is normal and if the parameters are independent of one another.
We were advised by a statistician to use mixed effects anova and directed us to your work as our data is longitudinal (using space instead of time).
Can you please guide us on how we can use your software for our analysis since your software seems to be tailored for neuro-imaging data.
Thanks much.
Hussain Fatakdawala http://www.bme.ucdavis.edu/marculab/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer