 |
 |
I would look at the SPM.mat files produced on different computers. In looking at the different fields, you will be able to see what is different between your analyses. 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 Tue, May 15, 2012 at 7:14 PM, Anthony Ang <[log in to unmask]> wrote: > Dear Simon, > > Thank you very much for your help and suggestions. > > We had tried a larger sample size (1 patient vs 20 controls) on our > dataset. This produced white/grey boxes on the Win PC installed with > matlab2008/spm5 and black boxes on other computers. The screenshots were > based on a previous research dataset that showed a definite temporal > cluster in 4 patients, particularly in 1 patient. Hence, 5 controls were > compared against 4 patients (pg 1) or 1 patient (pg 2, 1st Fig) to validate > whether the steps are working or not. Identical steps, design matrix (-1 > 1) and unequal variances for both patient and control groups were performed > in all cases. The outcome is that we consistently obtained white/grey boxes > on that "special" computer but black boxes on other computers. Our aim is > to get rid of those black boxes first before talking of getting any result > from the cluster analysis. > > I don't really understand the GUI issue you have raised. > > I am looking forwards to your further suggestions or suggestions from > anyone on helping to resolve this problem. Thank you very much. > > Best regards > Anthony > > > On Fri, May 11, 2012 at 7:30 PM, Eickhoff, Simon <[log in to unmask] > > wrote: > >> Dear Anthony >> >> >> >> >> >> There are several aspects that may contribute to your problem. >> >> >> >> >> >> First of all, your sample is very small, containing only 4/5 subjects per >> group, which obviously compromises the reliability of your results. If you, >> in such small sample, find that changing the design / software version >> slightly, you should be very cautious. >> >> >> >> >> >> The incongruency between the images on the first side should indeed >> relate to the "white/grey" boxes. These relate to the non-sphericity >> correction when estimating the design. Is it possible, that in one version >> you modelled unequal variances, in the other you didn't? >> >> >> >> The version shown on the second page of your document seems quite >> different: Here the first group only contains a single subject, whereas the >> second group contains 8 subjects. This seems a mistake. >> >> >> >> Finally, the error on the second slide (right) seems to relate to the GUI >> rather than an error in estimation. >> >> >> >> >> >> Best regards >> >> Simon >> >> >> >> >> >> >> ------------------------------ >> *Von:* SPM (Statistical Parametric Mapping) [[log in to unmask]]" im >> Auftrag von "Anthony Ang [[log in to unmask]] >> *Gesendet:* Freitag, 11. Mai 2012 07:01 >> *An:* [log in to unmask] >> *Betreff:* Re: [SPM] Problem with 2-Sample T-Test Analysis on PET images >> >> Dear SPM Experts, >> >> I encounter problems replicating the result of 2-Sample T-Test Analysis >> on PET images. The PET images are normalized and subjected to T-Test using >> SPM5 and Matlab R2008b on a 32-bit WinXP Desktop. This produced a single >> cluster in the temporal lobe of patients (with p=0.005 and voxels=50) and >> white/grey boxes in the Design Matrix. Please refer to the pdf of >> screenshots here: >> https://docs.google.com/open?id=0B7LWtFTZ-QYiNlRadm1BZTNySUU >> >> The same results cannot be replicated on other computers (64-bits Linux, >> Mac, Windows) with the same dataset. We have tried SPM5, SPM8, standalone >> SPM8, Matlab 2008/2009/2012, the latest software versions, updates and >> patches for both SPM and Matlab but still could not reproduce the results. >> There is no single cluster in the temporal lobe. Also black boxes (plus >> white/grey boxes) are obtained instead of the white and grey boxes. We are >> told that something is wrong if black boxes are obtained. The person who >> obtained the temporal lobe cluster on the 32-bits, WinXP desktop also >> obtained the same result on her 64-bits Win7 laptop in which SPM5 and >> matlab2008 are installed. She failed to get the results when using SPM8 or >> ( SPM5+matlab2009) which is why she used SPM5 with matlab2008. We need to >> get this working on the Mac computers we are working with. >> >> Will greatly appreciate help to troubleshoot the problems as we have >> tried many things but still could not figure out what went wrong. Besides >> the screenshots, I have also attached the mat files for your reference >> below. Thank you very much. >> >> Links for the mat files: >> (1) spm5_matlab2008_winXPDesktop: >> https://docs.google.com/open?id=0B7LWtFTZ-QYiZnlCQ2RoM2tXbWc >> (2) spm8_matlab2012_win7Desktop: >> https://docs.google.com/open?id=0B7LWtFTZ-QYiY1lZbDV4c0xZNUk >> (3) spm8_matlab2012_mac: >> https://docs.google.com/open?id=0B7LWtFTZ-QYiX01pNVpjRDA1RkU >> >> >> Best regards >> Anthony Ang >> >> >> >> ------------------------------------------------------------------------------------------------ >> >> ------------------------------------------------------------------------------------------------ >> Forschungszentrum Juelich GmbH >> 52425 Juelich >> Sitz der Gesellschaft: Juelich >> Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 >> Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher >> Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender), >> Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, >> Prof. Dr. Sebastian M. Schmidt >> >> ------------------------------------------------------------------------------------------------ >> >> ------------------------------------------------------------------------------------------------ >> >> Kennen Sie schon unseren neuen Film? http://www.fz-juelich.de/film >> Kennen Sie schon unsere app? http://www.fz-juelich.de/app >> > >