Frequently Asked Questions (FAQ)

This page provides a repository of frequently asked questions.

What should the dimensions of my image be?

  1. When performing the analysis of Patient_02, the initial dimensions of the ictal and interictal scans are, respectively, 128x128x43 and 128x128x41.
  2. After realignment (i.e. after completing step 1.3 on the ISAS website), the dimensions of the ictal and interictal scans are, respectively, 128x128x41 and 128x128x41.
  3. After completing normalization (i.e . after completing step 1.4 on the ISAS website), the dimensions of the ictal and interictal scans are, respectively, 91x109x91 and 91x109x91.
  4. The image dimensions do not change after masking or smoothing (steps 1.5 and 1.6).
if your analysis produces different dimensions after any of the above steps, some possible solutions...
  1. Please make sure that your spm_normalize_ui.m file has been modified per the directions at
  2. Please make sure that line 15 in your healthy_normals_location.m actually points to the healthy normal database downloaded from ISAS ( e.g. 'C:\ISAS\Healthy_Normals\').
  3. In general, when looking for files using the SPMget window, make sure to clear the view filter so that you can see all files in the current folder.

I'm doing a group analysis, several of my scans have different dimensions, and I'm getting errors. What should I do?

It is likely that a group analysis encompassing several patients will include images of several different dimensions. SPM will occasionally have trouble with these images. One possible solution is to change the dimensions of all images used in your analysis to the dimensions of one image. This can be done using chg_vx.m which you can find on the downloads page. After typing chg_vx.m at the Matlab prompt, two SPMget windows will pop up. In the first window, please select the image whose dimensions you would like all other images to be transformed. In the second SPMget windows, select all images that you would like to transform. The resized images will have a 'i' at the end of their name (e.g. if Patient2_ICTAL.img is transformed, a new image called Patient2_ICTALi.img will be available.)

Do I need to flip dicom images before running it through SPM?

It is important to run the sample analyses and see if you get the same results as on website. Then as a double check of L-R with your own data, you should analyze a scan with an obvious L-R asymmetry so you can double check that L and R are coming out where they should. This is the only way to know for sure that you've got it right, and well worth checking for clinical data. Once you've got things set, then as long as your scans are always in the same format and you do the analyses the same way you should be all set.

How do I flip images before running ISAS?

One simple method of flipping images before using ISAS is to use spm_flip.m which you can find on the downloads page. After typing spm_flip in the matlab prompt, an SPMget window will pop. Select all images that you would like to flip. Select Done. All of the flipped images will have an 'f' before their names (e.g. if Patient2_ICTAL.img is flipped, a new image called fPatient2_ICTAL.img will be available.).

What is the difference between ISAS using SPM and BioImage Suite?

ISAS BioImage Suite provides a custom user interface for the analysis and eliminates any compatibility problems between different versions of MATLAB and SPM2. In addition, the registration algorithm in BioImage Suite has been shown to perform better on multimodal neuroimaging data when compared to the algorithm used by SPM2. The quality registration into common space is critical in the prevention of skull artifacts in the analysis. However, ISAS for both BioImage Suite and SPM follow the same processing step and should provide similar results for most studies.

Back to Home || Jump to Top || Forward to Links/References