Hi Ye

you need to give recon-all a T1-weighted anatomical (e.g. mprage, SPGR/FLASH, etc...), not the fMRI data.

cheers Bruce On Wed, 26 Jun 2013, ye tian wrote:

Dear Bruce,

Recon-all exits with the following error

ERROR: input(s) cannot have multiple frames! $SUBJECTS_DIR/Barba50527_2/mri/orig/001.mgz has 439 frames

If I open 001.mgz using freeview, I obtain the image in the attachment saved as freeview 001.mgz.png.

In order to make sure that my problem does not come from further upstream, I am including the major steps that I took to get to this error.

MAJOR STEPS

STEP0:  My inherited data has names BARBA50527.MR.MEDITATE_MEDITATION.#.#.year.month.day.#.#.#.#.#.IMA where each # is a string of numbers.

STEP1:  The images can be grouped into five categories, according to the numbers in the names that occupy the positions of the red #:    1.1-1.3,    2.1,   3.1-3.160,    4.1-4.439,    5.1-5.439

[1] Note: Images from groups 4 & 5 are very different from the images of groups 1, 2 and 3. A link to a  youtube video of all the images can be seen in note[1].

STEP2:  recon-all -s Barba50527_2                -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.4.1.2010.07.27.10.52.51                  .437500.56148038.IMA               -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.5.1.2010.07.27.10.52.51                  .437500.56148502.IMA recon-all completes without error.

[2] Note: other attempts leads to different errors. See note [2] for details.

STEP3: recon-all -s Barba50527_2 -autorecon1

Exits with the error message at the beginning of the email. The number 439 seems to come from the number of files in each group.

USEFUL COMMENTS

http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg23764.html

I will do it for 001.mgz and 002.mgz and will let you know how it goes.

QUESTIONS

1. Do I really need to manually parse the sequence of images? If so, how?
2. I don't really need to mri_convert my files into other formats, e.g.,

.dcm, do I? 3. Each file from groups 4 and 5 have multiple images. Do I need to cut them into sub-images, each with one brain?

Thank you very much!

Sincerely, Ye

NOTES: Note [1] A video of the data can be found at http://www.youtube.com/watch?v=W8ZFGvGvHVc&list=UUfGdQq3j1gqPAc4z6Spmu2g

On the top left corner:  the white text is:    current frame number / total frame number the yellow text is:   the name of the frame

Note[2] recon-all -s Barba50527_2                -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.1.1.2010.07.27.10.52.51                  .437500.56141898.IMA Completes without error recon-all -s Barba50527_2 -autorecon1 Warning: only one run found  Error: MRIsample(): source matrix has zero determinant; matrix is:  EOF

econ-all -s Barba50527_2                -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.1.1.2010.07.27.10.52.51                  .437500.56141898.IMA               -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.2.1.2010.07.27.10.52.51                  .437500.56148502.IMA               -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.3.1.2010.07.27.10.52.51                  .437500.56148502.IMA               -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.4.1.2010.07.27.10.52.51                  .437500.56148038.IMA               -i /Path_to_data/BARBA50527.MR.MEDITATE_MEDITATION.5.1.2010.07.27.10.52.51                  .437500.56148502.IMA error.log: matrix dimensionality