ok, found it. It's very, very low contrast. What flip angle and TE did you use. I can't visually distinguish gray from white in many regions such as motor cortex, so I can't imagine that you can get good thickness estimates over much of the brain. You could still use them for functional analysis, or possibly for thickness in other regions that have longer T1 such as frontal or anterior temporal. You can try creating an expert options file and adding the entry:
mri_normalize -gentle
then running
recon-all -expert expert.opts -s sub -all
cheers Bruce
On Mon, 21 May 2012, Meng Li wrote:
the name of the file uploaded is sub.rar.
--
At 2012-05-21 21:02:46,"Bruce Fischl" fischl@nmr.mgh.harvard.edu wrote:
what is the name of the file you uploaded? On Mon, 21 May 2012, Meng Li wrote:
Hi professor, The subject has been uploaded. Can you give me some advice to troubleshoot the outputs. Thanks. Meng Li -- At 2012-05-20 21:27:28,"Bruce Fischl" fischl@nmr.mgh.harvard.edu wrote: If you put the gzipped and tarred subject dir on our FTP site I'll take a look Cheers Bruce On May 20, 2012, at 12:41 AM, ?? limengsecret@126.com wrote:
Hi professor, The attachment is the subject after running the recon-all command. MR data were obtained on a 1.5T Philips Achieva Nova Dual MR scanner. the sequence name is T1W.3D.FFE.SENSE and the voxel size is 0.90*0.90*1mm. how should I fix the poor quality surface reconstruction? Thanks a lot for your help. Meng Li At 2012-05-19 23:10:31,"Bruce Fischl" fischl@nmr.mgh.harvard.edu wrote: >p.s. looks like an intensity normalization failure. Can you give us details >of the acquisition? Sequence name, voxel resolution, etc....? If you upload >the subject I'll take a look > >cheers >Bruce > >On Sat, 19 May 2012, >limengsecret wrote: > >> Hi Professor, >> The image blow shows the problems which occur. >> color lines representing the surfaces (pial and white) were not along the >> folding pattern, and the distance between the two sufaces was rather short. >> How to deal with these problems? >> [IMAGE] >> Thanks. >> >> Meng Li >> >> >> ____________________________________________________________________________ >> limengsecret >> >> From: Bruce Fischl >> Date: 2012-05-19 22:08 >> To: ?? >> CC: freesurfer >> Subject: Re: [Freesurfer] how to edit the pial surface >> Hi Meng >> >> 1. Not positive I understand, but I don't think it should matter. >> >> 2. we'll need to see some data to diagnose this. >> >> cheers >> Bruce >> >> >> On Sat, 19 May 2012, ?? wrote: >> >> > >> > Hi FreeSurfer Experts, >> > >> > I had 2 question about the output troubleshooting. >> > >> > ? >> > >> > 1)?? when I used the mri_convert command, the format of the input image I >> > choosed was 4D nii format, but not the original dicom format, which would >> > affect the result of reconstruction? >> > >> > 2)?? When I checked the quality of output using tksurfer, I found the >> > color line representing the surface (pial and white) were not along the >> > folding pattern, and the distance between the two sufaces was rather shor >> t. >> > Whether or not it was caused by the bad GM/WM contrast? What is the best >?? ?way >> > to deal with them? >> > >> > ? >> > >> > Thanks. >> > >> > ? >> > >> > Meng Li >> > >> > >> > >> > >> > >> >> >> The information in this e-mail is intended only for the person to whom it i >> s >> addressed. If you believe this e-mail was sent to you in error and the e-ma >> il >> contains patient information, please contact the Partners Compliance HelpLi >> ne at >> http://www.partners.org/complianceline%C2%A0.%C2%A0If%C2%A0the%C2%A0e-mail%C... >> rror >> but does not contain patient information, please contact the sender and pro >> perly >> dispose of the e-mail. >> >> >> __________________________________________________________________________________________________________________________________
[icon_att.gif] ?QQ????????? [fu_rar.gif] Sub.rar (262MB, 2012?06?19? 12:40 ??)???????http://mail.qq.com/cgi-bin/ftnExs_download?k=0d636362224297d81b9673551e64001...
c520a561c5055060a49060409534e5a0f02041a535400500e0051
070753545b3863326444014d105916320a&code=1ccb8d27&t=exs_ftn_download