Hi Andrew, Thank you for your very helpful tutorials. I have been trying to create a mask as per JHU-ICBM-DTI-81 atlas in fslview. But i found that the probability maps are not displayed in FSL view for this atlas (they are displayed for other atlases like JHU White-matter-tractography atlas) and hence "save as" saves all the tracts instead of only the tract of interest. Is there any way to create a mask as per JHU-ICBM-DTI-81 atlas ? thank you.
Hi Andrew! Thanks for your help, I've used these instructions to save ROIs from fsl with no problem. I'm trying to save ROIs from functional analysis but only the top ~20% of activations within a cluster. The Marsbar 'get SPM cluster' automatically writes ROIs as binary. The SPM 'save current cluster' seems to maintain statistics but fslmaths thresholding doesn't work with this (I've tried, 10, 50, 99!), any suggestions? Many thanks!
Hi Laurel, It can get pretty tricky trying to transfer data across packages; I know! One way around this may be to first save the binarized image with SPM, and then multiply that by the contrast image you are working with. That will give you voxels with different values, but only within that ROI. Then try the FSL tools on that. Best, -Andy
Hi Andrew! Thank you for your detailed tutorials! I'm trying to save the regions, but the result is always the binary mask, while I want the non-binary mask of that region. Could you please help me with that? Thanks.
Hi Noor, I would use fsleyes, as that is a more recent imaging viewer. When you load a probabilistic atlas and save a particular region, it should be non-binary. Best, -Andy
@@AndrewJahn Hi Andy! Thank you so much for your reply! I am using fsleyes, and yes, the region is non-binary, but I can not save it as is. When I save it, only the binary region gets saved, which is just the mask. Do you happen to know a way for me to get the non-binary region saved? Thanks!
Hi Andy, Greetings, Your tutorials helps me a lot, as I do not have any background of computing skills. I'm much interested in DTI analysis. I have used ExploreDTI to compute DTI scalars. Yet, I would like to work with FSL. As per the instructions given in your DTI analysis tutorial using 'dtifit' FA maps are generated. In ExploreDTI it is possible to extract DTI Scalar values for a particular ROI (say, Corpus Callosum Splenium). Can we do this - extracting the FA values for a particular ROI in the FA map generated using FSL by using the above posted video? - Joy
Hey Joy, I don't see why not. It's not the intuitive to extract data from an ROI using featquery, since it requires a FEAT directory; I would look into using a tool like AFNI's 3dmaskave (which works with NIFTI images). Best, -Andy
Hi Andrew, thank you for your tutorials, they're wonderful! I have a question regarding fsl atlases in general. Do you know if there is a way to add my own atlas (downloaded from PET analysis software and it's in MNI space) within fsl?
Hey Jlingo, I haven't done that before, but my guess is that you'll have to include it in a directory within the $FSLDIR/data/atlases folder. You'll probably need to create an XML-linked file for it as well, describing whether it is probabilistic or a label atlas, what the labels refer to, etc. For the technical details I would write to the FSL listserv. Best, -Andy
Hi, Andrew, Your video series are awesome, very helpful to my research! However, I met some problems in creating the masks from Atlas by using my own image. When I used the standard images from the database, I got no problem but whenever I used my own image, it cannot mask the area appropriately. Have you met any problems for creating masks for certain images, what problems would it be for your guess? Also, is there any easy tool that you can save the coordinates of those masked voxels? Thank you so much for your help!
+Qiongge Li Hi Qiongge, Hey, thanks! It's good to know that they help. What kind of image are you trying to use? Is it a contrast from another analysis, or something else? For saving coordinates, I don't know of any FSL tool that allows you to do that; in AFNI there is 3dmaskave and 3dROIstats, which, if used with the -verbose option (which I think is the default), it outputs coordinates along with whatever information you are dumping from the mask. Best, -Andy
Thank you so much for the help, Andrew! I am going to try it out. For the mask creating, I had the .nii anatomical fMRI data, but I got error creating masks like attached image. I do not know what has happened. Another question is that when you save the mask, it gives you a .nii file but I could not save it as .nii but a image file which is strange. Do you know why? Thank you again for all your help!
+Andrew Jahn I am sorry I just realized that there is no way to upload a screenshot here. By the way, is this an appropriate place to ask questions, I saw you have a blog and a youtube channel~
+Qiongge Li Hi Qiongge, yes, people ask questions here all the time! If you want to send me personal files, you can send it to Andrew dot Jahner at gmail dot com. I have no idea whether typing it like that prevents spam robots from figuring out what my actual email address is, but try it. -Andy
Hey Mahsa, It has a Talairach daemon atlas (open up a standard template, then click on Tools->Toolbars->Atlas Tools, and then the Structures button that appears near the bottom of the screen). However, you could also probably create masks from SPM and as long as they have the same orientation and dimensions as the FSL templates (which I think they should, if they're both in MNI space), it should work. -Andy
Hi SBUPM, I'm not familiar with white matter tractography, so I cannot say too much about that. However, I do not believe that there are any hard and fast rules for what thresholds to use when creating masks from atlases; just be able to defend why you choose to use either a more liberal or more conservative threshold. Best, -Andy
Hi Andrew, I'm trying to create a mask analogous to your video... however, every time I use "save as" to create the ROI, it saves as a picture file and not .nii.gz. Any thoughts as to why this may be? Thank you, Hayley
Hi Andrew, Thanks for the tutorial its very usefull, i have a question, i need create a mask with a standard MNI_T1_3mm_Brain, so i cant create the mask with the atlas that come in default fsl, can you help me? Thanks!
Hi Alexis, Where did you get the standard MNI_T1_3mm_Brain template? I don't see that in FSL's data/standard directory; did you create it yourself? What exactly are you creating? If it's a mask that you want to resample to a different voxel size, use flirt with your smaller-voxel image as a reference: flirt -in 3mm_mask -out 3mm_to_2mm_mask.nii.gz -ref 2mm_template.nii.gz -applyxfm -init transformMatrix.mat -Andy
Hi Andy I get the MNI 3mm brain out of fsl in internet because the directory of that template doesn't exist. I wanna create a binary ROI using the fsl atlas with 3mm brain template So, its posible resample a roi of 2mm to 3mm using the reference template 3mm? Thanks
Hello, I followed your way and created a mask from ''Tapetum L'' of JHU ICBM-DTI-81 White-Matter Labels atlas. After that, when I opened the mask that I created, I noticed it wasn't the same region that I saw before I saved it. Can it be a bug? I checked the coordinate systems and something that came my mind. However, I didn't understand what it was occured. Can I create mask with script and how?
Hi Siim, What does the output look like? FSL will save whatever is highlighted in the GUI, so make sure to click on it before selecting "Save As." Best, -Andy
I noticed the problem, it doesn't related to FSL GUI. ''JHU ICBM-DTI-81 White-Matter Labels atlas'' shows all regions and only a mask can be generated by save as. I have to seperate the regions of the mask by looking the intensity value because those values give us the region names and locations written in xml file. Finally, I understood JHU ICBM-DTI-81 White-Matter Labeling approach even if it is different from MNI structural atlas labeling. Thank you for your answer.
Hi Amartei, sometimes those fsleyes issues are difficult to debug; just to make sure, are you using the more up-to-date fsleyes viewer, or fslview? I would recommend upgrading to the former if possible. Best, -Andy
@@AndrewJahn Thank you for the reply, I tried downloading the new fal but it will not allow me to share my directory and I have tried absolutely everything. I also don't know who to reach out to (FSL support team) or don't see any contact information.
you're the bomb. please never take down these vids
Very helpful, thank you very much for these FSL-tutorials.. A
Hi Andrew, Thank you for your very helpful tutorials. I have been trying to create a mask as per JHU-ICBM-DTI-81 atlas in fslview. But i found that the probability maps are not displayed in FSL view for this atlas (they are displayed for other atlases like JHU White-matter-tractography atlas) and hence "save as" saves all the tracts instead of only the tract of interest. Is there any way to create a mask as per JHU-ICBM-DTI-81 atlas ? thank you.
great tutorial
Hi Andrew! Thanks for your help, I've used these instructions to save ROIs from fsl with no problem. I'm trying to save ROIs from functional analysis but only the top ~20% of activations within a cluster. The Marsbar 'get SPM cluster' automatically writes ROIs as binary. The SPM 'save current cluster' seems to maintain statistics but fslmaths thresholding doesn't work with this (I've tried, 10, 50, 99!), any suggestions? Many thanks!
Hi Laurel,
It can get pretty tricky trying to transfer data across packages; I know! One way around this may be to first save the binarized image with SPM, and then multiply that by the contrast image you are working with. That will give you voxels with different values, but only within that ROI. Then try the FSL tools on that.
Best,
-Andy
Thanks a lot !!!! Really helpfull
Hi Andrew! Thank you for your detailed tutorials! I'm trying to save the regions, but the result is always the binary mask, while I want the non-binary mask of that region. Could you please help me with that? Thanks.
Hi Noor,
I would use fsleyes, as that is a more recent imaging viewer. When you load a probabilistic atlas and save a particular region, it should be non-binary.
Best,
-Andy
@@AndrewJahn Hi Andy! Thank you so much for your reply! I am using fsleyes, and yes, the region is non-binary, but I can not save it as is. When I save it, only the binary region gets saved, which is just the mask. Do you happen to know a way for me to get the non-binary region saved? Thanks!
Hi Andy,
Greetings,
Your tutorials helps me a lot, as I do not have any background of computing skills. I'm much interested in DTI analysis. I have used ExploreDTI to compute DTI scalars. Yet, I would like to work with FSL. As per the instructions given in your DTI analysis tutorial using 'dtifit' FA maps are generated. In ExploreDTI it is possible to extract DTI Scalar values for a particular ROI (say, Corpus Callosum Splenium). Can we do this - extracting the FA values for a particular ROI in the FA map generated using FSL by using the above posted video?
-
Joy
Hey Joy, I don't see why not. It's not the intuitive to extract data from an ROI using featquery, since it requires a FEAT directory; I would look into using a tool like AFNI's 3dmaskave (which works with NIFTI images).
Best,
-Andy
Hi Andrew, thank you for your tutorials, they're wonderful! I have a question regarding fsl atlases in general. Do you know if there is a way to add my own atlas (downloaded from PET analysis software and it's in MNI space) within fsl?
Hey Jlingo, I haven't done that before, but my guess is that you'll have to include it in a directory within the $FSLDIR/data/atlases folder. You'll probably need to create an XML-linked file for it as well, describing whether it is probabilistic or a label atlas, what the labels refer to, etc.
For the technical details I would write to the FSL listserv.
Best,
-Andy
Good idea, thanks for getting back to me!
Hi, Andrew,
Your video series are awesome, very helpful to my research! However, I met some problems in creating the masks from Atlas by using my own image. When I used the standard images from the database, I got no problem but whenever I used my own image, it cannot mask the area appropriately. Have you met any problems for creating masks for certain images, what problems would it be for your guess?
Also, is there any easy tool that you can save the coordinates of those masked voxels?
Thank you so much for your help!
+Qiongge Li Hi Qiongge,
Hey, thanks! It's good to know that they help.
What kind of image are you trying to use? Is it a contrast from another analysis, or something else?
For saving coordinates, I don't know of any FSL tool that allows you to do that; in AFNI there is 3dmaskave and 3dROIstats, which, if used with the -verbose option (which I think is the default), it outputs coordinates along with whatever information you are dumping from the mask.
Best,
-Andy
Thank you so much for the help, Andrew! I am going to try it out. For the mask creating,
I had the .nii anatomical fMRI data, but I got error creating masks like attached image. I do not know what has happened.
Another question is that when you save the mask, it gives you a .nii file but I could not save it as .nii but a image file which is strange. Do you know why?
Thank you again for all your help!
+Andrew Jahn I am sorry I just realized that there is no way to upload a screenshot here. By the way, is this an appropriate place to ask questions, I saw you have a blog and a youtube channel~
+Qiongge Li Hi Qiongge, yes, people ask questions here all the time! If you want to send me personal files, you can send it to Andrew dot Jahner at gmail dot com. I have no idea whether typing it like that prevents spam robots from figuring out what my actual email address is, but try it.
-Andy
Dear Andrew
Does FSL contain BA3 and BA4 areas completely(like spm)?
Could I Create Masks from Atlas in SPM?
Thank you in advance for your help.
Hey Mahsa,
It has a Talairach daemon atlas (open up a standard template, then click on Tools->Toolbars->Atlas Tools, and then the Structures button that appears near the bottom of the screen). However, you could also probably create masks from SPM and as long as they have the same orientation and dimensions as the FSL templates (which I think they should, if they're both in MNI space), it should work.
-Andy
Hi SBUPM,
I'm not familiar with white matter tractography, so I cannot say too much about that. However, I do not believe that there are any hard and fast rules for what thresholds to use when creating masks from atlases; just be able to defend why you choose to use either a more liberal or more conservative threshold.
Best,
-Andy
Hi Andrew,
I'm trying to create a mask analogous to your video... however, every time I use "save as" to create the ROI, it saves as a picture file and not .nii.gz. Any thoughts as to why this may be?
Thank you,
Hayley
THANKS SIR :) its a great help :)
Hi Andrew, Thanks for the tutorial its very usefull, i have a question, i need create a mask with a standard MNI_T1_3mm_Brain, so i cant create the mask with the atlas that come in default fsl, can you help me? Thanks!
Hi Alexis,
Where did you get the standard MNI_T1_3mm_Brain template? I don't see that in FSL's data/standard directory; did you create it yourself?
What exactly are you creating? If it's a mask that you want to resample to a different voxel size, use flirt with your smaller-voxel image as a reference:
flirt -in 3mm_mask -out 3mm_to_2mm_mask.nii.gz -ref 2mm_template.nii.gz -applyxfm -init transformMatrix.mat
-Andy
Hi Andy
I get the MNI 3mm brain out of fsl in internet because the directory of that template doesn't exist.
I wanna create a binary ROI using the fsl atlas with 3mm brain template
So, its posible resample a roi of 2mm to 3mm using the reference template 3mm?
Thanks
Yes, it is possible; try the command below. AFNI's 3dresample can also do the same thing, if you feel comfortable with using AFNI commands.
-Andy
Hello, I followed your way and created a mask from ''Tapetum L'' of JHU ICBM-DTI-81 White-Matter Labels atlas. After that, when I opened the mask that I created, I noticed it wasn't the same region that I saw before I saved it. Can it be a bug? I checked the coordinate systems and something that came my mind. However, I didn't understand what it was occured. Can I create mask with script and how?
Hi Siim,
What does the output look like? FSL will save whatever is highlighted in the GUI, so make sure to click on it before selecting "Save As."
Best,
-Andy
I noticed the problem, it doesn't related to FSL GUI. ''JHU ICBM-DTI-81 White-Matter Labels atlas'' shows all regions and only a mask can be generated by save as. I have to seperate the regions of the mask by looking the intensity value because those values give us the region names and locations written in xml file. Finally, I understood JHU ICBM-DTI-81 White-Matter Labeling approach even if it is different from MNI structural atlas labeling.
Thank you for your answer.
Good to hear!
Hello, everytime I attempt to select the probalistic mapping and add it on the FSL view crashes. How do I fix that?
Hi Amartei, sometimes those fsleyes issues are difficult to debug; just to make sure, are you using the more up-to-date fsleyes viewer, or fslview? I would recommend upgrading to the former if possible.
Best,
-Andy
@@AndrewJahn Thank you for the reply, I tried downloading the new fal but it will not allow me to share my directory and I have tried absolutely everything. I also don't know who to reach out to (FSL support team) or don't see any contact information.
@@Megafunnyhead1 Hi Amartei, I would check out this link: www.jiscmail.ac.uk/cgi-bin/webadmin?A0=fsl
Best,
-Andy