From susumu at mri.jhu.edu Fri Feb 1 17:43:30 2008 From: susumu at mri.jhu.edu (susumu) Date: Fri, 01 Feb 2008 17:43:30 -0500 Subject: [Announcements] Message from Susumu Message-ID: <6i7avk$enj2a@ipex2.johnshopkins.edu> Dear DtiStudio users, AS you may already know, we established a new website, www.mristudio.org last year. In this site, there are two ways to communicate with us and get support. 1) mristudio-users at mristudio.org : This is an email list and the message is delivered to everybody in the list. The activity is also archived as threads, which can be seen anytime. This is a great way to ask questions to the community including us. You can also see the archives ("Community" -> "Mailing List") for past Q & A activities. We are currently modifying the archive so that questions can also be directly posted in the archive instead of sending emails to everybody. Please use this method when you have questions, comments, and suggestions. To join this mailing list please click the above link or http://lists.mristudio.org/mailman/listinfo/mristudio-users. Here you can subscribe to this mailing list. 2) Support at mristudio.org : This email will be read by us and responded individually. Many of you tend to choose this method when there is a question. This is a great way to get support from us, but one issue is the Q & A activities cannot be seen by other users. For the common questions, we move them to "FAQ" sections. If we feel that the questions could be of interest for many users, we anonymize the questions and forward them to "mristudio-users at mristudio.org " so that they can be read by many users and also archived in the website. Sincerely, Susumu Mori -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.mristudio.org/pipermail/announcements/attachments/20080201/7ba7fae2/attachment.html From susumu at mri.jhu.edu Fri Feb 1 17:55:04 2008 From: susumu at mri.jhu.edu (susumu) Date: Fri, 01 Feb 2008 17:55:04 -0500 Subject: [Announcements] Tutorial for DtiStudio/MriStudio Message-ID: <6i7avk$ensci@ipex2.johnshopkins.edu> Dear DtiStudio Users, Recently, we announced two new software programs, RoiEditor and Landmarker. While these are very powerful programs, they are difficult to learn with so many functions. We are trying to make our manual(www.mristudio.org -> "Documentation" -> "User Manual") more complete, but they are still lagging behind. We are currently preparing for tutorial courses in following times and locations. 1) Johns Hopkins Medical Campus in Feb. This is for a testing of the tutorial course, open to only local users (I don't want users to spend $$$ to come, only to find it doesn't work). 2) Kyoto University, Japan, 3/2. 3) ISMRM, Toronto, 5/3 or 5/4. Is there any user who is working in Toronto and willing to provide a large conference room? This is how it works; 1) Users are required to download DtiStudio, RoiEditor, and Landmarker to their own laptops and bring them to the tutorial 2) Users are required to download sample data from our website (currently in "Download" section of the website). 3) Users are required to follow web-based tutorials which will be posted soon. These #1-#3 steps will make sure that you have working versions of our programs and sample data. In this way, all we have to do is to book a room and prepare a lot of electric outlets. The tutorial will be free unless the room costs us a lot. The tutorial will be very unformal without a strict schedule, but it will have 1-hour PPT presentation about DTI and the software, followed by actual step-by-step instructions. I'm thinking about 20 people / session. At this point, we just want to have some idea about how many users are interested in the tutorial. So any comments, suggestions, and information would be welcome. Should you submit your email, please send it to MriStudio_users at mristudio.org or support at mristudio.org. Thanks, Susumu Mori -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.mristudio.org/pipermail/announcements/attachments/20080201/88a952ae/attachment-0001.html From susumu at mri.jhu.edu Thu Feb 14 14:10:16 2008 From: susumu at mri.jhu.edu (susumu) Date: Thu, 14 Feb 2008 14:10:16 -0500 Subject: [Announcements] Second Acconuncement for Tutorial Message-ID: <6i7deu$nkn6c@ipex1.johnshopkins.edu> Dear DtiStudio Users, As I announced few weeks ago, we will have tutorials for new functions of DtiStudio and two new programs (Landmarker and RoiEditor) for image registration and atlas-based automated image analyses. Now the times and locations are set as follows; 2/25: 1-5 pm: Darner conference room in Turner Concourse, Ross Building, Johns Hopkins University, School of Medicine. 3/2: 1-5 pm: Kyoto University, Department of Radiology (http://www.kuhp.kyoto-u.ac.jp/~diag_rad/) 5/4: 1-5 pm: Toronto Western Hospital, 399 Bathurst St. Toronto, ON, Canada To attend the tutorial, it is required to bring own laptops with the current versions of DtiStudio, Landmarker, and RoiEditor, as well as our sample data. Please go to www.mristudi.org -> "WorkShop" and follow the instruction to prepare your laptops. We want to limit the attendees up to 30 for each session. If you are interested in, please express your interest by sending emails to support at mristudio.org. We will send subsequent emails to our "mristudio_users" mailing list to avoid all registered users from getting unnecessary emails. If you are interested in, please register and subscribe to "mristudio_users" at www.mristudio.org -> "Mailing List". Thanks, Susumu Mori From susumu at mri.jhu.edu Thu Mar 27 18:49:57 2008 From: susumu at mri.jhu.edu (susumu) Date: Thu, 27 Mar 2008 18:49:57 -0400 Subject: [Announcements] About gradient table and oblique planes Message-ID: <6i7avk$1c2f2a@ipex2.johnshopkins.edu> Dear DtiStudio/MriStudio users, Thank you for using our software. Recently, I got a question about issue of gradient table and oblique plane. We have been trying to forward this message several times. I want to reiterate because this is very important issue; PLEASE BE CAREFUL WHEN YOU ARE USING OBLIQUE IMAGING. If you are using GE or latest Siemens operating system, you can forget about this message. If you are using Philips and are using "overplus" option "off", you can forget about this message too. If you are using Philips and "overplus = on" or Siemens with older operating system, you are affected. This issue is, YOU HAVE TO RECALCULATE GRADIENT TABLE BASED ON THE OBLIQUE ANGLES because the gradient table is not dynamically rotated with the imaging planes. If you are using Siemens + Dicom/Mosaic, you need to check "Rotate gradients if applicable" box before you load the images. This automatically load oblique info and recalculate the table. If you are using Philips, Jonathan Farrell made a nice software to recalculate and create a new table in the DtiStudio format. You can get it from http://godzilla.kennedykrieger.org/~jfarrell/software_web.htm#DTI_gradient_t able_creator. Either way, it is a good idea to scan the same person twice; one with non-oblique and the other with severe oblique, process the data, and do fiber tracking to see everything is calculated correctly. You never know when scanner companies change something important. Please let me know if you have any questions. Sincerely, Susumu Mori -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.mristudio.org/pipermail/announcements/attachments/20080327/5aeddbc5/attachment.html From support at mristudio.org Thu Apr 24 15:30:46 2008 From: support at mristudio.org (support at mristudio.org) Date: Thu, 24 Apr 2008 15:30:46 -0400 (EDT) Subject: [Announcements] release of Landmarker 1.2 and ROIEditor 1.2 Message-ID: <4146.162.129.28.87.1209065446.squirrel@webmail.mristudio.org> Mristudio users, We have released the new versions of Landmarker and ROIEditor. Please go to our website to download the latest versions, Landmarker 1.2 and ROIEditor 1.2. These two versions include a new atlas called JHU_MNI_SS. This is our single-subject atlas in the ICBM-152 space with extensive segmentation. It is preferable to use this single-subject atlas as a target for highly non-linear warping algorithms, such as LDDMM used in Landmarker. The probabilistic map, ICBM-DTI-81 atlas, remains an ideal target for AIR-based linear or low-order non-linear transformation, but LDDMM requires sharp definition of each structure for accurate transformation. Therefore, the single-subject atlas is preferable. Our DTI atlas in the Talairach coordinate is also updated with more extensive segmentation. Other modifications are as follows; >Software registration for Landmarker 1.2 has changed. The users who don?t have administrator authorization now can run Landmarker on their computers. >Automated intensity histogram matching is added in Landmarker 1.2. >Skull stripping algorithm (beta version) is added to ROIEditor 1.2 >Tensor data can now be loaded to ROIEditor 1.2 >For each ROI drawn in ROIEditor, values of all pixels can be exported as a text file. This feature allows generation of histogram for each ROI. If you have any questions about these two new programs, please contact us at www.mristudio.org or support at mristudio.org.. Thanks! Mristudio Support From susumu at mri.jhu.edu Thu Jul 24 19:00:39 2008 From: susumu at mri.jhu.edu (susumu) Date: Thu, 24 Jul 2008 19:00:39 -0400 Subject: [Announcements] Tutorial on 8/11 Message-ID: <73alti$347ktk@ipex2.johnshopkins.edu> Dear MriStudio Users, We are currently planning to have a tutorial on 8/11 in Hopkins, Baltimore. This schedule is still tentative but we would like to do a rough head-counting. As usual, we require attendants to bring their own laptops and pre-condition them by visiting www.mristudio.org->Workshop. It's free but attendants have to cover all travel expenses. This time, it will become a whole day (10-17:00) course with 3-5 assistants. If you are interested in the tutorial, 1) please express your interest by sending an email to support at mristudio.org 2) please register at Mailing List -> MriStudio-Users because the following announcements will be made only to the mailing list. Thanks, Susumu Mori -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.mristudio.org/pipermail/announcements/attachments/20080724/78dae375/attachment.html From johnlsherman at hotmail.com Fri Jul 25 02:01:38 2008 From: johnlsherman at hotmail.com (John Sherman) Date: Fri, 25 Jul 2008 00:01:38 -0600 Subject: [Announcements] Tutorial on 8/11 In-Reply-To: <73alti$347ktk@ipex2.johnshopkins.edu> References: <73alti$347ktk@ipex2.johnshopkins.edu> Message-ID: Thank you for the email. I am interested in attending t he tutorial on August 11th. I have to arrange some things but it is most likely that I will attend. I am interested in working with data from a 3.0 Tesla G.E. HD MRI system and from a 3.0 T Phillips MRI system. Thanks John Sherman, M.D. Date: Thu, 24 Jul 2008 19:00:39 -0400From: susumu at mri.jhu.eduTo: announcements at mristudio.orgSubject: [Announcements] Tutorial on 8/11 Dear MriStudio Users, We are currently planning to have a tutorial on 8/11 in Hopkins, Baltimore. This schedule is still tentative but we would like to do a rough head-counting. As usual, we require attendants to bring their own laptops and pre-condition them by visiting www.mristudio.org->Workshop. It's free but attendants have to cover all travel expenses. This time, it will become a whole day (10-17:00) course with 3-5 assistants. If you are interested in the tutorial, 1) please express your interest by sending an email to support at mristudio.org 2) please register at Mailing List -> MriStudio-Users because the following announcements will be made only to the mailing list. Thanks, Susumu Mori -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.mristudio.org/pipermail/announcements/attachments/20080725/b650d0d3/attachment-0001.html From support at mristudio.org Fri Aug 1 17:43:44 2008 From: support at mristudio.org (support at mristudio.org) Date: Fri, 1 Aug 2008 17:43:44 -0400 (EDT) Subject: [Announcements] (no subject) Message-ID: <2815.162.129.28.202.1217627024.squirrel@webmail.mristudio.org> Dear Mristudio users, We have released Landmarker 1.3 and ROIEditor 1.3. Please go to our website to download them. The new features are as follows: Landmarker 1.3 1. ROI-based manual intensity matching Before LDDMM, intensity profiles need to be adjusted between the template and the subject if non-quantitative images (such as b0 and T2-weighted images) are used. Currently there are manual and automated histogram matching functions are available in Landmarker. There is also an option for remote automated histogram matching when the images are sent to our server. However, because you can not check the quality of the matching, we recommend to use the matching functions in Landmarker. The new manual intensity matching is based on intensities inside the ROI. In this approach, users are asked to define corresponding regions between the template (atlas) and the subject data. High intensity (e.g. CSF), medium intensity, and low intensity regions should be specified to capture the entire intensity range both in the atlas and the subject. Then the intensity profiles are matched between the two. This type of matching is necessary because image intensity of non-quantitative images is arbitrary and there could be a large difference between the template and the subject, which confuses LDDMM. This is, in theory, not necessary when quantitative maps are used such as FA and Trace. 2. All the loaded images can be resampled at one time. When image resampling functions are used (buttons in the "image" section), they were applied only to the active images. Now, there is an option to apply them to all loaded images. 3. These is a manual scaling when the data format of an image is changed to byte. Before LDDMM, all image formats have to be converted to the byte format. This can be done by the "BFW" button in the "image" section. In the older version, this conversion was automatically done by taking the maximum intensity = 255. With the new version, you can manually specify the conversion to the 0-255 range. When quantitative maps such as FA and Trace are converted to byte for LDDMM, absolute scale should be used to achieve the optimum intensity matching. 4. Invert an AIR linear matrix when a xx.air file is provided. Using Landmarker, we can do "inverse" transformation, in which our atlases are transformed to each subject data. LDDMM provides K-map (forward) and H-map (reverse), but AIR provides only forward (Image.air). This function can invert the AIR transformation. 5. Jacobian calculation of a LDDMM map. A new function to calculate Jacobian from LDDMM results (K-map) was added. This extract volume change information from a K-map in a pixel-by-pixel manner. 6. Disable tensor reorientation. When transformation is applied to a tensor field, Landmarker performs tensor-reorientation whenever rotation components are involved. However, in certain circumstances, users may want to turn off this function. For example, when we do B0-susceptibility distortion correction, we believe tensor should not be re-oriented because the distortion is just pixel-mislocation, which is different from cross-subject transformation. ROIEditor 1.3 1. The data format of a ROI data file created by ROIEditor is changed to WORD. In the old version, it was byte, which could hold only up to 256 ROIs. Now you can delineate more than 36,000 ROIs! 2. Hide all ROIs and show all ROIs. In the old version, the show/hide of ROIs had to be controlled for each ROI individually. Using this new function, all ROIs can be shown/hid at once. 3. Calculate the RGB of color images. RGB values of color maps can now be quantified for each ROI. 4. Double click a ROI and the corresponding ROI object box in the ROI object list will be highlighted. It was sometimes difficult which ROI in the image correspond which ROI in the ROI list. Now, by clicking an ROI in the image, you can identify the ROI object in the ROI list. 5. Users can decide if zero is treated as no data in statistical calculation. When there are pixels with "0" intensity in ROI, sometimes you don't want to include the pixels for statistics (e.g. masked out pixels). You can choose whether you want to include such pixels or not. If you have any questions about these two new softwares, please send your email to Mristudio-users or support at mristudio.org. Thanks! Mristudio Support From support at mristudio.org Fri Aug 1 17:45:13 2008 From: support at mristudio.org (support at mristudio.org) Date: Fri, 1 Aug 2008 17:45:13 -0400 (EDT) Subject: [Announcements] release of Landmarker 1.3 and ROIEditor 1.3 Message-ID: <2821.162.129.28.202.1217627113.squirrel@webmail.mristudio.org> Dear Mristudio users, We have released Landmarker 1.3 and ROIEditor 1.3. Please go to our website to download them. The new features are as follows: Landmarker 1.3 1. ROI-based manual intensity matching Before LDDMM, intensity profiles need to be adjusted between the template and the subject if non-quantitative images (such as b0 and T2-weighted images) are used. Currently there are manual and automated histogram matching functions are available in Landmarker. There is also an option for remote automated histogram matching when the images are sent to our server. However, because you can not check the quality of the matching, we recommend to use the matching functions in Landmarker. The new manual intensity matching is based on intensities inside the ROI. In this approach, users are asked to define corresponding regions between the template (atlas) and the subject data. High intensity (e.g. CSF), medium intensity, and low intensity regions should be specified to capture the entire intensity range both in the atlas and the subject. Then the intensity profiles are matched between the two. This type of matching is necessary because image intensity of non-quantitative images is arbitrary and there could be a large difference between the template and the subject, which confuses LDDMM. This is, in theory, not necessary when quantitative maps are used such as FA and Trace. 2. All the loaded images can be resampled at one time. When image resampling functions are used (buttons in the "image" section), they were applied only to the active images. Now, there is an option to apply them to all loaded images. 3. These is a manual scaling when the data format of an image is changed to byte. Before LDDMM, all image formats have to be converted to the byte format. This can be done by the "BFW" button in the "image" section. In the older version, this conversion was automatically done by taking the maximum intensity = 255. With the new version, you can manually specify the conversion to the 0-255 range. When quantitative maps such as FA and Trace are converted to byte for LDDMM, absolute scale should be used to achieve the optimum intensity matching. 4. Invert an AIR linear matrix when a xx.air file is provided. Using Landmarker, we can do "inverse" transformation, in which our atlases are transformed to each subject data. LDDMM provides K-map (forward) and H-map (reverse), but AIR provides only forward (Image.air). This function can invert the AIR transformation. 5. Jacobian calculation of a LDDMM map. A new function to calculate Jacobian from LDDMM results (K-map) was added. This extract volume change information from a K-map in a pixel-by-pixel manner. 6. Disable tensor reorientation. When transformation is applied to a tensor field, Landmarker performs tensor-reorientation whenever rotation components are involved. However, in certain circumstances, users may want to turn off this function. For example, when we do B0-susceptibility distortion correction, we believe tensor should not be re-oriented because the distortion is just pixel-mislocation, which is different from cross-subject transformation. ROIEditor 1.3 1. The data format of a ROI data file created by ROIEditor is changed to WORD. In the old version, it was byte, which could hold only up to 256 ROIs. Now you can delineate more than 36,000 ROIs! 2. Hide all ROIs and show all ROIs. In the old version, the show/hide of ROIs had to be controlled for each ROI individually. Using this new function, all ROIs can be shown/hid at once. 3. Calculate the RGB of color images. RGB values of color maps can now be quantified for each ROI. 4. Double click a ROI and the corresponding ROI object box in the ROI object list will be highlighted. It was sometimes difficult which ROI in the image correspond which ROI in the ROI list. Now, by clicking an ROI in the image, you can identify the ROI object in the ROI list. 5. Users can decide if zero is treated as no data in statistical calculation. When there are pixels with "0" intensity in ROI, sometimes you don't want to include the pixels for statistics (e.g. masked out pixels). You can choose whether you want to include such pixels or not. If you have any questions about these two new softwares, please send your email to Mristudio-users or support at mristudio.org. Thanks! Mristudio Support From support at mristudio.org Mon Oct 13 14:34:30 2008 From: support at mristudio.org (support at mristudio.org) Date: Mon, 13 Oct 2008 14:34:30 -0400 (EDT) Subject: [Announcements] Important Bug Report for Landmarker Message-ID: <2080.162.129.28.202.1223922870.squirrel@webmail.mristudio.org> Dear MriStudio users, Recently, it came to our attention that there is a bug in the registration algorithm of Landmarker. It is related to the tensor re-orientation and affects the tensor/vector data after tensor transformation. As a result, the vector data can not be used for fiber tracking. The bug does not affect scalar values such as FA and eigen-values generated from the transformed tensor data. We have fixed the bug and uploaded the updated version on the server for you to download. Below, we copied the description in the "ChangeLog" in the site. We are sorry for this inconvenience and hope this bug will not affect your research significantly. Should you have any questions, please contact us. Sincerely, Susumu Mori ChangeLog Description: IMPORTANT: A bug was found in the tensor / vector reorientation algorithm of all the previous Landmarker versions, Version 1.0.1, Version 1.1.1, Version 1.1, Version 1.2 and Version 1.3. 1) A bug affects the linear and non-linear registration of tensor data (*.d). 2) When the transformation involves shearing operations, the tensor orientation was not properly re-oriented. 3) This affects tractgraphy after registration. 4) This does not affect scalar values such as FA, ADC, and eigenvalues. 5) We strongly encourage you to download Version 1.4 and reprocess your registered images. For the reprocessing, you simply need to reapply the stored transformation matrices using the new version (Version 1.4) because transformation matrices are not affected by the bug.