MEDx 3.4.3 Release Notes

MEDx 3.4.3 is the third version upgrade to MEDx 3.4. The new functionality in this release is described below.

New Operating Systems Supported

• MEDx 3.4.3 is now officially supported on the Solaris 9 and RedHat Linux 9 operating systems. The list below shows all supported operating systems for MEDx 3.4.3:

  RedHat 7, 8, 9
  Solaris 7, 8, 9
  Irix 6.5+

DICOM Reader Accepts DICOM Mosaic Image Format

• The DICOM reader in MEDx v3.4.3 can now read MR images output from the Siemens scanner that are in Mosaic format. The format of these images, also known as Siemens Syngo, support multiple images in each file.
  
  Documentation: For more information on this feature, please see Chapter 8 of the MEDx User's Guide.
  
  Location of feature in the MEDx GUI: The DICOM reader module can be found under the Image->Open Image->File Format->DICOM menu.

Analysis of Brain Lesions (ABLe) Enhancements

• MNI Template brain used for registration. The ABLe module has been upgraded to include the use of the MNI 152 subject brain as a template for MR registration. When the option to “Analyze the whole brain with VOTL database” is chosen, the input brain is registered to the canonical MNI template, which is an average of 152 spatially normalized brains. This assumes the input volume is an MRI. If CT, the input brain is registered to a CT template, which has previously been aligned to the MNI template. A transformation from MNI space to Talairach space (Talairach/Tournoux 1988) is then applied. This transformation was previously created by registering (with a 12 parameter affine model) a gray and white matter image mask of the MNI brain to a gray and white matter image mask derived from the VOTL database.

• Single and Two Group Analysis. The Single Group Analysis option is only available in VOTL Database mode. The user may determine a minimum number of overlapping lesions (one from each subject) required to be included in a group report. The group report will show all anatomic structures where at least this many subjects had a lesion. Once the report is complete, a colorwash image is created, representing the fusion of the common lesion areas on top of a MRI template image in Talairach space. The Two Group Analysis option is only available in VOTL Database mode and is designed for a dissociation analysis. With this type of analysis, two separate groups of subjects must be determined a priori, usually based on some common deficit. The deficit is different for the two groups. The purpose of this analysis is to generate three reports. The first report represents the overlap of lesions common to all subjects in Group A and none in Group B and lists these common structures based on a query to the VOTL database. The second report represents the overlap of lesions common to all subjects in Group B and none in Group A and lists these common structures based on a query to the VOTL database. The third report represents the overlap of lesions common to all subjects in Group A and Group B and lists these common structures based on a query to the VOTL database. Once the report is complete, a group of colorwash images is created, representing the fusion of the common lesion areas on top of a MRI template image in Talairach space.

• Behavior Analysis. With this type of analysis, an ascii file containing a matrix of behavior scores (in columns) and subject names (in rows) is required input. The maximum number of patients is 32. The format of the file is shown below. Tabs can separate the entries.

  PATIENTS	Behavior1	Behavior1
  Pat1	.75	.90
  Pat2	.85	.65
  Pat3	.77	.88

  The program will sort through the lesion images and will determine pixel by pixel which images have a lesion at that location and which do not. A t-test is then performed comparing the Behavioral measures between the lesion and non-lesion groups for each pixel. The result is a t-map image where the intensity of each pixel is equal to the students’ t-value for that pixel’s behavior t-test. A t-map image is generated for each behavior and these images are then placed into a MEDx group page. An option exists to determine which subjects had lesions and which did not for any chosen voxel in the t-map image.

• Load Brodmann Areas from VOTL Database. This button is used to load a set of regions of interest (ROIs) representing all Brodmann areas by the VOTL database. The only requirement is that the image be in Talairach space based on the ABLe registration option. Therefore, the Brodmann Areas can be overlaid on the input brain that has been registered to Talairach space or on any of the t-map images generated from the Behavioral analysis. An option also exists to load a subset of Brodmann areas.

Documentation: For more information on these new features, please see Section 17.8 in the MEDx User's Guide.

Location of feature in the MEDx GUI: The Analysis of Brain Lesions module can be found under the Volumetric menu in the toolbox.

New Mx Commands for 3D Contour Properties

• With the new MxGetContourProperties and MxSetContourProperties commands, the user can set and get the name of a 3D contour within a script. This functionality was unavailable in previous versions of MEDx.

  Documentation: For more information on this feature, please see the MEDx ImageScript page.

Utility Scripts

• ReadVarian.tcl: This script will allow the user to read in MR data output from the Varion scanner.

• ReadSDT.tcl: This script will allow the user to read in MR data output from the Varian scanner in the SDT format.

• WriteAIR.tcl: This script will allow the user to extract motion parameters (translations and rotations) after running AIR 3.08 within MEDx.

• AccessDICOMManager.tcl: This script will allow you to access the DICOM Manager's contents.

• DICOM_Save_ByCloningKeys_MR.tcl: Automatically preserve/transfer
  header information from your existing DICOM MR images into DICOM MR
  images you create with MEDx.
  

  Location of feature in the MEDx Distribution: These utility scripts can be found in the $PXHOME/tcl directory where $PXHOME is the MEDx installation directory.
  

MEDx 3.4.2 Release Notes

MEDx 3.4.2 is the second version upgrade to MEDx 3.4. The new functionality in this release is described below.

Additions to MEDx DICOM Capabilities - DICOM Reader

• DICOM Nuclear Medicine images can be opened into MEDx starting with version 3.4.2. Each of the eight image categories can be opened. These categories are Static, Whole Body, Dynamic, Gated, Tomo, Gated Tomo, Recon Tomo, and Recon Gated Tomo.

• A new, filename-based user option has been added in version 3.4.2 to enable the user to open occasional MR and CT data for which the options available in 3.4.1 are not ideal or at times not suitable. A primary example of such data is temporally consecutive volumes, such as fMRI volumes, where temporal relationship information has not been well-defined in the image headers in the form of separate Acquisition Numbers or separate unique Series identifiers or separate Temporal Position Identifiers. Due to the absence of this information, the new option enables the user to utilize file names to supervise the number of volumes to be opened and the slice order.

• A new option for CT and MR images has been added to the reader that enables the user to select whether a volume containing one slice should be opened as a 2D Image or as a Volume.

• A new option has been added to the reader that enables the user to specify a list containing DICOM file names and/or directory names (when using MxOpenImageAsDicom, or by multi-selecting files in Open Image) along with the choice to create or not create a Group to represent the distinct items in this list. The user did not have the choice to not create this Group in version 3.4.1 because it was always automatically created. The user also did not have the choice in 3.4.1 to specify a combination of file names and directories.

• New optional arguments have been added to MxOpenImageAsDicom to add support to ImageScript for the user options shown in the Open Image panel of DICOM Defaults that are used for customizing the behavior of the DICOM reader.

• A new ImageScript command called MxDICOMGetDefaults has been added to make the values of some DICOM Defaults available in ImageScript.

• In version 3.4.1, when opening more than one file at once (such as when opening a directory containing multiple files), the reader did not open an image if it appeared multiple times under different file names, but did not notify the user that a duplicate image scenario was being encountered. In 3.4.2, when this scenario is encountered, the user is notified, along with the suggestion that the DICOM Manager can be used to address it. Please see the notes below on the new interactive tool in the DICOM Manager for more information.

• In version 3.4.2, we have improved our documentation of the DICOM reader whose features have grown considerably since 3.4. Please find this documentation in the form of new text, charts, and tables in Chapter 8 of the MEDx 3.4.2 User's Guide.

• The following defects found in version 3.4.1 have been repaired in 3.4.2.

  1. Various problems related to saving and opening folders containing DICOM data.

  2. When opening DICOM data, header keys whose DICOM Value Representation is SQ (Sequence), DT (Date Time), or FL (Floating Point Single) were not being properly placed into the MEDx image header information.

Additions to MEDx DICOM Capabilities - DICOM Manager

• In order to bring new images into the DICOM Manager, it is no longer required to use the Add button. Starting with version 3.4.2, the use of this button is merely recommended. Images can now be directly placed (moved or copied) into any location under the DICOM Manager directory.

• In version 3.4.1, if an image were present multiple times under different file names under the DICOM Manager directory, the DICOM Manager occasionally became corrupted and nonfunctional. This has been repaired, and the duplicate image scenario addressed, in 3.4.2 by adding an interactive dialog box that appears automatically when needed that enables the user to select which image copy among the multiple copies should be the one to be used by the DICOM Manager and to subsequently move the non-selected copies out of the DICOM Manager to a user-specified directory. This interactive tool automatically appears also when using the Add button, if it is determined the Add would add an image more than once.

• In version 3.4.1, although a single Study could be sent/"pushed" to another DICOM application, it was not possible to send/"push" a single Series or Image. This feature has been added in 3.4.2.

• Two new buttons have been added to enable the user to copy images from the DICOM Manager to a user-specified directory or to move images out of the DICOM Manager to a user-specified directory.

• Occasionally, a Series in DICOM can contain more than one kind of DICOM object, for example both MR and CT. A new (fourth) browser has been added to the DICOM Manager to enable the user to distinguish between the different kinds while browsing the contents of the Series. This browser becomes visible only when a Series does contain more than one type of object.

• A new column for Image Type has been added to the Image browser.

• The following defects found in version 3.4.1 have been repaired in 3.4.2.

  1. In the DICOM Manager, if multiple Studies were consecutively deleted without pressing the Update DICOM Manager button until after all Studies were deleted, a list of Studies incorrectly continued to be shown in the Study browser after that update, improperly suggesting some or all of the deletions did not take place.

  2. In the DICOM Manager, there were occasions where opening an entry in the Image browser, and not in the Study and Series browsers, was generating an error.

Additions to MEDx DICOM Capabilities - DICOM Query/Retrieve

• In versions through 3.4.1, MEDx included a menu called Type of DICOM System (located in DICOM Defaults) that served to configure in one step the mode and look of the DICOM Query/Retrieve window to be compatible with the query/retrieve behavior of the specific DICOM conformant product selected in the menu. In version 3.4.2, the functionality of this menu has been enhanced. When using this one-step tool in 3.4.2, you can now select from approximately 120 DICOM conformant products, including scanners, PACS systems, workstations, toolkits, and miscellaneous other systems of varying scale.

• Querying by Patient Name has been improved to increase the likelihood of a Patient Name match, by automatically issuing consecutive queries, if needed, that vary how exactly the user-specified last, first, and/or middle name components are spliced, with and without the use of wild card and/or delimiter characters.
  
  Documentation: For more information on these features, please see the MEDx User's Guide, Chapter 8.
  
  Location of features in the MEDx GUI: The DICOM utilities can be found under the Image->DICOM menu in the MEDx folder. The DICOM defaults settings can be found under the DICOM Defaults option under the Setup menu.

Integration of the Talairach database (VOTL) into the Analysis of Brain Lesions (ABLe) module

• The Analysis of Brain Lesions module has been upgraded to include the use of the Volume Occupancy Talairach Label (VOTL) database from Jack Lancaster's group
  at the University of Texas Health Science Center in San Antonio. Now lesion location can be quantified using the atlas derived from Damasio and previously contained in Able (including only 11 slices from the brain) or the VOTL database which includes all slices of the brain in Talairach space. The reporting facilities for single subject and group analysis of brain lesion images have been improved and include percentage of brain regions taken up by the lesion as well as percentage of the lesion in each brain region.
  
  Documentation: For more information on this feature, please see the MEDx User's Guide, Chapter 17 Section 17.8
  
  Location of feature in the MEDx GUI: The Analysis of Brain Lesions module can be found under the Volumetric menu in the toolbox.

Additions to the Talairach Daemon Interface

• In previous versions of MEDx, the Talairach Daemon (RIC/UTHSCSA) was used to obtain anatomic structures based on input Talairach coordinates. Access to the Talairach Daemon required an internet connection. In version 3.42, MEDx no longer accesses the Talairach Daemon over the Internet, but instead accesses a comparable database known as the VOTL (volume occupancy Talairach labels) which is installed locally with the MEDx distribution. This VOTL database was provided by Prof. Jack Lancaster from the RIC at the University of Texas Health Science Center in San Antonio. Queries to this database can be structured to include a search range, or nearest gray matter structure. Input coordinates can be entered via a text file, or by the location of markers on an image. The input coordinates may be either in classic Talairach space or MNI-Talairach space.
  
  Documentation: For more information on this feature, please see the MEDx User's Guide, Chapter 31 Section 31.14
  
  Location of feature in the MEDx GUI: The Talairach Database Interface module can be found under the Functional or Volumetric menu in the toolbox.

Option for conducting perfusion analysis without arterial input function fitting

• Absolute Cerebral Blood Volume (CBV) calculations require the normalization of the relative CBV maps by an estimate of the arterial input volume. While previous calculations automatically fit the arterial input function (AIF) to a gamma variate, the new feature provides the option of conducting the analysis without a fit. In the event of irregular arterial input functions which may complicate the fitting process, the "No Fit" option can enhance the accuracy of the calculations.
  
  Documentation: For more information on these features, please see the MEDx User's Guide, Chapter 33.
  
  Location of features in the MEDx GUI: The Perfusion Analysis utilities can be found under the Functional->Perfusion Analysis menu in the toolbox.
  

Modification of the DES Reader and Integration of DES Writer into MEDx

• The MEDx 3.4.2 DES Reader has been upgraded to fully support the opening of images of the “descriptor file” format. This format is supported by programs such as Alice and Cheshire. The DES Reader now allows users to read groups of DES volumes in addition to a single volume.
  
  A new DES Writer, an important and powerful tool, has been integrated into MEDx 3.4.2. Images of various formats may first be loaded into MEDx and then saved in DES format. This module is capable of writing out an image, an image volume, or a group of image volumes in DES format.
  
  Documentation: For more information on this feature, please see the MEDx User's Guide, Chapter 5, Section 5.5.24 for the DES Reader and Section 5.7 for the DES Writer
  
  Location of feature in the MEDx GUI: The functionality of the DES Writer can be found under Image ->SaveImageAs ->DES or Image -> SaveGroupAs -> DES

Integration of the Philips Reader into the Open Image menu

• With the new PHILIPS reader, the user does not even have to specify the image type explicitly. A simple click on Apply in the Auto Detect Format, is sufficient for MEDx to recognize the datatype of the specified filename whereby the file is displayed automatically.

"Seeding” of Dialog boxes with ImageScript functions

• ImageScript is the scripting (macro) language in MEDx. A subset of the ImageScript function calls pertains to activating dialog boxes in MEDx. To pop-up a dialog box, one calls the command MxDialogxxx, where xxx refers to the dialog box name (e.g. MxDialogOpenImage). As of MEDx version 3.42, an additional argument to the ImageScript command MxDialogxxx provides the user with the capability to pre-set values in the dialog box and to reconfigure buttons. The user has to simply specify the configuration file (tcl code), which is automatically sourced, implementing the user specifications. This feature provides the user with autonomy to adapt the functionality of the existing dialog boxes for a particular application. It is required that the user has some knowledge of the Tcl code for the dialog box under question.
  
  Documentation: For more information on this feature, please see the MEDx Help menu found in the File->ImageScript Execution dialog box.

Logging of the MxPause ImageScript command

• The logger enables the user to keep track of a particular MEDx session whereby simply sourcing the logged file becomes sufficient for repeating the exact same analysis at a later time. The new Log Pause feature allows the user to insert comments into the logged file and pause the running script. Possible prompts such as "Please specify graphics" for example could help clarify the analysis.
  
  Documentation: For more information on this feature, please see the MEDx Help menu found in the File->ImageScript Execution dialog box.

FSL Scripting commands

• Scripting commands, alternately referred to as Mx commands, are a distinguishing feature of MEDx as they enable access to all MEDx features through function calls. The scripting commands empower the user, enabling him to add features of his own to MEDx. The current release of MEDx provides Mx calls for FSL functions.
  
  Documentation: For more information on this feature, please see the MEDx User's Guide, Chapter 34.
  
  Location of feature in the MEDx GUI: The FSL modules can be found under the FSL menu in the toolbox.

Option for unwarping of echo-planar images (EPI) through FUGUE

• Distortions in echo-planar images are successfully removed through FUGUE – FMRIB's utility for geometric unwarping of EPI images. A map of the residual magnetic field is calculated through two images acquired with an echo offset. The phases of these images are initially unwrapped prior to the calculation of the fieldmap which in turn is used to unwarp the EPI images. The MEDx implementation of FUGUE while true to its original FSL roots has an additional user-interface to facilitate the various acquisition specifications.

Multichannel automatic segmentation option with nonuniformity correction through FAST – FMRIB's Automated Segmentation Tool.

• The FAST technique which comes as part of the FSL package, accurately differentiates between gray matter (GM), white matter (WM), cerebrospinal fluid (CSF), and lesions without any user interaction. The method provides a distinct advantage over other segmentation techniques due to the intrinsic correction scheme it provides for intensity variations resulting from RF inhomogeneities. The process is fully automated and the outputs include a bias-field corrected input image as well as probabilistic and/or partial volume segmentation maps.

Independent Component Analysis (ICA) through the Melodic option of the FSL package

• The ICA based MELODIC option of FSL provides the MEDx user with the opportunity of conducting model-free analysis. In non-exploratory approaches, the response to activation is expected to conform to a user-defined waveform. Yet in compromised regions of the brain, the response may not adhere to this expectation. Exploratory techniques like ICA are not model-based. Thus unexpected temporal responses to activation can be detected through ICA as the technique extracts distinct temporal components and the associated spatial maps taking into account the temporal response from all pixels.
  

MEDx 3.4.1 Release Notes

MEDx 3.4.1 is the first version upgrade to MEDx 3.4. The new functionality in this release is described below.

Major Feature List

• Interpretation and Presentation of DICOM Images in MEDx. Starting with version 3.4.1, DICOM images read into MEDx are interpreted based on the content of the images describing the relationships between the images in the DICOM Information Model. This improved mechanism for interpreting DICOM images is one of the new major features in 3.4.1 and replaces the file name and directory name driven mechanism previously used for interpreting DICOM images in versions of MEDx through 3.4. Simply place all of your DICOM files immediately in and/or somewhere under one directory and then open that directory in MEDx.
  
  MEDx now contains support for opening DICOM images that are not DICOM Part 10 images.
  
  In addition to the new DICOM object interpretation mechanism, MEDx now contains several object-specific major and minor improvements and repairs. Selected improvements and repairs are as follows:

  • MEDx can now read DICOM PET images. PET images of the same series can either be opened as a group of 3D volumes acquired at different temporal durations, or as a group of group of 2D images such that the 2D images in a group of 2D images are at the same spatial location and have been acquired at different temporal durations. Temporal information is automatically extracted when reading the PET images so that it may be used to generate time activity curves by using the Time Stamp Entry and Time Activity Curve dialog boxes.

  • MR images of the same series can now be opened also as a group of volumes where each volume corresponds to a different time point as determined by the value of the Temporal Position Identifier attribute. This feature is useful for dynamic and functional MR data that contains this attribute.

  • Some image orientation and display problems that appeared for some CT and MR images have been repaired in 3.4.1.

  • The ImageScript function MxOpenImageAsDicom has been repaired and improved to reflect this latest DICOM reader in MEDx 3.4.1.

  • The values of the DICOM attributes in the images opened can now be accessed more intuitively in MEDx 3.4.1. For each attribute, the name of the key in Image Header Info corresponding to it now contains the DICOM text description of the attribute followed by the DICOM hexadecimal tag designation of the attribute. For example, PatientID_0010_0020. The values of the attributes in the images opened are accessible also by using the ImageScript function MxGetHeaderInfo.

  • Secondary Capture images could previously be opened in MEDx only if they were RGB images that were color-by-pixel (RGBRGB…). MEDx can now also open color-by-plane (RRR…GGG…BBB) RGB, as well as monochrome Secondary Capture images.

  • All meaningful values for the (Bits Allocated, Bits Stored, High Bit) triplet are now supported in MEDx. The meaningful values for this triplet are as follows. Bits Stored must be less than or equal to Bits Allocated. High Bit must be less than or equal to Bits Allocated minus 1 and greater than or equal to Bits Stored minus 1.

• Importing and Exporting DICOM images in MEDx. DICOM transfer syntaxes are used when images are transferred over the network via DICOM. The default DICOM transfer syntax is called Implicit VR Little Endian. MEDx now supports two other DICOM transfer syntaxes: Explicit VR Little Endian and Explicit VR Big Endian. The support for these "explicit VR" transfer syntaxes in which the data type of DICOM attributes is explicitly encoded into an image enables MEDx to preserve vendor-private DICOM attributes through image transfers, as well as to interpret these attributes correctly when the images are opened.

• Managing DICOM images in MEDx. A new tool called the DICOM Manager is available in MEDx. The DICOM Manager enables users to browse, open, delete, send, and manage all of their DICOM data in one place.

Minor Feature List

• Latest SPM'99 Support. We have revised the SPM'99 exectuables to include the latest changes and fixes from the SPM group.

• Perfusion Analysis. The perfusion analysis module now provides an option for manual selection of arterial input pixels.

MEDx 3.4 Release Notes

This document describes new functionality added to MEDx and included in version 3.4. A CD-ROM including the version 3.4 software binaries will be mailed to current MEDx subscription holders. PDF files of the MEDx user's guide will be included on the CD-ROM. For those with Internet access, the software and User's Guide are available via anonymous ftp. Contact medxsales@sensor.com for further downloading and license information.

These release notes are split into two main sections: major features and minor features. A brief description is provided for each feature as well as a reference to the MEDx User's Guide for a more full explanation.

Major Feature List

• SPM'99. Over the last year, the incorporation of SPM-99 into MEDx has been the most-requested new feature. We have built an executable version of SPM-99 that will be distributed with MEDx 3.4. This version of SPM-99 will be started from a menu within MEDx, and will have the same look and feel as the interpreted Matlab version. It will not require a Matlab license to run; it will, however, require a valid MEDx 3.4 or higher license.
  
  Folder->Toolbox->SPM->SPM99
  See the MEDx User's Guide, Chapter 32, Section 32.1

• FSL. This new module is a collection of functional and structural brain image analysis tools developed by the Image Analysis Group at Oxford University under the leadership of Dr. Stephen Smith. The structural tools consist of an automatic brain extraction tool (BET), a Linear Registration Tool (FLIRT), and a noise reduction technique based on Smallest Univalue Segment Assimilating Nucleus (SUSAN). The Easy Analysis Tool (FEAT) provides a very straightforward and easy-to-use interface for the fMRI data analysis process. This interface enables the user to specify pre-processing steps such as motion correction, spatial filtering, global intensity normalization, and temporal filtering. The user can also specify a paradigm and the statistical method for individual or group analyses. The functional analyses employ a General Linear Model and a model-free approach for cases where no a priori knowledge of the expected response exists. FSL also includes tools for analyzing event-related fMRI experiments. The FSL module provides a powerful, fast, and robust alternative set of tools that complement and enhance the existing MEDx tool set.
  
  Folder->Toolbox->FSL
  See the MEDx User's Guide, Chapter 34

• The Analysis of Brain Lesions module is a tool to characterize lesions in MRI or CT images of the adult human brain. Module functionality includes:

• Determination of the lesion volume

• Spatial normalization of the input brain into Talaraich space

• Percent of total brain volume taken up by the lesion

• Quantification of Brodmann Area intersections based on Damasio Atlas

• Reporting routines allowing for group lesion display

Folder->Toolbox->Volumetric->Analysis of Brain Lesions
See the MEDx User's Guide, Chapter 17 Section 17.8

Minor Feature List

• MR Parameter Estimation. This module will estimate relaxation and diffusion parameters from various MR acquisition schemes. Summary Estimates of MR parameters T1-relaxation, T2-relaxation and Apparent Diffusion Coefficient (ADC) can be useful for differentiating physical tissue characteristics in MRI data. Different tissue types have characteristic T1 and T2 values. Values outside the normal range for a given tissue type may indicate pathology. True T1 and T2 maps also yield better tissue segmentation than T1 or T2-weighted images. Calculated ADC maps from diffusion weighted images estimate the diffusion coefficient of water at each voxel. These maps can be useful for identifying tissue at risk after stroke. Chi-square (x2) maps measure goodness of fit of the model to (inversion, saturation or multi-echo) data.
  
  Folder->Toolbox->Functional->MR Parameter Estimation
  See the MEDx User's Guide, Chapter 17

• New Option in Talairach Daemon Interface. A new option has been provided in the Talairach Daemon Interface for those running MEDx on Sun Solaris or SGI IRIX platforms. This new option, "Load by Markers" allows the user to select voxels on the volume by clicking with the left mouse button. The Talairach coordinates of these marker locations are automatically loaded into the Talairach Daemon Interface.
  
  Folder->Toolbox->Functional->Talairach Daemon Interface Folder->Toolbox->Volumetric->Talairach Daemon Interface
  See the MEDx User's Guide, Chapter 31, Section 31.14

• Flip Volume. A new option has been added to the Image and Transformation menus (in 3D mode) allowing the user to manipulate the orientation of the volume. This module provides for flipping the orientation of the volume from left to right, or top to bottom. In addition, the user may reverse the slice order of the volume.
  
  Folder->Toolbox->Transformation->Flip Volume
  Folder->Image-> Flip Volume
  See the MEDx User's Guide, Chapter 19, Section 19.3

• Single-Group t-Test. An option is now available to perform a single group t-test in order to perform a group analysis of fMRI data. The user provides a group of contrast images and a mean with which to compare.
  
  Folder->Toolbox->Functional->Group Statistics->Within Group

MEDx 3.3 Release Notes

Major Feature List

• MRI Perfusion. Estimates of Cerebral Blood Flow (CBF), Cerebral Blood Volume (CBV) and Mean Transit Time (MTT) are computed from several different algorithms based on MR perfusion imaging of the brain with a Gadolinium contrast agent. Many of these methods use gamma-variate fitting of the ‘first pass response’ at every pixel and numerical deconvolution of the input tissue curve to find the residue. This module also includes methods to automatically segment arterial voxels. The perfusion module is the first add-on module available for MEDx. It does not come with the MEDx 3.3 upgrade but can be purchased separately (medxsales@sensor.com).

  Folder->Toolbox->Functional->Perfusion

  See the MEDx User's Guide, Chapter 33

• Estimation of Spatial Smoothness. Methods to estimate spatial smoothness of a statistical image have been added. The output of this option is used for proper z-map threshold calculations in the Critical Threshold option and for Cluster Detection of z-maps. This option as well as the Critical Threshold and Cluster Detection have been included in one dialog box named Final Significance.

See the MEDx User's Guide, Chapter 31, Section 31.13

• Interactive Brain Lesion Analysis. A module has been added via the ImageScript scripting language to characterize lesions in MRI or CT images of the adult human brain. This module performs the following tasks: determination of the lesion volume as well as percent of total brain volume, spatial normalization of the input brain into Talairach space, quantification of Brodmann Area intersections and a series of reporting routines allowing for group lesion display. This project was a collaborative effort between Medical Numerics, Inc. and the CNS lab of the National Institutes of Neurological Disorders and Stroke at the National Institutes of Health, as well as the Diagnostic Radiology Department at the National Institutes of Health.

  File->ImageScript Execution->iblap.tcl

• MR Parameter Estimation. A module has been added via the ImageScript scripting language to generate T2-relaxation, T1-relaxation and Apparent Diffusion Coefficient (ADC) image maps, given image data derived from certain MR pulse sequences (multi-echo, inversion recovery, saturation recovery, diffusion weighted). Output maps include chi-square for goodness of fit measurements.

  File->ImageScript Execution->mr_params.tcl

  MR Parameter Estimation page

• Talairach Daemon Interface. The Talairach Daemon, developed at the Research Imaging Center at the University of Texas Health Science Center at San Antonio, is a high-speed database server for querying and retrieving data about human brain structure over the Internet.  A MEDx interface has been developed to query this Daemon. Talairach coordinates can be entered into the user interface manually or by loading an ascii file (e.g. output from MEDx local min/max reporter).

Folder->Toolbox->Functional->Talairach Daemon Interface

See the MEDx User's Guide, Chapter 31, Section 31.14

• Save/Send as DICOM. DICOM part 10 format has been added to the list of export image formats. The following modalities are supported: MR, CT, CR. The user may export a single slice, a volume, or group of volumes as a study. This study may be sent via DICOM protocol to any specified DICOM destination.

  Folder->Image->Save Image As->DicomFolder->Image->DICOM->Send

  See the MEDx User's Guide, Chapter 5, Section 5.7.2, and Chapter 8, Section 8.2.4.

Minor Feature List

• Polynomial de-trending. Polynomial de-trending In previous versions of MEDx, for each voxel, only a linear trend in fMRI data could be detected and removed from the time series. In this version, the user may specify the order of the de-trending algorithm. The order must be less than the number of scans in the time series.

• Improved Paradigm Editor. The Paradigm Editor is used to describe how the scans for an analysis of functional imaging data correspond to the task conditions of your experiment (e.g. The first 10 scans of your experiment correspond to an activation state). Two major changes to the Paradigm Editor have been included in this version. (1) The Apply Paradigm dialog box has been replaced by a button in the Paradigm Editor dialog box. (2) A contrast field has been added. This contrast field is used to specify weights for each epoch. The saved paradigm file, containing these contrasts (weights) can serve as input for the Correlation Analysis dialog box. If 0 contrasts are specified, this dialog box behaves just as it did in previous versions of MEDx.

• Lower and Upper Threshold in Statistics. Lower and Upper thresholds are now included in the Statistics dialog box. In addition, interactive statistics will now use the optional parameters. This new feature will be beneficial if one wants to calculate the volume (or other statistic) for voxels within an image that lie between a specified threshold range.

• Improved Spatial Filtering GUI

• MxWriteAIR, MxCombineAIR. Two new ImageScript functions have been created to access the WriteAir and CombineAIR binaries provided by Roger Woods at UCLA. These functions allow the user to write to disk the Automated Image Registration (AIR) transformation file as well as the ability to concatenate the AIR transformations into one file.

• B-Short/B-Float File Readers. Two new file readers have been made available. With these readers, the MEDx user may now open B-Short and B-Float image files. Format options allow the user to specify the x,y,z voxel size.

• Option to disable the Progress window. Entering the ImageScript command MxShowGasGauge 0 in the ImageScript Interactive Shell will disable the progress window. Henceforth, the user will not see status, in the form of percentage complete, for any of their MEDx operations. This option is saved to disk as part of the user's defaults. The progress window will remain in a disabled state until the command MxShowGasGauge 1 is entered into the ImageScript Interactive Shell.

• Snap to Closest Slice after panning. The middle mouse button may be used to pan (translate) a volume seen in lightbox mode. Typically this option is used when zooming up on a slice in the volume to view small structures. A new option has been added to the Display menu to re-format the lightbox after panning. The slice closest to the upper left corner will be placed in the upper left corner of the lightbox.

• MxScroll Slice number. A new ImageScript command has been developed to allow for scrolling to a desired slice of a volume in the lightbox view. Entering the ImageScript command MxScroll $volume $slice_number will place the slice number defined in $slice_number of the volume defined in $volume in the upper left-hand corner of the lightbox viewer.

  File->ImageScript Execution

• Time Stamp Entry. A new dialog box has been included in the Measurement menu to allow scan acquisition time and duration information for image data. This information will be stored in the Image Header and will be used when computing Time Activity Curves.

• Byte Swap and LPS Orientation options for AVW images. Two new options are available for AVW (Analyze header) image data. The user may save their images in AVW format and specify the byte order. This is convenient if the user is saving an image on a machine with LSB first (PC Linux) and wants to later open the image data on a machine with MSB first (Sun Solaris). An additional option places LPS orientation information (Left-Posterior-Superior) in the AVW header when an image is saved. Henceforth, this image data will be opened into MEDx with the correct orientation.

• New X-Axis Label for Waveform Editor. The number of scans in a time series of data is now displayed as part of the user interface for constructing a model to be used for correlation or multiple regression.

Top