MIPAV change list pre-1.0

No changes listed.

• Now reads and writes ICS images
• Redirect output to log file
• More file formats are available in the file filter. Check the options after installing this version
• added ability to hold onto more than just the last 4 loaded images
• DICOM reader includes more advanced DICOM sequence processing. DICOM save may be adversly affected.

• Shortcut saving a user-defined LUT button added to the LUT dialog
• Updates to RayCast renderer: faster, among other improvements
• Seperable kernel added as option for blurring; option is much faster, but consumes a great deal more memory
• Improved shear warp renderer
• DICOM reader has some fixes for problems introduced because of 4D support
• Additional license information added to the about-file
• Previous bundled linux JRE had some problems; it has been changed, but is still being tested

• Shortcut saving a user-defined LUT button added to the LUT dialog
• Updates to RayCast renderer: faster, among other improvements
• Seperable kernel added as option for blurring; option is much faster, but consumes a great deal more memory
• Improved shear warp renderer
• DICOM reader has some fixes for problems introduced because of 4D support
• Additional license information added to the about-file

• XML schema includes thumbnails
• Improve ray cast volume renderer and image capture for movie imaging
• Improved Surface Renderer
  • Image capture for Movie
  • Opacity function options and R/W
  • Clipping plane for eye
• Improved Macro robustness
• DICOM reader can now handle GE FastCard 4D datasets
• Improved Image Browser
  • Display of thumbnails
  • Faster and uses less memory
• 3.5D OAR registration
• More documentation in the Help files
• Improvments to the secure file transfer and chat GUI
• Secure file transfer and chat uses a new version of the SSL implementation

• The bundled JVM is now Java 1.4.2.01, with Web Services 1.2.1 for XML support and Java 3D 1.3.1.
• To improve program responce, we are now pre-processing source code using Dash-O as an optimization tool.
• Coherence diffusion and regularised isotropic diffusion support:
  • 2D
  • 2.5D
  • 3D
  • both color and greyscale
• Crop uses less memory
• Scripts can now save data from the Data Window
• Maximum Intensity Project and Digital Reconstruction Radiograph modes have been added to the interactive volume renderer.
• Improved Surface Renderer
• Added surface smoothing method that does not shrink, leaving total volume approximatly the same.
• OAR registration is more robust
• Added a "Global Data" area to the output Data Window
• Added ability to read LZW Compressed TIFF and LSM file formats
• Added Brightness and Contrast controls to DICOM and Image browsers.
• Refined Co-localisation for Microscopy

• We again support the MAC OS X platform!
• Major improvements to the surface visualization tool.
  • Added 3D and 2D texture VOLUME visualization with opacity control.
  • Improved coordination with tri-planar view. For example, selecting a position on a surface in the surface viewer automatically updates the tri-planar view to the same location.
  • Opacity of surfaces, volume and cut planes can be selected by the user.
  • Improved visualization recording for making movies.
  • Stationary light has been added.
  • New toolbar for quick access to tools
  • Many more improvements.
• Added ability to read and write SPM formatted files (similar to Analyze format).
• Added ability to write simple AFNI formatted files. We already read AFNI formatted files.
• Improved image cropping utility. Origin information correctly updated.
• Improved add image margin utility. Origin information correctly updated.
• Changes to the Surface extraction algorithm (see algorithms -> extract surface):
  • retains image coordinate space information
  • triangle consistency check is implemented to insure all triangle are all either CW or CCW.
• Improved the DICOM receiver to catch images from the Kodak AutoRads. The Clinical Center had recently upgraded to a new version of software on the AutoRads.
• Improved general functionality to the image transform algorithm GUI.
• The bounding box of a VOI now displays the origin, width and height in both pixels and units of measure (eg. mm) and position information.
• From the animate player added ability to save color AVIs so that VOI contours would retained. Option to still save in greyscale with RLE is available. Grey scale RLE encoded AVI are much smaller but you lose VOI information. Also ability to save in other compression formats including: MPEG4 version 2, MPEG, and Quicktime. New compression formats require a separate installation of Java Media Framework.
• Autocorrelation and autocovariance calculation algorithms added.
• Colocalization algorithm added. This algorithm is based on Costes and Locketts' method.
• Improved scripting tool.
• Reduced memory usage in Utility -> Convert Type. This utility is used to convert from one image type (byte) to another image type (float) and its data range.
• Improved Utility remove slices. User can now specify a range.
• OAR 2.5D (2D time series) registration now uses much less memory.
• Save specified DICOM tags into our XML file format.
• Improved Livewire segmentation, faster, more robust and ability to remove last achor point with a right mouse click
• Subsample by 2, 4, or 8 utility now available.
• New algorithm to tranform image so that each image dimension is a power of 2. Used with volume visualization tool

• Secure file transfer functionality added to the file menu. Use SSL java library to setup a secure link using RSA128 encryption. File transfer includes automatically zipping the file(s), securely sending the zipped file, unzipping the file and saving the received files on the receiver's local drive.
• New non-linear smoothing method that smoothes along an optimal path.
• Upgraded and validated DICOM query and retrieve functionality on Kodak and AGFA PACS achieves.
• Found and fixed memory leak in utility to remove image slices.
• Improve macro recording.
• More algorithm documentation
• Added memory check. MIPAV resets memory allocation to 100M during reinstallation. Upon starting MIPAV for the first time after reinstalling, memory is settings are checked and the user is notified if the old memory allocation is different from the 100M. The user then can select the old memory settings.
• AVI writing improvements. RLE option decreases file size.
• Many other small improvements and bug fixes.

• Fixed memory leak in the utility to convert a _3D_ image to 4D image set.
• Added a new algorithm to evolve a triangulated surface using the GVF.
• Fixed a bug that could hinder saving 4D datasets in the MIPAV XML format.
• Slider value fixed for 4D images and the lightbox view.
• Faster DICOM file reader, for some images speed increase by as much as ~200%.
• Two new adaptive filters smoothing filters. Algorithms -> filters (spacial) -> Adaptive noise reduction and Adaptive smooth. See MIPAV help for explanation of the algorithms. For 2D images.

• Enhancements
  • New web page!
  • Online help using JavaHelp for many algorithms -- more to come. For example see Algorithm -- Filter (spacial) - Gaussian Blur.
  • Link two images together so that they scroll together. Works also in the Tri-planar view.
  • New MIPAV file format based on XML. The XML Schema for the file format is in the file image.xsd. This file is located in the mipav install directory.
    • The XML formatted header file is stored in a separate file and stores a link to the image file.
    • The XML file is a text format easily viewed with any text editor as well as internet browsers.
    • The MIPAV Schema is flexible and allows for many optional tags. For example. The users may (optional) add [0 -- infinity] parameters that have the form Name, Description, Value type, Value, and Date/time.
    • Saves LUT, matrix, resolution, axis orientation information.
    • Link to another image so that the linked image is automatically opened. See menu: Image->Attributes->edit attributes
  • Crop and Add image margins utilities now correctly retain and/or update important image tag information like image origin, image orientation, dimensionality, etc.
  • Modified internal parameters of the brain extraction algorithm to be more robust and improve segmentation (less tissue included anterior -- near the eye).
  • Levelset segmentation added
  • When the cursor is in an image, right click and hold the button and move the mouse to adjust window and level of the image. Level is changed by moving the mouse up(+) and down(-). Window is changed by moving the mouse right (+) and left(-).
  • A utility function to convert a series of images loaded as a _3D_ dataset into 4D dataset.
  • Added utility to save a 4D image to a series of 3D volumes and save a 3D image to a series of 2D images.
  • Undo last VOI change.
  • VOIs are now saved in one of two formats *.voi and *.xml. See voi.xsd schema that describes the new VOI format.
  • Watershed segmentation algorithm automatically saves the gradient magnitude image to a file. The next time the Watershed algorithm is called, it will automatically load gradient mag. image that that it doesn't need to be recalculated.
  • Image and DICOM browser are faster and have a better GUI.
  • Option to smooth a contour using a Bspline. See menu VOI -> smooth VOI.
  • Two options added to OAR3D registration:
    • Subsample image can be turned off -- only select for small images.
    • Option to skip search -- jumps directly to last optimization. Only choose this option if the images are approximately aligned (~5 degrees, 10 mm). This makes the registration twice has fast but accuracy or ability to converge could be reduced.
  • OAR 2.5D and 3.5D registrations algorithms where rigid registration was requested, a graph of the resultant translations and rotations is produced.
  • OAR 3.5D added an option to select a reference volume or register to an average volume
  • Improved the speed of 2D, 2.5D and 3D OAR registration by 25% - 30%.
  • A button has been added to the top toolbar to propagate a contour of a VOI to all slices.
  • A middle mouse click in the 2D image frame will cause the triplanar view to update to that location (if open).
  • Remove GUI bugs for VOI boundary evolution algorithms.
  • Option to ask user before closing frame -- see Help->Mipav Options.
  • Image frame registry displays image frames and images and allows the user to bring a selected frame to the front.
  • Updated DICOM anonymization to include the option toanonymize UIDs.
  • Upgraded to newer DICOM3 element dictionary.
  • Manual (semi-manual) registration tool now allows the user to select two 2D images to register or a 2.5D (2D images where the 3rd dimension is time) time series. Many GUI improvements also.
  • Trim parameter (Set from VOI menu ) now works as expected.
• Bugs fixed
  • Read LUTs from a read-only file (e.g. CD).
  • Fixed a bug in Histogram Equalization algorithms
  • VOI retrace mode no longer randomly adds a new point at the origin
  • VOI - bring to front, bring forward, send to back sent backward now work correctly.
  • Bug fix to thin-plate-spline registration method.
  • Found and fixed a number of memory leaks.

• Added VOI statistics calculator (VOI -> Statistics Generator menu). This dialog allows the users to select all or some VOIs and which statistics to calculate on the selected VOIs. The output is to a file and the log window in tab delimited format. The user is able to indicate weather the statistics can be calculated on a VOI basis or for each contour of a VOI
• File menu now records last four images open for quickly opening often used images.
• DICOM image browser for browsing (thumbnail view), sorting, and opening DICOM images.
• Simple background subtraction algorithm. User identifies background with a VOI and the average or median value is subtracted from the image.
• 4D (3.5D) time series registration based on MCFLIRT (FSL group in UK) called TSOAR.
• Improvements to simplify plugins
• Surface viewer:
  • Added ability to adjust opacity of the surfaces.
  • Fixed a smoothing surface bug.
• Added Adaptive Skeleton Climbing algorithm used to generate surfaces to the algorithms menu.
• Numerous smaller improvements and bug fixes.

• Added 2D and 3D registration for color images (least squares cost function)
• Improvements to image browser (e.g. open full image from thumbnail image)
• Analyze color images can now be read and written
• Fixes to DICOM anonymization tool
• Improved DICOM reader robustness
• Fixed bug is threshold dialog

• File format improvements and additions
• Read Bruker format
• Read Biorad format
• Read Ziess 510 format
• Read Freesurfer image and surface formats
• DICOM reader can now read lossless JPEG
• Load and save array of images in separate files of type JPEG, PICT, TIFF
• Improved DICOM directory anonymization
• Microcat images can be opened by either selected the .log file or one of the image files .ct
• VOI:
  • Save and Open all for VOIs
  • Select VOI on mouse release -- fixes problem of accidentally moving VOI when selecting it.
  • VOI polygon either open or closed based on selection of last mouse click (last point or first point of the VOI).
  • Euclidian length of VOI ployline is now displayed
• View:
  • Simple window and level tool so that the complete Histogram LUT tool does not need to be opened.
  • Tri-planar view updates image views and position information when selecting a point on a surface in the Surface Render viewer.
  • Lightbox view has many new options for extracting slices and dynamically updating the display of the images.
  • New Image browser. User chooses a directory and a directory tree is generated. The user is then able to scan the directories and quickly view images with relevant image information displayed (dimensionality, voxel size, etc. ).
  • Improved contrast mapping, especially with CT images.
• Algorithms:
  • Added MRI "deskulling" algorithm (based on the Brain Extraction Tool "BET")
  • Advanced image calculator: combine images of sophisticated functions using calculator GUI.
  • Added Histogram matching of two images using their probability density function.
  • Improved adaptive histogram equalization - faster
  • Convert RGB to gray with weightings input by the user option.
  • FFT 2.5D option added
  • Improvements to OAR (Flirt variant) registration
    • VOI option to identify sub regions to be registration.
    • Fix weightings options
    • Improved accuracy and speed
  • Upgraded to AIR 5.07
  • Livewire added (interactive VOI generation). User can select one of three functions to optimize using Dykstra's method. (2D)
    • Gradient information,
    • Laplacian medialness.
    • Intensity.
  • Fuzzy connectedness region grow method is now 2x faster.
  • Principle components filter for reducing noise.
• Many more improvements and bug fixes.

• Lightbox improvements.
  • Faster and uses less memeory
  • Select and delete image slices
• File format improvements
  • Improved DICOM sequence reader
  • Read MicroCat CT images (select *.log file)
  • Improved MINC writer (faster and uses less memory)
  • Improved MINC reader (more robust)
  • Improved Siemens reader (more robust)
  • Improved MRC file reader
• Surface renderer
  • Improved lighting controls
  • Capture to AVI movie
  • Surface volume calculation
  • View bounding box option
  • Adjust background color
  • Smoothing surface option
  • Save surface to VRML format
  • Added toolar button to invoke surface renderer
• Added Fuzzy Connectedness region grow algorithm
• Simple wavelet denoising algorithm
• Polyline VOI now displays the length.
• Paint to VOI option now available in VOI menu
• Improved screen capture tool (under file menu)

• 2D and 3D image registration (intensity based) using a choice of least squares, normalized mutual information, mutual information or cross correlation cost function.
• Tri-planar upgrades
  • Radiologically correct orientation of DICOM / AFNI / GE / MINC / SIEMENS formatted images
  • Improved speed.
• Read GE formatted images
• Read and write Interfile format (A nuclear medicine format)
• Improved MINC image format including MINC label files.
• Local normalization algorithm using to reduce intensity gradient across images
• GVF snakes
• Improved surface renderering with the ability to capture to AVI movie format.
• DICOM anonymous tool - anonymize whole directories of DICOM formatted images
• Macro toolbar
• Improved logging
• Transform option include skews
• Improved adaptive histogram equalization
• More interpolation options: 3rd, 5th, or 7th order Lagrangian
• Many user-interface improvements.

• Major tri-planar upgrades:
  • Correct orientation of DICOM formatted images
  • Ability to put images into Talairach orientation (AFNI-like)
  • Overlay Talairach grid
  • Display DICOM or Talairach cursor position
• Read AFNI images
• Read Siemens' Magnetom Vision format (view all header information).
• Read Metamorph formatted tiff-extended images
• Read/write MINC region of interest files
• Read/write FITS images
• Read/write/modify transformation matrices (MIPAV or FSL format)
• Surface extraction and decimation (triangulated mesh) from VOI.
• Surface rendering tool (triangulated meshes) and orthogonal image viewer.
  • Ability to add multiple surfaces and adjust properties of surfaces
  • Control lighting
  • Zoom/rotate/translate surfaces and record mouse function for playback.
• 3D height surface of 2D images.
• Mathematical Morphology improvements. Now supports edge extraction and skeletonization.
• Graphing component now supports interactive time-activity-curves from point or contour VOIs.
• Image transformation now supports the transformation of VOIs.
• Improved Analyze header viewer.
• Ability to load two images into a single frame with alpha blending without first saving image to a file.
• Newer version of documentation.

• MIPAV bundled with Java 1.3.1
• MRI Inhomogenity correction - N3 algorithm is now available (Sled et.al.)
• Adaptive histogram equalization has been added. Significantly improves image contrast
• Checker board for two image display
• Movable window which display a second image over another image. Great tool for checking registration results.
• Read Cheshire formatted images and ROIs
• Automatically determines big or little endian ordering when reading Analyze images
• Image -> attributes -> show header menu item now displays image endianess
• Change endianess using check box found in Utilities convert type dialog
• Enhanced TIFF reader to read "tiled" images
• Improvements to the 2D semi-manual registration tool
• 2D/3D Thin-plate spline landmark registration algorithm added
• Save DICOM tags to a text file.
• 3D orthogonal image viewer using Java 3D
• windowed sinc interpolation option available in algorithm transform
• Improved memory monitor display
• Utility to extract 3D datasets from 4D datasets in any dimension (i.e. extract an image slice from the same location over time).
• Non-maximum edge suppression (i.e. Canny edge) algorithm added

• Fixed the annoying menu problem for Windows OS
• Added the ability to save the Lightbox configuration parameters
• User preferences are now saved in the user's directory. Now users no longer need to share a preference file
• Added Dzung Pham's Fuzzy C-Means algorithm for gray, white, and CSF segmentation
• Added additional functionality to the AVI reader
• Read/write Quick time files
• More speed-ups for transform algorithm
• New toolbar buttons for quick access to tri-planar and lightbox views.
• Improved Analyze file format reader
• Users are now able to open a series of 3D images (individual 3D Analyze formatted, consecutively ordered images) into to a 4D dataset in Analyze format. (use File -> open -> multi-file)

• DICOM
  • Query/Retrieve improvements (faster & less memory)
  • Now reads multiframe and RGB images
• AVI reader/writer (uncompressed, RLE8)
• Added 4D gradient magnitude and dual-threshold filters
• Added Gaussian blur, gradient magnitude, and median filters for RGB images. Each color channel (r, g, and/or b) can be processed independently and is selected by the user in the respective algorithm dialog.
• "Painting" of complex images and RGB images is now supported. Painting complex images (Freq. domain) allows the user to remove specific artifacts.
• Improved dialog for region grow (paint can) in the paint toolbar
• Bug fix for Morphology operations for odd shaped images
• Significant speed improvements to image transformation algorithm.
• Added algorithm that quantifies image data based on a mask image. This method is available in the Algorithms -> quantification using mask menu. An automated version is in LUT dialog via the Threshold button.

• Tri-planar view:
  • Added 3D crop feature
• All views (Lightbox, Tri-planar, 2D planar ) support 3D image with a 4D overlay image. For example: 3D structural MRI and a 4D functional MRI
• Added 4D Gaussian Blur option
• Rotate and Insert image utilities now update DICOM image tags more accurately
• Added Utility function to swap the 3rd and 4th dimensions
• DICOM reader/writer improvements:
  • Anonymize function purges more tags per Supplement 55 of the DICOM spec.
  • Better use of tag information when displaying the image
• Added "Memory allocation" function to adjust the amount of memory that MIPAV uses. Default is 100MB. Be sure to restart MIPAV after changing these values. (The mipav.lax will no longer need to be manually changed)
• Image header information has been moved from Help -> About Image to Image -> Attributes -> View header
• General improvments to the interaction of image views. For example, double clicking in the light box view on a specific image will automatically update the 2D and tri-planar view to that image.
• DICOM images can now be registered based on Image position and orientation tags.

• Tri-planar view will now:
  • allow point VOI adjustment in any of the three views - used for Landmark registration
  • improved display speed
• DICOM reader improvements:
  • now able to read DICOM (via secondary capture images exported from the Cheshire program.
  • improvements to better handle compliant and somewhat non-compliant DICOM images
  • Improved flexibility when exporting non-DICOM images to DICOM images via secondary capture.
• Point VOI improvements:
  • when reading and writing to a file
  • group points in a single VOI
  • improved labeling

• Tri-planar view will now display crosshairs in all three views
• 4 new preset LUTs: red, green, blue, and gray ramp with blue in first location and red in the last location
• Utility to add image margins to an image 2D and 3D
• Animate viewer:
  • will now process and display 4D images
  • allow interpolation when constructing image of a different size for display
• More/updated documention
• Fixed bug that would not allow RGB images to be converted to grayscale
• Improvements to image resample function:
  • aspect ratio lock/unlock
  • size slider

• Tri-planar view will now
  • display 4D images
  • overlays two images with alpha blending
  • improved image alignment
  • painting in all three views
• DICOM Q/R improvements
  • if study and/or series number is missing study and series descriptions now used
  • general improvements to GUI functions
• Improvements to the protractor VOI
• More/updated documention
• Minor bug fixes

• MIPAV is now bundled with JDK 1.3 for Win32, UNIX, Linux but not MAC. MIPAV should now run much faster on the SUN and Linux platforms
• Added protractor to VOI tool bar for measuring angles
• Added a new VOI - landmark registration algorithm which is presently used to register 2D images over time.
• Tri-planar view now has the capability to align images. This is one step to aligning volumes in Talairach space. More functionality to come.

• Fixed bug that affected the display of images on the SUN platform

• Improved DICOM Q/R table to improve layout.
• DICOM file reader more robust when reading images with inaccurate "pixel-padding value".
• Working on tri-planar view
• Working on Fuzzy C-means segmentation
• Added more documentation - see User's Guide web page
• Other minor bug fixes

• Fixed bug introduced in V0.98b that caused DICOM Q/R and send problems
• VOI size can now be interactively modified using the VOI's bounding box
• Lightbox view now using less memory
• Added more functionality to Tri-planar view - still adding more

• Fixed bug that caused MIPAV not to execute if the mipav.preference file was missing
• Quick LUT - new toolbar button to draw a box in the image. The min and max of the area bounded by the box are used to remap the LUT. Saves time so one does not need to open the full LUT dialog.
• Manual registration now reads/writes landmark points from a file and uses them as imput to a least squares algorithms for 2D registration.
• Read DICOM images in order by: Position if x,y,z dimensions are all spacial. If position is the same for each image slice (i.e. x,y, time) then Order by Instance Number. If the Instance Number is not in the header then Order by the file order in the directory
• Improved MINC file writer
• Interactive levelset VOI now works with RGB images
• Uses less memory in Cine loop.
• Paint feature more robust on various platforms

• Improved DICOM file reader/writer.
• Improved netCDF/MINC file reader/writer.
• Any file type (i.e. DICOM, raw) to MINC dialog.
• Added install/remove plugins methods.
• Added orthogonal (tri-planar) viewer - more functionality to be added.
• Fixed bug that would not allow RGB images to be displayed in Lightbox viewer.
• Added function under Utilities menu to rotate 3D volume 90 degrees in the 3 orthogonal directions (x,y,z).
• Fixed FFT bug.
• Improved Watershed algorithm to use a little less memory.
• Added (2D) manual registration tool under Algorithms -> Registration.
• Added method to control the automatic trimming of VOI points (see VOI menu).
• Improved DICOM sanitization (more tag options).
• Added ability to disable DICOM receiver (see Help menu).

• Debugged and tested query and retrieve on Kodak/Cemax PACS
• Tested and verified MIPAV as a receiver of images from the Kodak/Cemax Autorad device.
• Improved Query and retrieve from GE PET scanner.
• Improved DICOM PET file reader.
• MIPAV to MIPAV DICOM send/receive function.
• DICOM - Verification function added. Used to test host connectivity.
• Fixed bugs in Graph module.
• Added animation viewer for showing movies.
• NetCDF and MINC file reader / writer.
• Flip images ( x or y axis).
• Fixed bugs in image transformation.
• Fixed bug when generating VOI using interactive level set.
• Mask to VOI function added. (Converts mask image (binary - short) to VOIs).
• Added function to convert 3D gray-value dataset to 3D RGB dataset.
• Paint region grow now uses (First In First Out) FIFO data structure. Also added parameters to limit size and/or distance from the seed point.
• Paint toolbar paint intensity level now displays the correct value.
• Paint can (i.e. region grow) now has one level of undo.
• Add Bspline snake algorithm (2D and 2.5D).
• Fixed 3D RGB (TIFF) write bug.
• Misc. bug fixes and speed improvements.