Analyze 8.1: New Features and Enhancements

Module Specific Enhancements
New Support Features

Module Specific Enhancements

Volume Render Module

- Interpolated Rays option: The 'Interpolated Rays' option improves the quality of visualizations generated when anisotropic volumes are being scaled interactively, for data in oblique orientations, and for thin object. Rendering rays are cast using floating point calculations to interpolate the voxel contributions along each ray by weighting the contributions of voxels in each voxel neighborhood. This will significantly increase rendering computation time.

- Fix Tool Related Volume Correction option: The Fix tool in the Volume Render module has been enhanced with a new slice correction option. The new option ‘Copy from Related Volume’ provides the ability to use a related volume as input for the Fix Tool, allowing images to be copied from the Related Volume to the Loaded Volume.

Registration Modules

- Support for simultaneous transformation of associated object maps: Object maps associated with the match data set can be loaded and registered simultaneously with the data.

Tree Analysis Module

- Improved Tree Analysis algorithm: The algorithm has been updated with an increased tree level limit. The enhancement increases the tree level limit from 702 to 2054.

Diffusion Tensor Imaging (DTI) Add-On Module

Extended support for more than 6 directional gradient diffusion volumes:

- 4-D mulitvolume support: Multivolume input support has been added to the DTI module to allow for easier loading of mulitple directional gradient diffusion volumes.

- Extended support for reading gradient directions: When the volumes are loaded into the DTI module, Analyze will attempt to read the gradient directions from the header. If the gradient directions are found, they will be displayed on the volume selection button. If they are not found, gradients can be entered either in the interface or by importing a text file with the encoding directions.

- White matter fiber tracking and visualization (tractography): The DTI module has been enhanced with Fiber Tractography, which provides the ability to visualize white matter tracts.

- Tracking specific anatomy via use of object maps (ROIs): The DTI module now provides the ability to load object maps in order to specify regions of interest for fiber tracking.

- Visualization with multiple orthogonal slices: The DTI module provides interactive display of intersecting orthogonal sections with 3-D display of fibers.

New Platform Support

- Intel-based Mac platform support: The Analyze 8.1 release is the first version of Analyze available with full Intel Mac support. This port of Analyze runs with MacOS X 10.4+.

- Windows 64-bit platform support: The Analyze 8.1 release is the first version of Analyze available with true Windows 64-bit platform support. This port of Analyze runs with Windows XP x64.

- New Windows Vista platform support: The Analyze 8.1 release is the first version of Analyze available for the Windows Vista operating system.

New Support Features

Analyze Viewlets

Analyze Viewlets are tutorial movies that can help you become familiar with the Analyze interface and its basic and advanced operations. More than 50 viewlets have been created for the Analyze 8.1 release and they are accessible directly from the "Help" menu. They are also available on the Analyze Viewlets page

Live Online Technical Support, Demonstrations and Training

With the release of Analyze 8.1 we now have the ability to do live, interactive technical support, demonstrations and training with our live remote-assistance tool. This tool gives our Analyze Application Support staff the ability, with your permission, to log on to your system to demonstrate functionalities within Analyze, provide technical support if issues arise, and train users in specific areas of interest. During an online session, our staff can view your desktop and share control of your mouse and keyboard. You are in full control of your computer at all times, having overriding control of your mouse and keyboard and can end the screen-sharing session at any time. Your session is completely secure with proprietary compression technology and 128-bit AES encryption.

In our testing throughout this year, the live remote assistance tool has completely transformed our service offering. This resource is available to our customers who hold a Maintenance Agreement at no additional cost. For customers who do not have a Maintenance Agreement, the cost is $250.00 USD per hour with a minimum one-hour charge.

New Maintenance Agreement Benefits

In addition to unlimited priority technical support, discounted workshop training and free access to upgrades, maintenance agreement holders get up to four hours of free access (per maintenance agreement) of live online support, training, and consulting. Please contact Brandon Pennington by e-mail at brandon@analyzedirect.com or by phone at +1 (913) 338 2527 ext. 201, for details on how to obtain a maintenance agreement for your Analyze license.

Analyze 7.0 Enhancements

Here's a list of all the new features and enhancements that were previously built into Analyze version 7.0.

New Modules and Optional Add-Ons
DICOM Tool - A New DICOM Image Management Module
MRI Diffusion Tensor Imaging (DTI) Add-On
Image similarity metric computation Add-On
Predictive T2-weighted MR image generation Add-On
3D Brain Atlas Add-On

General Enhancements
Enhanced implementation of Object Maps
Parametric volume visualization
Enhanced colormap definition for color tool

Module Specific Enhancements
File Management
Load As
Volume Render
Spatial Filters
Watershed Segmentation
Object Extractor
Voxel Registration / Surface Matching Registration
Point-to-Surface Registration
Region Of Interest
Object Counter
Software Structure / Distribution Enhancements

New Modules and Optional Add-Ons

DICOM Tool - A New DICOM Image Management Module
DICOM Tool is a completely new functionality for the import and management of DICOM images. DICOM Tool replaces the current BIRPacs functionality, including the Analyze client, the database, and the DICOM receiver. The client includes a much-improved interface to data selection and loading, including an image viewer for examining the images prior loading. The underlying database is new, providing much improved cross-platform compatibility, and new tools allow direct management of the receiver component directly from the client interface on all platforms. Information for migrating existing BIRPacs databases and stored images is also provided. DICOM Tool is not an optional add-on, and is directly integrated into and immediately accesible from Analyze 7.0.

MRI Diffusion Tensor Imaging (DTI) Add-On
The Diffusion Tensor Imaging (DTI) Add-On computes and saves the primary diffusion maps from diffusion-encoded MRI data. Seven datasets are imported including the baseline volume and six gradient encoded volumes. The diffusion tensor is estimated from these volumes and standard diffusion maps are generated including: apparent diffusion coefficient (ADC), relative anisotropy (RA), fractional anisotropy (FA), and volume ratio (VR) maps. In addition, a color-encoded volume is generated to display direction of the principal diffusion vectors at each voxel.

Volume Metrics Add-On: Image Similarity Metric Computation
The VolumeMetrics Add-On provides sophisticated tools to assess the quality of multiple associated volume images, including those derived from image registration and segmentation techniques. This Add-On allows the user to:
- assess the similarity of two co-registered three dimensional volumes with/without ROI using parametric (15 measures) and non-parametric (12 measures) similarity indices.
- compute directional change detection between two volumetric images using a fast and robust change detection technique.
- assess the concordance of different segmentation results with ground truth using 47 popular binary similarity metrics.
- to quantitatively assess the inter- and intra-rater agreement of multiple segmentations.
- compute statistical shape atlases from multiple segmentations.

MR T2 Projection Add-On: Predictive T2-weighted MR image generation
The MR T2 Projection Add-On optimizes the display of T2 contrast in MR images using a simple, fast, robust, three-dimensional and prior-information-driven technique by rapidly projecting T2 weighted MR Images across (TR,TE) space. The projections may serve as a preprocessing step to improve the quality of further image analysis and visualization with T2 MRI images.

Mayo 3D Brain Atlas Add-On
The Mayo 3D Brain Atlas Add-On provides a unique integration of the new Mayo 3D anatomical atlas of the human brain and the Analyze biomedical imaging software system for the purpose of using the defined anatomical components in the brain within Analyze modules. The Mayo 3D Brain Atlas provides four key steps in integration of the brain atlas:
1. Interactive AC-PC based alignment of volume image data to the Talairach-Tournoux coordinate and proportional grid system.
2. Interactive scaling of the proportional grids to improve the spatial registration of the 3D Mayo Brain Atlas with the volume image data.
3. Single or multiple point query of Talairach/volume coordinates to identify the nearest Talairach label to the queried point.
4. Output of atlas-derived anatomical component maps for use with other Analyze modules.

General Enhancements

Enhanced implementation of Object Maps
- Support for 4-D Object Maps

Parametric volume visualization
- Parametric information encoded from additional related volume
- Integrated with 2D and 3D displays
- 2D parametric map overlays
- 3D parametric map rendering (integrated with other object map controls)

Enhanced colormap definition for color tool
- Simple mechanism to establish colors for ranges of values, i.e., set 0-10 to be red, 11-50 to green, 51-100 to blue, etc.

Module Specific Enhancements

File Management Modules
New image file format support
- Siemens/CTI MicroPET/MicroCAT format
- NIfTY neurofunctional image file format

Load As Module
- Added slider to Interactive Subregion tool to set beginning and ending slices.

Volume Render
- New tissue map control interface for volumetric compositing
- Advanced thresholding techniques for both global threshold determination and multilevel thresholding (presets)
- Added interactive orthogonal section display to Ortho Sections tool rendering

Spatial Filters
- Added distance transforms - chamfer, Euclidean, etc.

Watershed Segmentation
- T2-weighted MRI preprocessing for brain segmentation

Object Extractor
- T2-weighted MRI preprocessing for brain segmentation

Voxel Registration / Surface Matching Registration
- Added user-selected landmark correspondence registration to manual mode

Point-to-Surface Registration
- Integrated piecewise point-to-surface registration
- User-defined association of points to corresponding base objects
- Output of multiple transformation matrices

Region Of Interest
- Added area and volume measurements (uncalibrated and calibrated) for threshold ranges - Above, Between, Below thresholds
- Added 'Entropy Measures' for single volume analysis

Object Counter
- Extended to allow 3D object counting
- 3D objects segmented by size (connectivity) and shape (sphericity, etc.)

Software Structure / Distribution Enhancements
Analyze Software Structure Changes
- Upgraded to latest release of Tcl/Tk 8.4.11
- Upgraded to latest release of ITK

New platform support
- Itanimun II 64-bit version (PC_LINUX64)

Integrated Analyze distribution assistance
- Automated Patch Update Tool

Analyze 6.0 Enhancements

Here's a list of all the new features and enhancements that were previously built into Analyze version 6.0.

The Insight ToolKit (ITK)
ITK Filter Module
ITK Segment Module
ITK Seg 3D Module
ITK Registration Module
Macintosh (MacOS X) compatible
Continued enhancement to DICOM image support
Load/Load As enhancements
Import/Export enhancements
Multiplanar Sections enhancements
Oblique Sections enhancements
Volume Render enhancements
Region Of Interest (ROI) enhancements
Brain Atlas enhancements

The Insight ToolKit (ITK): The Insight ToolKit (ITK) is an open-source toolkit of advanced filter, segmentation and registration algorithms contributed by leading scientists and institutions around the world under sponsorship by the National Library of Medicine as part of their Visible Human Project. ITK is seamlessly integrated into Analyze 6.0 (and AVW 6.0) to provide to researchers the powerful functionality of ITK via a set of interactive Analyze modules. ITK functionality has been integrated into the Analyze/AVW environment at two levels. First, key ITK filter, segmentation and registrations functions are wrapped as AVW functions, making them directly accessible to AVW developers using the same API as for AVW function calls. Second, custom Analyze modules are now available from the main Analyze menu to interactively run the ITK algorithms. These new Analyze modules include:

ITK Filter Module: Adds both linear and non-linear filters to the collection of spatial filters offered in Analyze. These filters can be used for smoothing, gradient calculation, and feature extraction via distance transforms or edge detection. Specific filters include: bilateral, binary min/max curvature flow, binomial blur, B-spline decomposition, curvature anisotropic, curvature flow, Danielsson distance, derivative, discrete gaussian, anisotropic gradient, gradient magnitude, recursive gaussian gradient magnitude, min/max curvature flow, and zero crossing edge detection.

ITK Segment Module: Provides an interactive framework for applying sequential processing steps using ITK functions for segmentation of sets of 2-D slices based on region growing and level set methods. Filtering steps can be applied to preprocess the images, including anisotropic diffusion and flow filters, along with either gradient magnitude or recursive gaussian edge detection. Seed-based techniques can then be selected for target structure definition, including seeded region growing, fuzzy connectedness, a fast marching algorithm and watershed region extraction. Further mask-based algorithms can be applied, including shape detection techniques, active contour definition, and Canny or Laplacian edge extraction. Standard morphological operators can be applied to the segmented regions to further refine the segmentation, including erode, dilate, open, close, and connected component analysis.

ITK Seg 3D Module: Implements all of the functionality of the previous module in a full 3D volume image segmentation context. Each algorithm is extended to apply to the entire 3D volume image for full 3D segmentation using region growing and level set methods, with associated pre- and post-processing algorithms.

ITK Registration Module: Using the interactive multipanel registration interface developed for other Analyze registration modules, the ITK registration module implements the mutual information-based ITK registration algorithm for multimodal 3D volume image registration and fusion.

Macintosh (MacOS X) compatible: The Analyze 6.0 release is the first version of Analyze available on the Mac platform. This port of Analyze runs with MacOS X 10.3 and uses the native Mac interface (via Tcl/Tk) as the standard user interface for Analyze on the Mac. This port is still a beta release, as there are remaining Tcl/Tk-based issues that have yet to be handled by the Tcl/Tk community on the Mac platform.

Continued enhancement to DICOM image support: including the following:
Improved selection of closest orthogonal orientation from the direction cosine information in the DICOM header.
Support added for 24-bit RGB run length encoded (RLE compressed) images stored in DICOM format.
All lossless JPEG transfer syntaxes are now supported.
Robust handling of images that contain sequence information.

Load/Load As enhancements: including the following:
If the maximum and minimum values in the header for the volume image being loaded are equal, a prompt is offered to calculate the true max/min during loading or set these to specific values.
Clicking the right mouse button over the Input Maximum or Input Minimum fields in the Intensities panel will provide an option to Calculate Max/Min for the volume being loaded.
Loading of multivolume image files in non-transverse (sagittal/coronal) orientations is now supported.

Import/Export enhancements: The Volume Tool and the associated volume wrapper file format now supports the inclusion of individual 3D volume image files to form 4D data sets.

Multiplanar Sections enhancements: The Scan Tool contains a new feature to interactively report the voxel coordinate and value at the position of the cursor in the image display window.

Oblique Sections enhancements: Smoother, predictable transitions in oblique images are now computed for obliques generated along a traced curve (following a trace generated by Curved Sections). This keeps sections from flipping and/or rotating significantly as they are generated along the curve.

Volume Render enhancements: A new tool called the Fix Tool has been added to the Manipulate tools in Volume Render. This tool allows a variety of correction methods for missing or corrupt images that occur in a volume, including removing corrupt images, copying from an adjacent uncorrupted image, and interpolation from a range of uncorrupted images near the corrupt image.

Region Of Interest (ROI) enhancements: including the following:
A new tool called Propagate Regions has been added to the Tools menu to permit shape-based interpolation of objects (region sets) throughout the volume. The main purpose for this tool is to allow sparse definition of region sets throughout the volume (i.e., every other slice) that can then be interpolated into complete 3D regions (objects) using shape-based interpolation.
Additional support has been added to allow multivolumes to be used as the Related Volume, with controls for linked volume selection between the original and related volumes or independent volume selection for these multivolumes.

Brain Atlas enhancements: The Brain Atlas module now supports the use of object maps associated with the volume image being registered atlas. This permits the transformation of an object map associated with the input volume into the Brain Atlas registration space, providing both the transformed volume image and the transformed object map as outputs for further use.

Analyze 5.0 Enhancements

Here's a list of all the new features and enhancements that were previously built into Analyze version 5.0.

New 3D Volume Registration Interface
New Surface Rendering Module
New Object Counter and Analysis Module
64-bit Analyze
New Analyze License Manager
Improved Manual Object Segmentation Tools
Integrated Watershed Segmentation
Volume Render Presets
Multiple Object Stereologic Measurement
Improved ROI Region Definition
Continued enhancement to DICOM image support
More colorful interface

New 3D Volume Registration Interface: Reintegration of the 3D volume registration algorithms in a more intuitive, interactive, exceptionally fast interface, including improved visualization of the fused registration results.A snapshot of the new interface is shown here - click on the image for a larger view. Note that the interface now includes Transverse, Coronal, and Sagittal images for both the Base and Match volumes and the registered and fused images. The linked cursor is available in any window for scrolling through the images or identifying particular points of interest.

New Surface Rendering Module: Improved visualization and interactive manipulation of extracted surfaces integrated with the Analyze 'Object Map' concept for direct viewing of structures.

New Object Counter and Analysis Module: Automated segmentation using connected component analysis techniques and integrated mensuration of characteristic properties for the segmented objects.

64-bit Analyze: True 64-bit compatibility on 64-bit architecture systems to support visualization and analysis of very large volume images (>2Gb). This currently includes 64-bit support on Linux, Sun and SGI platforms.

New Analyze License Manager: As part of infrastructure support for Analyze, a new license manager application, BIRLic, has been developed by the BIR to provide better support for electronic license management and distribution. BIRLic replaces the current FlexLM license management system.

Improved Manual Object Segmentation Tools: Enhancement of the manual segmentation tools (i.e., Image Edit, Slice Edit) to allow for multiple object definition as a single-pass operation. This means that users can switch between different objects in the object map while making edits on the same slice. In previous versions of these editing modules, only a single object could be defined in a single pass. Integration of ‘Smart Edge’ automated boundary definition functionality with Spline definition – i.e., movement of spline control points automatically regenerates ‘Smart Edge’ boundary. The 'Smart Edge' boundary detection algorithm was actually added with version 4.0 for traced regions. The technique has now been extended to the definition of spline curves as well.

Integrated Watershed Segmentation: Direct integration of watershed segmentation techniques in Analyze segmentation modules, including Object Extractor and Morphology. Watershed Segmentation allows "skull stripping" of the brain from MRI datasets to be performed quickly and accurately.

Volume Render Presets: User-definable presets for volume rendering parameters, including thresholds, colors and assignment to specific volume information dependencies.

Multiple Object Stereologic Measurement: Improvements in Stereology to sample multiple structures as a single-pass operation (assigning sampled points to specific objects).

Improved ROI Region Definition: New ‘Split’ option added to the Edit Tool to support direct manual tracing to split an existing region into two separate regions (objects). The 'Edit Region' tool now has a check box called 'Split' that allows the edit trace to be used to split a region into two new regions, rather than just editing the boundary of the existing region. Click on the images below to see a demonstration of this option.

Before Split	During Split	After Split

Continued enhancement to DICOM image support:
Direct import of DICOM files into BIRPacs allowing the BIRPacs database to be populated with DICOM images transferred directly to the system via non-DICOM mechanisms (i.e., by direct copy rather than DICOM Send or Query/Retrieve); Generation of DICOM-compliant media using the DICOMDIR directory standard;
Additional supported image file formats:
Direct support for the standard AVI format for both input and output: primarily useful for output of rendered movie sequences direct to an AVI format file for use with other third-party products that support the AVI format (PowerPoint, etc.). (Note: PC Windows only)
New support for Bruker MRI format (Paravision) integrated with a tool for volume identification and selection.

A much more ‘colorful’ interface, and due to popular demand...color icons!!

Analyze 4.0 Enhancements

Here's a list of all the new features and enhancements that were previously built into Analyze version 4.0.

Major New Modules/Functionality

BIRPacs Add-On
Virtual Endoscopy
Surface Extractor
Application Critical 2D and 3D Registration
Coronary Artery Calcium Scoring
Interactive Tree Analysis
Talairach-Tournoux Brain Atlas Add-On
General Improvements

Module-Specific Enhancements

Load/Load As
Import/Export
Volume Render
Movie
Image Calculator
Spatial Filter
Image Edit
Morphology
Surface Extractor
2D Registration
3D Registration
Region Of Interest
Tree Analysis
Publish

Major New Modules/Functionality (top)

BIRPacs (top)

New DICOM communications for Analyze - BIRPacs provides the communications necessary for Analyze to receive, query, retrieve, and send images to other DICOM entities, eliminating the need to separately store DICOM image files and sort them into volumes for loading into Analyze. The features of BIRPacs include:
Volume image selection using Patient/Study/Series hierarchy.
Direct loading of series into Analyze as 3D volume images.
Rapid access to DICOM header information to review specific scan details.
BIRPacs database and DICOM receiver process running independent of Analyze.
Custom database queries from BIRPacs database (sortable on any information).
DICOM server administration for configurable DICOM entity name and port number.

Virtual Endoscopy (top)

New Virtual Endoscopy module with interactive endoscopic navigation interface and intuitive localization images, including 3D endoscopic object recognition and visualization. Includes improved sequencing ability for endoscopic fly-through sequence generation.

Surface Extractor (top)

Integration of former Tiler and Contours functionality into new, enhanced Surface Extractor module, with completely new and improved interface and additional output options to support connectivity to finite element analysis programs (via PATRAN format) and rapid prototyping systems (via binary STL output and using real-world scaled coordinates).

Application Critical 2D and 3D Registration (top)

New, application-critical algorithms added to 2D and 3D Registration methods, including:
Non-rigid 2D and 3D registration using multiresolution, piecewise-linear algorithm;
Nine degree of freedom NMI registration (scale space search);
3D 'point cloud' to surface registration (non-image based point coordinates);
Isocoordinate registration for anisotropic volume images (automatically resampled base).

Coronary Artery Calcium Scoring (top)

New application-specific module for Coronary Artery Calcium Scoring from EBCT image data, including output of Agatston score, total calcific area, and total calcific volume on a per-lesion, per-artery segment, and total heart basis.

Tree Analysis (top)

Important enhancements to Tree Analysis functionality to allow interactive editing of extracted tree information using rendered images from multiple viewpoints, and a new diagramatic tree map display to elucidate complex branching relationships.

Brain Atlas (top)

Implementation of Talairach-Tournoux Brain Atlas for registration to other volume images using an interactive Analyze tool and output as an Analyze image file or Analyze object map for further atlas-based interactions.

General Improvements (top)

Based on improved AVW 4.0 software library.

Support for new Windows ME and XP operating systems.

New, interactive method for establishing and managing user-specific default settings for module parameters, including:
'One-click' default setting to copy current parameter value as default.
'Defaults' editor for assignment of option-specific default settings.
Control of global window defaults, including size and position.

Image canvas print ability added to all Analyze modules for direct printing of the contents of any display canvas and tools windows.

Module-Specific Enhancements (top)

Load / Load As: (top)

Multiple image file loading via selection of multiple files in the file selection box (Load As - PC version only) and subsequent loading of all the volumes selected.

Default colormap control in Intensities panel (Load Colormap option) for loading all data with specific default colormap.

Per-slice Shear Factor correction for shifting slices during loading (see gantry tilt info in Import/Export).

New shape-based interpolation option for resizing of volume images during loading of binary volume images or object maps as volumes.

Support added to manage 64-bit float (double float) numeric value representation.

Import/Export: (top)

Continued enhancement to DICOM image format support, including:
Enhanced JPEG-compressed DICOM image management.
Support for eFilm DICOM images added.
Support for multiframe X-ray Angiographic (XA) cine loops added.

Automatic recognition and correction for CT gantry tilt FOV shearing effect, restoring slice-based registration along the acquisition axis. This is keyed by the presence of a value in DICOM Acquisition Gantry/Detector Tilt field (0018|1120).

For DICOM CT volumes, Image Rescale Intercept (0028|1052) and Image Rescale Slope (0028|1053) are now used to rescale the CT values, usually for the purpose of automatic rescaling to Hounsfield units. NOTE: This will occur for all CT data that is imported from DICOM data, so CT value ranges will be significantly different that with previous versions of Analyze (actually handled correctly now as Hounsfield units).

Improved Raw Data import tool, including examination of actual data values for exploration of image data format.

Volume Render: (top)

Direct rendering of difference volumes (created by subtracting loaded volume and related volume).

'Change View' rotation interaction improved by adding the ability to regenerate the rotated image upon mouse button release (previously only when cursor moved outside of the window).

Movie: (top)

Ability to output multipanel movie frames for 4-D data to an image sequence stored in the Analyze Workspace added to the Movie module. This allows saving of multipanel sequences for conversion to other format types (such as Quicktime) for use with other movie players outside of Analyze.

Image Calculator: (top)

Improved support of keyboard use for entry of numbers, operators, and actions ('Enter' same as '=', Backspace for deleting entries).

Spatial Filter: (top)

Added 'sticks' filter to provide an improved method for noise reduction. This filter uses circularly-symmetric rotational line segments as multiple filter kernels, outputing numeric combinations of the properties of voxel sets processed by these kernels (maximum, minimum, average max, average min, etc.). This directionally-dependent filter has been used successfully in noise reduction for ultrasound images.

Added low pass filter support for 24-bit RGB images.

Image Edit: (top)

Added ability to restrict the Autotrace region growing process by an enclosing object from a loaded object map, providing regional growth control based on specific objects for improved Autotrace functionality.

Morphology: (top)

A new watershed operator has been added, providing an algorithm which will segment connected structures in a binary volume using ultimate erosion and regrowth to determine dividing lines between the objects.

Surface Extractor: (top)

Completely new interface design for surface extraction algorithms and contour extraction, with improved management of controling parameters for each algorithm and easy output file format selection and saving of surfaces.

New Marching Cubes implementation includes the ability to use volume image-based gradients for normal computation for surface polygons (produces much smoother- appearing surfaces).

New output surface format support added for:
PATRAN - to support connectivity to Finite Element Analysis (FEA) applications.
Binary STL - to improve connectivity to rapid prototyping (model building) systems (note that both binary STL and ASCII STL are now supported).

The output of surface coordinate information can now be done in either the discrete voxel coordinates in the volume image, or the coordinates can be scaled using the voxel size information into coordinates in the original measurement space of the image acquisition.

2D Registration: (top)

A mutual information-based, piecewise-linear non-rigid registration algorithm has been implemented for improved 2D slice-to-slice registration capabilities. The algorithm is based on hierarchical subdivision of the image space, with beginning and ending grid sizes as determined by the user, where within each grid section the subregions are registered using standard normalized mutual information techniques. The image is transformed at each hierarchical subdivision level by combining the transformations determined for each grid section as a weighted, linear combination of the transformations in the neighborhood of the point being transformed. The implementation also includes manual control capabilities for the transformation in each grid section, allowing interactive correction or initial condition setting with visual feedback.

3D Registration: (top)

Several improvements have been implemented in the 3-D volume-to-volume normalized mutual information registration capability in Analyze, including:
Isocoordinate base volume registration - the base volume is now automatically interpolated to be anisotropically sampled (cubic) to provide a more stable basis for the resampling of the match volume during the registration process. The Cursor Link tool can also show this base volume in the cubic resampled space.
The fundamental search algorithm for normalized mutual information now permits searching the scale dimensions of the volume, implementing a full nine degree of freedom registration algorithm. Controls for these additional three degrees of freedom can be found in the MatchLab interface.

A new feature has been added to allow the registration of a 3D 'point cloud' to a 3D binary surface using the chamfer distance transform algorithm in Surface Matching. The 'point cloud' can consist of coordinates derived from any source, such as mechanical tracking and coordinate sampling devices, or from landmark positions as sampled directly from other volume images using modules within Analyze.

As with 2D registration, a mutual information-based, piecewise-linear non-rigid registration algorithm has been added to the 3D volume-to-volume registration capabilities of Analyze. This algorithm is based on the hierarchical subdivision of the 3D volume space into successively smaller subvolumes, where the beginning and ending subvolume size are under user control. Each subvolume is registered based on standard normalized mutual information registration techniques. At each hierarchical level, the independent subvolume registration transformations are combined on a specific voxel basis to provide the non-rigid transform for each voxel. These transformations are combined through a weighted, linear combination of the transforms for subvolumes in the neighborhood of the voxel being transformed. The non-rigid registration can be preconditioned by a standard, rigid-body normalized mutual information registration.

Region Of Interest: (top)

A 'Linked Cursor' feature has been added to ROI (left button panel) when both an input volume and a related volume are loaded. This permits simultaneous investigation of both volumes by linked cursor selection of a common point and output of the coordinates of that point and the values from both the input and related volumes.

Added ability to restrict the Autotrace region growing process by an enclosing object from a loaded object map, providing regional growth control based on specific objects for improved Autotrace functionality.

Tree Analysis: (top)

Many important improvements have been made to the Tree Analysis module, most all in response to requests and suggestions by users of this module:
Multiple viewpoints have been added for viewing the extracted tree(s).
Renderings from the original volume image can be combined with tree display.
Using multiple viewpoints, editing tools now allow the interactive manipulation and editing of the tree branching patterns and connectivity associations between trees.
Tree labeling has improved to manage trees of greatly increased complexity.
An extracted, diagramatic representation of the tree branching relationships has been implemented to provide an unambiguous reference for the entire tree.
An image volume can be regenerated by application of the BAP values at each tree node to build up a representation of the tree based on these measured values.

Publish: (top)

Many improvements have been implemented in the Publish module, most with a focus on creation of images for publication (rather than for direct printing). These enhancements include:
Object grouping / ungrouping.
Multiple object alignment tools
Interactive color setting for all objects.
Improved intensity control for embedded images, including independent controls for each image and the ability to copy settings to other images.
Added support for multivolume images, with volume selection for embedded images.
Fonts and graphics now handled internal to Publish (rather than via the external Ghostscript application), providing greatly improved font characters and graphic object representation.
Simplified operation of the Page Setup window, including optimal setup for constructing of output images (rather than printer pages).