|
|
| (256 intermediate revisions by 4 users not shown) |
| Line 1: |
Line 1: |
| − | <html><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"><meta name="GENERATOR" content="Microsoft FrontPage 4.0"><meta name="TEMPLATEBASE" content="mipav_help"><meta name="LASTUPDATED" content="12/04/08 15:56:01"><link rel="StyleSheet" href="document.css" type="text/css" media="screen"><title>Optimized Automatic Registration 3D</title></head><body link="#3366CC" vlink="#9999CC" text="#000000" alink="#0000CC" bgcolor="#FFFFFF">
| + | == [[Preface]] == |
| − | {| width="100%" border="0" cellpadding="0" cellspacing="0"
| + | *[[Scope of this guide]] |
| − | | <img border="0" src="images/mipavfinallogo.gif">
| + | |
| − | | align="right" valign="top" | <a href="NEIRegistrationPlugInBuildMPMaps.html"><img src="images/sprev.gif" border="0" alt="Previous"></a><a href="RegistrationAlighnPatientPosition(DICOM).html"><img src="images/snext.gif" border="0" alt="Next"></a>
| + | |
| − | |}<hr align="left"><blockquote><h2 >
| + | |
| − | Optimized Automatic Registration 3D
| + | |
| − | <a name="wp999048"></a></h2><p >
| + | |
| − | The Optimized Automatic Image Registration method in MIPAV determines a transformation that minimizes a cost function, which represents the quality of alignment between two images. The standard registration problem of computing a transformation that best aligns a reference image to a target image is formulated as a mathematical optimization problem of determining the global minimum of a cost function. The method evaluates the cost function at several different image resolutions, starting with the lowest resolution. Each step of increasing resolution utilizes the previously determined optimal transformation as a starting point and further refines its values. This method usually works well with the images of the same modality.
| + | |
| − | <a name="wp1826176"></a><h3 >
| + | |
| − | Background
| + | |
| − | <a name="wp1846818"></a></h3><p >
| + | |
| − | The main approach for determining an optimal transformation is to calculate a cost function, determine how it changes as individual transformation parameters are varied, and find the parameters that minimize the value of the cost function. For instance, Figure 1 shows a plot of a sample cost function for a selection of transformation parameters; in each plot, a single parameter is varied while all other are kept constant.
| + | |
| − | <a name="wp1846834"></a><p >
| + | |
| − | Another fundamental step in this registration method is to resample and interpolate the reference and target images at several different resolutions (starting with a coarse voxel size and gradually increasing the resolution), while searching for the <em>min</em> cost function.
| + | |
| | | | |
| − | <a name="wp1846835"></a>
| + | == Recent changes in documentation == |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | |
| − | |-
| + | |
| − | | width="9%" valign="top" | <img src="images/tipicon.gif" align="right">
| + | |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" |
| + | |
| − | An advantage of this multi-resolution approach is that the initial optimization considerably reduces computational cost, since the number of sample points is substantially less. In addition, the overall alignment is easier to find at a coarse resolution. Improved accuracy of the transformation paremeters is achieved using the highest resolution image that has not been subsampled.
| + | |
| | | | |
| − | |}
| + | This section is reflects the recent changes in documentation. It is mostly for Olga to help her tracking changes. Please don't pay any special attention to it. |
| − | </br><p >
| + | |
| − | The resampling process may influence the computed value of the cost function; therefore, several different interpolation methods are incorporated with this technique. These methods are discussed in more detail later in<a href="OptimizedAutomaticRegistration3D.html#wp1848350">"Resampling Interpolation" </a>.
| + | |
| − | <a name="wp1846854"></a><p >
| + | |
| − | Taking into account the resampling and interpolation criteria, the Optimized Automatic Registration method begins searching for an optimum transformation using the Powell method at a coarse resolution. Once a transformation has been determined that minimizes the cost function, the transformation is further refined by progressively increasing the resolution.
| + | |
| − | <a name="wp1846859"></a><p >
| + | |
| − | The algorithm offers various resampling interpolation methods combined with the Correlation Ratio, Least Squares, Normalized Cross Correlation, and Normalized Mutual Information as a cost function.
| + | |
| | | | |
| − | <a name="wp1846860"></a>
| + | {{Special:RecentChanges/limit=10, hideminor, hidebots, hideliu}} |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0" | + | |
| − | |-
| + | |
| − | | width="9%" valign="top" | <img src="images/tipicon.gif" align="right">
| + | |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" |
| + | |
| − | See also <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a>.
| + | |
| | | | |
| − | |}
| + | == Getting Started Quickly with MIPAV == |
| − | </br><div >
| + | <ol type="1"><li value="1">[[Introducing MIPAV]] |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | **[[Introducing MIPAV#MipavPlatformIndependence | Platform Independence]] |
| − | |+<div >'''Figure 24. The plots of the Correlation Ratio cost function versus some of the individual parameter values. In each plot (A, B, C, and D), a single parameter is varied while the other are kept constant'''<a name="wp1846875"></a></div>
| + | **[[Introducing MIPAV#MipavCapabilities | Understanding MIPAV capabilities]] |
| − | |-
| + | **[[Introducing MIPAV#JavaPlugIns | Extensibility with Java plug-ins]] |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/CostFunctions1.jpg" border="0" hspace="0" vspace="0"></div>
| + | **[[Introducing MIPAV#MipavFeatures | Sampling of MIPAV's features]] |
| − | <a name="wp1846873"></a></div>
| + | *[[MIPAV mailing list]] |
| − | |}<a name="wp1846857"></a></div><h4 >
| + | *[[Installing mipav#MipavNews|Viewing MIPAV News and Updates]] |
| − | Image types
| + | |
| − | <a name="wp1856604"></a></h4><p >
| + | |
| − | You can apply Optimized Automatic Registration method to color, grayscale and black-and-white 2D and 3D images.
| + | |
| − | <a name="wp1856605"></a><h4 >
| + | |
| − | Outline of the method
| + | |
| − | <a name="wp1856574"></a></h4><h5 >
| + | |
| − | Pre-optimization steps
| + | |
| − | <a name="wp1847085"></a></h5><ol type="1"><li value="1"> The minimum resolution of the reference and target images is determined. For more information, refer to <a href="OptimizedAutomaticRegistration3D.html# wp1847303">"Blurring" </a>.<a name="wp1847111"></a><li value="2">Both images are resampled and interpolated to create high resolution isotropic voxels. The following interpolation methods are available: trilinear (a default), B-spline 3-rd and 4-th order, 4, 5, 6 Lagrangian, and Windowed sinc. Refer to <a href="InterplationMethods.html# wp999048">"Interpolation methods used in MIPAV " </a>.<a name="wp1847112"></a><li value="3">The center of mass (COM) for both images is then computed and one translation step is performed to align the COM. Refer to <a href="OptimizedAutomaticRegistration3D.html# wp1847495">"Calculating the center of mass" </a>.<a name="wp1847114"></a></ol><div > | + | |
| − | <a name="wp1847105"></a></div><h5 >
| + | |
| − | Optimization steps
| + | |
| − | <a name="wp1847211"></a></h5><p >
| + | |
| − | As a part of the transformation optimization process the images are subsampled by 8, 4, and 2 times.
| + | |
| − | <a name="wp1847212"></a><ol type="1"><li value="4"><em>levelEight optimization.</em> Both images are subsampled and interpolated, so that each image is 8 times smaller. The parameters of the transformation are then systematically varied, where the cost function is evaluated for each setting. The parameters corresponding to the smallest value(s) of the cost function are then used as the initial transformation for the next level in the optimization.<a name="wp1847285"></a><li value="5"><em>levelFour optimization</em>. The images are subsampled and interpolated, so that each image is 4 times smaller. And the transformation corresponding to the smallest value(s) of the cost function are determined and used as the initial transformation for the next level in the optimization.<a name="wp1847305"></a><li value="6"><em>levelTwo optimization.</em> Similar processing as the previous two levels except the images are, first, subsampled and interpolated, so that each image is 2 times smaller.<a name="wp1847306"></a><li value="7"><em>levelOne optimization</em> - 1mm resolution images are used in this step and the transformation is generalized to include 12 degrees of freedom.<a name="wp1847307"></a></ol><h5 >
| + | |
| − | Post optimization
| + | |
| − | <a name="wp1847308"></a></h5><ol type="1"><li value="8">Using the same interpolation methods as described in Step 2 and the optimal transformation determined above, the method transforms the reference image into the same coordinate system of the target image.<a name="wp1847309"></a></ol><p >
| + | |
| − | Following Sections 1.2-1.5 describe in detail the algorithms and methods used to perform Steps 1-8 of the optimization routine.
| + | |
| − | <a name="wp1847310"></a><h4 >
| + | |
| − | Pre Optimization
| + | |
| − | <a name="wp1847286"></a></h4><h5 >
| + | |
| | | | |
| − | Blurring
| + | *[[Supported Formats]] |
| − | <a name="wp1847303"></a></h5><p >
| + | **[[Supported Formats | Graphical and file formats supported by MIPAV]] |
| − | The registration algorithm first, determines the minimum resolution for each dimension of the reference and target images. The reference image is blurred to the resolution of the target image if one of the resolutions in a dimension of the target image is 50% or larger than the corresponding resolution in the reference image. Likewise, the target image is blurred to the resolution of the reference image if one of the resolutions in a dimension of the reference image is 50% or larger than the corresponding resolution in the target image. It is possible (but not likely) that both the reference and target images will be blurred.
| + | **[[Image formats descriptions | Image formats' descriptions]] |
| − | <a name="wp1847400"></a>
| + | **[[MIPAV configuration files]] |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | **[[FAQ: Understanding Image Basics#FilesExtensions | Image files supported by MIPAV]] |
| − | |-
| + | |
| − | | width="9%" valign="top" | <img src="images/tipicon.gif" align="right">
| + | |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" |
| + | |
| | | | |
| − | See also <a href="OptimizedAutomaticRegistration3D.html#wp1848350">"Resampling Interpolation" </a>.
| + | <li value="2">[[Installing mipav|Installing MIPAV]] |
| | + | *[[MIPAV system requirements]] |
| | + | *[[Installing mipav|Installing MIPAV]] |
| | + | *[[Installing mipav#MipavUpgrade|Upgrading MIPAV]] |
| | + | *[[Installing mipav#MipavUninstall|Uninstalling MIPAV]] |
| | + | *[[Quitting MIPAV]] |
| | | | |
| − | |}
| + | <li value="3">[[Getting Started Quickly with MIPAV]] |
| − | </br><h5 > | + | *[[Getting Started Quickly with MIPAV#StartingMIPAV | Starting MIPAV]] |
| − | Resampling
| + | *[[Getting Started Quickly with MIPAV#StartingMIPAV | MIPAV main window]] |
| − | <a name="wp1847453"></a></h5><p >
| + | *[[Getting Started Quickly with MIPAV#ManagingMemoryResources | Managing memory in MIPAV]] |
| − | Once this isotropic 1mm resolution images have been obtained, the subsampled versions are also created. The sub-sampling algorithm then simply keeps every n-th point (that is, 2, 4 or 8) on the lattice in each direction. Therefore, the new volume contains 1/n3 as many points as the original.
| + | *[[Allocating Memory in MIPAV]] |
| | + | *[[Opening and loading image files]] |
| | + | *[[MIPAV toolbars]] |
| | + | *[[Displaying MIPAV Toolbars]] |
| | + | *[[Displaying images]] |
| | + | **[[Displaying images#DefaultView| Default View]] |
| | + | **[[Displaying images#LightboxView| Lightbox View]] |
| | + | **[[Displaying images#CineView| Cine View]] |
| | + | **[[Displaying images#AdjustingMagnification| Adjusting Magnification]] |
| | + | **[[Displaying images#MagnifyingRegions| Magnifying Regions within Images]] |
| | + | **[[Displaying images#ImprovingContrast| Improving Image Contrast]] |
| | + | *[[Modifying image resolutions]] |
| | | | |
| − | <a name="wp1847469"></a>
| + | *[[Changing Image Contrast]] |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | **[[Changing Image Contrast|Adjust Window and Level]] |
| − | |- | + | **[[Changing Image Contrast|Improving contrast by generating and modifying histograms]] |
| − | | width="9%" valign="top" | <img src="images/noteicon.gif" align="right"> | + | **[[Changing Image Contrast|Using Lookup Table dialog box]] |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" |
| + | **[[Changing Image Contrast#PredefinedLuts|Applying color to images using predefined LUTs]] |
| − | Resampling can be performed for x, y, and z axes for 3D images, or for only x and y for 2D images.
| + | **[[Changing Image Contrast|Comparing images using alphablending]] |
| | + | **[[Changing Image Contrast|Restoring images to their original appearance]] |
| | | | |
| − | |} | + | *[[Creating new images]] |
| − | </br><h4 >
| + | **[[Creating new images| Image dialog box]] |
| − | Calculating the center of mass
| + | **[[Creating new images| Annotating images with text]] |
| − | <a name="wp1847495"></a></h4><p >
| + | |
| − | To calculate the center of mass (COM), the method uses the right hand convention in 3D coordinate systems (X, Y, Z). Thus, in the image space, the left hand corner of the image is set to (0,0,0). The x axis goes left to right, y axis goes top to bottom and z axis goes into the screen.
| + | |
| − | <a name="wp1847496"></a><ol type="1"><li value="1">To calculate the COM, we, first, define the characteristics function of an object in an image to be as described in <a href="OptimizedAutomaticRegistration3D.html#wp1857433">equation 7</a>:<a name="wp1857438"></a></ol><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 5pt; margin-left: 0pt; margin-right: 0pt; margin-top: 9pt; text-align: center; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | Equation 7
| + | |
| | | | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | *[[Saving and printing images]] |
| − | |+
| + | |
| − | |-
| + | |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/COMbPoints.jpg" border="0" hspace="0" vspace="0"></div>
| + | |
| − | <a name="wp1857449"></a></div>
| + | |
| − | |}
| + | |
| − | <a name="wp1857433"></a></div><ol type="1"><li value="2">Then, an area of an image can be calculated as <a name="wp1847516"></a></ol><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 5pt; margin-left: 0pt; margin-right: 0pt; margin-top: 9pt; text-align: center; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | Equation 8
| + | |
| − | <a name="wp1857499"></a></div><div >
| + | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | |
| − | |+
| + | |
| − | |-
| + | |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/COMArea.jpg" border="0" hspace="0" vspace="0"></div>
| + | |
| − | <a name="wp1857522"></a></div>
| + | |
| − | |}<a name="wp1857500"></a></div><div >
| + | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | |
| − | |+<div >'''Figure 25. Calculating the center of mass (COM)'''<a name="wp1857642"></a></div>
| + | |
| − | |-
| + | |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/COM_Calculating.jpg" border="0" hspace="0" vspace="0"></div>
| + | |
| − | <a name="wp1857640"></a></div>
| + | |
| − | |}<a name="wp1857512"></a></div><ol type="1"><li value="3">And the center of mass, denoted by (x<em>COM</em>, y<em>COM</em>)<a href="#wp1857627"><sup>1</sup></a> is given by the 1-st moments of the object: <a name="wp1857635"></a></ol><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 5pt; margin-left: 0pt; margin-right: 0pt; margin-top: 9pt; text-align: center; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| | | | |
| − | Equation 9
| + | *[[Delineating volumes of interest (VOIs)| Volumes of Interest]] |
| | + | **[[Delineating volumes of interest (VOIs)]] |
| | + | **[[Calculating VOI statistics]] |
| | + | **[[Calculating VOI statistics#VoiStatistics| VOI statistics dialog box]] |
| | + | **[[Reviewing VOI statistics]] |
| | + | **[[Calculating statistics on VOI groups]] |
| | + | **[[Reviewing VOI statistics|Generating graphs (intensity profiles)]] |
| | | | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | *[[ImageJ | Working with ImageJ]] |
| − | |+
| + | *[[Customizing MIPAV]] |
| − | |- | + | **[[Customizing MIPAV#MipavOptions | MIPAV Options dialog box]] |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/COM_X.jpg" border="0" hspace="0" vspace="0"></div>
| + | **[[Customizing MIPAV#SplashScreen | Showing and hiding the splash screen on start-up]] |
| − | <a name="wp1857569"></a></div>
| + | **[[Customizing MIPAV#DebuggingMipav | Debugging MIPAV]] |
| − | |}
| + | **[[Customizing MIPAV#ActionsHistory | Saving a history of actions on images]] |
| − | <a name="wp1857563"></a></div><div >
| + | **[[Customizing MIPAV#DefaultFileTypes | Choosing the default file types to display or save]] |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | **[[Customizing MIPAV#AddingShortcuts | Adding Shortcuts]] |
| − | |+
| + | **[[MIPAV configuration files]] |
| − | |- | + | |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/COM_Y.jpg" border="0" hspace="0" vspace="0"></div>
| + | |
| − | <a name="wp1857594"></a></div>
| + | |
| − | |}<a name="wp1857494"></a></div><h4 > | + | |
| − | Process
| + | |
| − | <a name="wp1857581"></a></h4><p >
| + | |
| − | The process used in this method performs the following:
| + | |
| − | <a name="wp1847517"></a><ul><ul><li >Calculates the COM for both images as described above in <a href="OptimizedAutomaticRegistration3D.html#wp1847495">"Calculating the center of mass" </a>; <a name="wp1847518"></a><li >Aligns the COM of both images, which is an initial transfomation that involves only translation;<a name="wp1847519"></a><li >For each resampled image (which is 8, 4, 2, and 1 times), determines the transform that minimizes the cost function.<a name="wp1847520"></a></ul></ul><h5 >
| + | |
| − | levelEight optimization
| + | |
| − | <a name="wp1847579"></a></h5><p >
| + | |
| − | levelEight optimization is primarily concerned with determining the rotational transformation that will lead to the globally optimal transformation, which corresponds to the global minimum of the cost function. levelEight optimization uses the lowest resolution images and coarse rotation angles evaluated at relatively large step sizes.
| + | |
| − | <a name="wp1847580"></a><h5 >
| + | |
| | | | |
| − | Optimization
| + | *[[Developing new tools using the API]] |
| − | <a name="wp1847620"></a></h5><p >
| + | |
| − | This Optimized Automatic Registration method is based on the empirical observation that finding the correct orientation, or rotation, is the most difficult task in image registration, since the three rotation parameters are highly coupled and most of the erroneous registrations that have been examined have happened primarily due to an incorrect orientation. Therefore, the search concentrates on the rotational part of the transformation.
| + | |
| − | <a name="wp1847621"></a><p >
| + | |
| − | Specifically, the reference image is oriented, interpolated, and the cost function evaluated for coarse rotations angles (-30, -30, -30), (-30, -30, -15), ..., (-30, -30, 30), where the values represent the amount of rotation in degrees about the x, y, and z axes respectively. Since there are three rotation angles and each angle can contain five different values, there are 125 possible angle configurations.
| + | |
| − | <a name="wp1847622"></a><p >
| + | |
| − | For each angle configuration, a 4-DOF local optimization is also performed to find the optimal translation and (global) scale. By default, the initial values are set to
| + | |
| − | <a name="wp1847623"></a><ul><ul><li >0 for translation<a name="wp1847624"></a><li >1 for scale. <a name="wp1847625"></a></ul></ul><p >
| + | |
| − | The best 20% of the cost values and/or the corresponding angle configurations (candidate local minima) are stored in a vector of minima that is used as a starting point for a further optimization, which uses a smaller step size over a narrower range of angles. Refer to <a href="OptimizedAutomaticRegistration3D.html#wp1848160">Figure 26</a>.
| + | |
| − | <a name="wp1847626"></a><p >
| + | |
| − | For each parameter setting corresponding to the top 20% of the cost function minima, the algorithm performs rotations over the fine grid (which is by default 15 degrees with 6 degrees step) and evaluates the cost function.
| + | |
| − | <a name="wp1848059"></a><p >
| + | |
| − | The method now finds all angle configurations that have the corresponding cost function values lower than their neighbors from the vector of candidate local minima.
| + | |
| | | | |
| − | <a name="wp1848060"></a><p >
| + | *[[Technical Support | Contacting MIPAV support]] |
| − | For each of these sets of parameters a 7-DOF optimization is then performed, storing the results of the transformation and costs <em>before</em> and <em>after optimization</em> in a vector of minima. See also <a href="OptimizedAutomaticRegistration3D.html#wp1858186">"Degrees of freedom" </a>.
| + | *[[MIPAV mailing list | Joining MIPAV mailing list]] |
| − | <a name="wp1859274"></a><p >
| + | |
| − | Since the relative costs of each candidate solution may change at higher resolutions, where structures become more sharply defined, the top 20% of the cost function minima points are considered for the next higher resolution (4) stage, rather than just a single best solution.
| + | |
| | | | |
| − | <a name="wp1848062"></a><p >
| + | == MIPAV Basics == |
| − | By default, the algorithm uses Trilinear interpolation to transform images to this new orientation.
| + | <li value="4">[[Understanding Image Basics]] |
| − | <a name="wp1848150"></a><div >
| + | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | |
| − | |+<div >'''Figure 26. levelEight optimization'''<a name="wp1848160"></a></div>
| + | |
| − | |-
| + | |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/OARLev8Scheme.jpg" border="0" hspace="0" vspace="0"></div>
| + | |
| − | <a name="wp1848157"></a></div>
| + | |
| − | |}<a name="wp1848161"></a></div><h5 >
| + | |
| − | levelFour optimization
| + | |
| − | <a name="wp1848179"></a></h5><p >
| + | |
| − | The levelFour optimization step uses the interpolated images subsampled by 4 to determine the transformation that minimizes the cost function starting with the transformations determined in levelEight.
| + | |
| | | | |
| − | <a name="wp1848180"></a><p > | + | <li value="5">[[Working with DICOM Images]] |
| − | The optimization determines a 7-DOF transformation that corresponds to the minimum value of the cost function. This transformation is then perturbed and the cost function is computed for these new settings. The perturbations correspond to six-degrees for each rotation parameter and a global scaling factor of 0.8, 0.9, 1.1, and 1.2. A type of transformation (translation and/or global scale) is specified by a user. It includes the following transformations: Affine-12 (a default), Rigid-6, Global Rescale-7, and Specific Rescale-9. Here, numbers indicate degree of freedom (DOF).
| + | *[[Browsing DICOM images|Browsing DICOM images]] |
| − | <a name="wp1848181"></a><p >
| + | *[[Sending and retrieving DICOM images|Sending and retrieving DICOM images ]] |
| − | A vector of parameters and top 20% of the <em>min</em> cost function values are considered for the next step, which involves images subsampled by 2.
| + | *[[Testing the connection|Testing the connection ]] |
| − | <a name="wp1856275"></a>
| + | *[[Posing queries and retrieving images|Posing queries and retrieving images]] |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | *[[Receiving and sending image files|Receiving and sending image files]] |
| − | |- | + | *[[Displaying and editing DICOM tag information|Displaying and editing DICOM tag information]] |
| − | | width="9%" valign="top" | <img src="images/tipicon.gif" align="right"> | + | *[[Protecting patient privacy using Anonymize|Protecting patient privacy using Anonymize]] |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" | | + | *[[Converting non-DICOM image files to DICOM format|Converting non-DICOM image files to DICOM format]] |
| − | See also <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms" </a> and <a href="OptimizedAutomaticRegistration3D.html#wp1858186">"Degrees of freedom" </a>
| + | |
| − | |} | + | |
| − | </br><h5 >
| + | |
| − | levelTwo
| + | |
| − | <a name="wp1856280"></a></h5><p >
| + | |
| − | In levelTwo, the method uses the images subsampled and interpolated by 2. The cost function is evaluated using the transformation parameters corresponding to the top 20% minima obtained from the levelFour. It finds the best minimum, and then optimizes it with the 7-DOF and then 9-DOF transformation. The method then returns the best minimum after optimization.
| + | |
| − | <a name="wp1848253"></a>
| + | |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | |
| − | |- | + | |
| − | | width="9%" valign="top" | <img src="images/noteicon.gif" align="right">
| + | |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" | <em>Note:</em> if a user has limited the degrees of freedom to six, as for the rigid transformation, there will only be one optimization run, with six degrees of freedom.
| + | |
| | | | |
| − | |}
| + | <li value="6">[[Visualizing Images]] |
| − | </br><h5 > | + | *[[Visualizing Images: Displaying images using the default view |Displaying images using the default view]] |
| | + | **[[Visualizing Images:Displaying images using the default view: Magnifying and minifying images|Magnifying and minifying images]] |
| | + | **[[Visualizing Images: Displaying images using the default view: Viewing two images together|Viewing two images together]] |
| | + | **[[Visualizing Images: Displaying images using the default view:Changing image brightness and contrast using LUTs|Changing image brightness and contrast using LUTs]] |
| | + | *[[Visualizing Images: Displaying images using the animate view |Displaying images using the animate view]] |
| | + | *[[Visualizing Images: Displaying images using the cine (movie) view|Displaying images using the cine (movie) view]] |
| | + | *[[Visualizing Images: Displaying images using the lightbox view |Displaying images using the lightbox view]] |
| | + | *[[Visualizing Images: Displaying images using the link to another image view|Displaying images using the link to another image view]] |
| | + | *[[Visualizing Images: Displaying images using the triplanar view|Displaying images using the triplanar view]] |
| | + | *[[Visualizing Images: Displaying images using the triplanar-dual view|Displaying images using the triplanar-dual view]] |
| | + | *[[Visualizing Images: Displaying images using the surface plotter view|Displaying images using the surface plotter view]] |
| | + | *[[Visualizing Images: Displaying images using the surface renderer view|Displaying images using the surface renderer view]] |
| | + | *[[Visualizing Images: Displaying images using the surface renderer view|Displaying images using the volume shear view]] |
| | + | *[[Visualizing Images: Displaying images using the surface renderer view|Displaying images using the volume renderer view]] |
| | | | |
| − | levelOne
| + | <li value="7">[[Segmenting Images Using Contours and Masks]] |
| − | <a name="wp1848308"></a></h5><p > | + | *[[Segmenting Images Using Contours and Masks: Using contours to segment a VOI|Using contours to segment a VOI]] |
| − | This step uses the unsubsampled interpolated images and computes the value of the cost function for each parameter setting determined in the previous levelTwo optimization. The parameters corresponding to the minimum value of the cost function with this resolution are then further optimized for transformations with 7-, 9-, and 12- DOF, unless a rigid (6-DOF) transformation is desired.
| + | **[[Modifying Contours]] |
| − | <a name="wp1848309"></a><h5 >
| + | *[[Segmenting Images Using Contours and Masks: Generating Masks|Generating masks]] |
| − | What is saved
| + | **[[Creating a mask using the Paint Grow Segmentation method]] |
| − | <a name="wp1848310"></a></h5><p >
| + | *[[Segmenting Images Using Contours and Masks:Converting VOI contours to masks|Converting VOI contours to masks]] |
| − | The method stores
| + | *[[Segmenting Images Using Contours and Masks: Converting masks to VOI contours and paints |Converting masks to VOI contours and paints]] |
| − | <a name="wp1848311"></a><ul><ul><li >The set of parameters or vector at which the minimum was reached; <a name="wp1848312"></a><li >The value of the cost function at that minimum; <a name="wp1848313"></a><li >And the matrix, which was the true input into the cost function and represents the transformation that gives the minimum cost of differences between the images.<a name="wp1848314"></a></ul></ul><h5 >
| + | *[[Segmenting Images Using Contours and Masks: Advanced paint and Power Paint tools|Advanced Paint and Power Paint tools]] |
| − | Additional registration parameters
| + | |
| − | <a name="wp1848315"></a></h5><p >
| + | |
| − | To better align the images, one can also include into calculation additional registration parameters, such as weighting coefficients and advanced parameters used for calculating the min cost function. For more information, refer to <a href="OptimizedAutomaticRegistration3D.html#wp1826716">"Applying the AlgorithmOAR 3D" </a>.
| + | |
| | | | |
| − | <a name="wp1848316"></a><h4 > | + | <li value="8">[[Analyzing Images]] |
| − | Resampling Interpolation
| + | *[[Analyzing Images|Calculating statistics for contoured VOIs ]] |
| − | <a name="wp1848350"></a></h4><p >
| + | *[[Calculating statistics on VOI groups|Calculating statistics on VOI groups]] |
| − | The Optimized Automatic Registration method resampes images using an interpolation scheme, where anisotropic voxels in the Z direction are resampled into isotropic 1mm cubic voxels. New voxel values are computed using a weighted combination of existing voxel values within a defined neighborhood of the new voxel location. The interpolation methods are available in this implementation of the registration technique, include Trilinear, B-spline 3-rd order, B-spline 4-th order, Cubic Lagrangian, Quintic Lagrangian, Heptic Lagrangian, and Windowed sinc.
| + | *[[Calculating the volume of masks|Calculating the volume of masks ]] |
| − | <a name="wp1848351"></a><p >
| + | *[[Generating graphs|Generating graphs]] |
| − | For mor information about the interpolation methods used in Optimized Automatic Registration, refer to <a href="InterplationMethods.html#wp999048">"Interpolation methods used in MIPAV"</a>.
| + | **[[Customizing the appearance of graphs - Modify graph dialog box]] |
| − | <a name="wp1913796"></a><h3 >
| + | **[[Changing the legends for functions]] |
| | | | |
| − | Cost Functions
| + | <li value="9">[[Changing Image Datasets Using MIPAV Utilities]] |
| − | <a name="wp1851326"></a></h3><p > | + | *[[Standard tasks provided through commands on the Utilities menu]] |
| − | A <em>similarity</em> or <em>cost function</em> measures the similarity between two images. During the Optimized Automatic Registration the adjusted image V is transformed using the transformation functions described above. And the similarity S(U;V<em>t</em>) between the reference image U and transformed image V<em>t</em> is then calculated. The Optimized Automatic Registration method searches for the transformation that gives <em>the smallest value of the cost function</em>, which we assume is the transformation that also gives the best alignment.
| + | *[[Recording utilities usage with the history feature]] |
| | + | *[[4 D tools]] |
| | + | *[[Adding image margins]] |
| | + | *[[Copying images using the Clone command]] |
| | + | *[[Converting image datasets to different data types]] |
| | + | *[[Correcting image spacing]] |
| | + | *[[Cropping images]] |
| | + | *[[Masking (filling) images]] |
| | + | *[[Flipping images]] |
| | + | *[[Image Calculator]] |
| | + | **[[Image Calculator: OR]] |
| | + | **[[Image Calculator: Advanced image calculator options]] |
| | + | *[[Image Math]] |
| | + | *[[Inserting slices into image datasets]] |
| | + | *[[Inverting the image]] |
| | + | *[[Matching images]] |
| | + | *[[Maximum Intensity Projection]] |
| | + | *[[Adding noise to images]] |
| | + | *[[Pad]] |
| | + | *[[Quantify Mask]] |
| | + | *[[Replacing pixel/voxel value in images]] |
| | + | *[[Rotating images]] |
| | + | *[[Slice tools]] |
| | + | *[[Inserting slices into image datasets]] |
| | + | *[[Subtract VOI Background]] |
| | + | *[[Standard tasks provided through commands on the Utilities menu]] |
| | | | |
| − | <a name="wp1852019"></a><p >
| + | <li value="10">[[Using Scripts (Macros) in MIPAV]] |
| − | The cost functions which are implemented in this method include:
| + | *[[Developing and using scripts]] |
| − | <a name="wp1852578"></a><ul><ul><li >Correlation ratio, refer to <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a> , <a href="CostFunctions.html#wp1852618">"Correlation ratio" </a><a name="wp1852590"></a><li >Least Squares, refer to <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a>, <a href="CostFunctions.html#wp1853027">"Least Squares" </a><a name="wp1913902"></a><li >Normalized Cross Correlation, see <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a>, <a href="CostFunctions.html#wp1853462">"Normalized cross correlation" </a><a name="wp1913908"></a><li >Normalized Mutual Information, see <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a>, <a href="CostFunctions.html#wp1853757">"Normalized mutual information" </a>.<a name="wp1913913"></a></ul></ul><h4 >
| + | **[[Setting up scripting]] |
| − | Powell algorithm
| + | **[[Planning scripts]] |
| − | <a name="wp1857913"></a></h4><p ><a href="http://math.fullerton.edu/mathews/n2003/PowellMethodMod.html" target="_window">The Optimized Automatic Registration method uses the Powell algorithm to find the global minimum of the chosen cost function. For more information about the Powell algorithm, refer to the following link </a><u>http://math.fullerton.edu/mathews/n2003/PowellMethodMod.html</u><a name="wp1857945"></a><h4 >
| + | **[[Recording scripts]] |
| − | Degrees of freedom
| + | **[[Running scripts]] |
| − | <a name="wp1858186"></a></h4><p >
| + | **[[Editing and deleting scripts]] |
| − | The number of independent pieces of information that go into the estimate of a parameter is called the degrees of freedom (DOF).
| + | **[[Examples]] |
| − | <a name="wp1914083"></a><p >
| + | *[[Combining scripts and other programs]] |
| − | For more information refer to <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms"</a>, <a href="CostFunctions.html#wp1858186">"Degrees of freedom" </a>.
| + | **[[Using the mipav command]] |
| − | <a name="wp1914087"></a><h4 >
| + | **[[Using Shell scripting to lessen typing]] |
| − | References
| + | |
| − | <a name="wp1914088"></a></h4><p >
| + | |
| − | Bjórn Hamre "Three-dimensional image registration of magnetic resonance (MRI) head volumes" Section for Medical Image Analysis and Informatics Department of Physiology & Department of Informatics University of Bergen, Norway.
| + | |
| | | | |
| − | <a name="wp1856361"></a><p ><a href="http://local.wasp.uwa.edu.au/ ~pbourke/other/interpolation/index.html" target="_window">Bourke Paul "Interpolation methods" http://local.wasp.uwa.edu.au/ ~pbourke/other/interpolation/index.html</a><a name="wp1856362"></a><p ><a href="http://www.cl.cam.ac.uk/teaching/1999/AGraphHCI/SMAG/ node4.html" target="_window">B-splines: http://www.cl.cam.ac.uk/teaching/1999/AGraphHCI/SMAG/ node4.html</a><a name="wp1856363"></a><p > | + | <li value="11">[[Developing Plugin Programs]] |
| − | Chapter 3: "Registration Methodology: Concepts and Algorithms" by Derek L.G.Hill and Philipe Batchelor in <em>Medical Image Registration</em> edited by Joseph V. Hajnal, Derek L.G.Hill, and David J. Hawkes,CRC Press, 2001, pp. 40-70.
| + | *[[Understanding plug-in programs]] |
| | + | *[[Using the API documentation]] |
| | + | *[[Developing plug-in programs]] |
| | + | *[[Creating a self-contained plug-in frame]] |
| | + | *[[Examples of MIPAV plug-ins]] |
| | + | **[[Plug-in CT_MD, a typical plug-in program]] |
| | + | ***[[PlugInAlgorithmCT_MD.java]] |
| | + | ***[[Plug-in CT_MD.java]] |
| | | | |
| − | <a name="wp1856364"></a><p > | + | <li value="12">[[Technical Support]]</ol> |
| − | Chapter 33 "Within-Modality Registration Using Intensity-Based Cost Functions" by Roger P. Woods in <em>Handbook of Medical Image Processing and Analysis</em>, Editor, Isaac N. Bankman, Academic Press, 2000, pp. 529-553.
| + | |
| − | <a name="wp1856365"></a><p ><a href="http://www.fmrib.ox.ac.uk/fsl/flirt/" target="_window">FLIRT, visit their homepage at http://www.fmrib.ox.ac.uk/fsl/flirt/</a><a name="wp1856366"></a><p >
| + | |
| | | | |
| − | Jenkinson, M. and Smith, S. (2001a) "A global optimisation method for robust affine registration of brain images". Medical Image Analysis, 5(2):143-156.
| + | == MIPAV Algorithms == |
| − | <a name="wp1856367"></a><p ><a href="http://www.fmrib.ox.ac.uk/fsl/flirt" target="_window">Jenkinson Mark and Stephen Smith "Optimisation in Robust Linear Registration of Brain Images" FMRIB Technical Report TR00MJ2. http://www.fmrib.ox.ac.uk/fsl/flirt </a><a name="wp1856368"></a><p > | + | *[[Understanding MIPAV capabilities]] |
| − | Josien P. W. Pluim, J. B. Antoine Maintz and Max A. Viergever. <em>Mutual information based registration of medical images: a survey</em>. IEEE TRANSACTIONS ON MEDICAL IMAGING, VOL. XX, NO. Y, MONTH 2003.
| + | *[[Developing new tools using the API]] |
| | + | *[[Using MIPAV Algorithms | Overview of MIPAV Algorithms]] |
| | + | **[[Cost functions used in MIPAV algorithms]] |
| | + | **[[Interpolation methods used in MIPAV]] |
| | + | **[[Degrees of freedom]] |
| | + | <div id="AutoCorrelation"></div> |
| | + | **[[Autocorrelation Coefficients]] |
| | + | **[[Autocovariance Coefficients]] |
| | + | **[[Barrel Distortion Correction]] |
| | + | <div id="DtiAlgorithms"></div> |
| | + | **[[DTI Color Display]] |
| | + | **[[DTI Estimate tensor | DTI Estimate tensor - under construction]] |
| | + | **[[DTI Pipeline]] |
| | + | <div id="EdgeDetection"></div> |
| | + | **[[Edge Detection: Zero X Laplacian]] |
| | + | **[[Edge Detection: Zero X Non-Maximum Suppression]] |
| | + | <div id="ExtractSurface"></div> |
| | + | **[[Extract Brain: Extract Brain Surface (BET)]] |
| | + | **[[Extract Brain: Extract Brain Surface (BSE)]] |
| | + | **[[Extract Surface (Marching Cubes)]] |
| | + | **[[Face Anonymizer (BET)]] |
| | + | <div id="FilterAlgorithms"></div> |
| | + | **[[Fast Fourier Transformation (FFT)]] |
| | + | **[[Filters (Spatial): Adaptive Noise Reduction]] |
| | + | **[[Filters (Frequency)]] |
| | + | **[[Filters (Spatial): Adaptive Path Smooth]] |
| | + | **[[Filters (Spatial) Anisotropic Diffusion]] |
| | + | **[[Filters (Spatial): Coherence-Enhancing Diffusion]] |
| | + | **[[Filters (Spatial): Gaussian Blur]] |
| | + | **[[Filters (Spatial): Gradient Magnitude]] |
| | + | **[[Filters (Spatial): Haralick Texture]] |
| | + | **[[Filters (Spatial) Laplacian]] |
| | + | **[[Filters (Spatial): Local Normalization]] |
| | + | **[[Filters (Spatial): Mean]] |
| | + | **[[Filters (Spatial): Median]] |
| | + | **[[Filters (Spatial): Mode]] |
| | + | **[[Filters (Spatial): Nonlinear Noise Reduction]] |
| | + | **[[Filters (Spatial): Nonmaximum Suppression]] |
| | + | **[[Filters (Spatial): Regularized Isotropic (Nonlinear) Diffusion]] |
| | + | **[[Filters (Spatial): Slice Averaging]] |
| | + | **[[Filters (Spatial): Unsharp Mask]] |
| | + | **[[Filters (Wavelet): De-noising BLS GSM]] |
| | + | **[[Filters (Wavelet): Thresholding]] |
| | + | **[[Fuzzy C-Means: Multispectral and Single Channel Algorithms]] |
| | + | <div id="HistogramAlgorithms"></div> |
| | + | **[[Histogram Equalization: Regional Adaptive]] |
| | + | **[[Histogram Equalization: Neighborhood Adaptive]] |
| | + | **[[Histogram Matching]] |
| | + | **[[2D Histogram]] |
| | + | **[[Cumulative Histogram]] |
| | + | **[[Histogram summary]] |
| | + | **[[Image Calculator]] |
| | + | **[[Manual 2D Series]] |
| | + | **[[Mask]] |
| | + | <div id="MicroscopyAlgorithms"></div> |
| | + | **[[Microscopy Colocalization Orthogonal Regression]] |
| | + | **[[Microscopy: Fluorescence Resonance Energy Transfer (FRET)-Acceptor]] |
| | + | **[[Microscopy: FRAP (Fluorescence Recovery After Photobleaching)]] |
| | + | **[[Microscopy: Fluorescent Resonance Energy Transfer (FRET) Bleed Through and Efficiency]] |
| | + | **[[Microscopy: Blind Deconvolution]] |
| | + | <div id="TransformationAlgorithms"></div> |
| | + | **[[Midsagittal line alignment]] |
| | + | **[[Morphology]] |
| | + | **[[Reslice-Isotropic Voxels]] |
| | + | **[[Shading Correction: Inhomogeneity N3 Correction]] |
| | + | **[[Standard Deviation Threshold]] |
| | + | **[[Subsampling images]] |
| | + | **[[Threshold]] |
| | + | **[[Transform to power of 2]] |
| | + | <div id="RegistrationAlgorithms"></div> |
| | + | **[[B-Spline Automatic Registration | Registration: B-Spline Automatic Registration ]] |
| | + | ***[[Detect folding]] |
| | + | ***[[User Dialogs in MIPAV]] |
| | + | **[[Optimized automatic registration 3D | Registration: Optimized automatic registration 3D and 3.5D]] |
| | + | **[[Registration: Landmark-Least Squares]] |
| | + | **[[Registration: Landmark-TPSpline]] |
| | + | **[[Registration: Manual 2D Series]] |
| | + | **[[Midsagittal line alignment | Registration: Midsagittal line alignment]] |
| | + | **[[Reslice-Isotropic Voxels | Registration: Reslice-Isotropic Voxels]] |
| | + | **[[Registration: Time Series Optimized Automatic Registration]] |
| | + | **[[Volume Renderer]] |
| | | | |
| − | <a name="wp1856370"></a><p ><a href="http://math.fullerton.edu/ mathews/n2003/PowellMethodMod.html" target="_window">Powell method to find the global minimum: http://math.fullerton.edu/ mathews/n2003/PowellMethodMod.html</a><a name="wp1856371"></a><p >
| + | == Glossary == |
| − | Stephen M. Smith , BET: "Brain Extraction Tool", FMRIB technical Report TR00SMS2b, FMRIB (Oxford Centre for Functional Magnetic Resonance Imaging of the Brain), Department of Clinical Neurology, Oxford University, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK.
| + | [[Glossary for MIPAV Help]] |
| − | <a name="wp1856372"></a><p >
| + | |
| − | Thomas M. Lehmann,* Member, IEEE, Claudia Gonner, and Klaus Spitzer. <em>Survey: Interpolation Methods in Medical Image Processing</em>. IEEE TRANSACTIONS ON MEDICAL IMAGING, VOL. 18, NO. 11, NOVEMBER 1999 1049.
| + | |
| | | | |
| − | <a name="wp1856373"></a><h3 >
| + | == Appendices == |
| − | Applying the AlgorithmOAR 3D
| + | <ol type="A"><li value="1">[[References]] |
| − | <a name="wp1826716"></a></h3><h5 >
| + | <li value="B">[[DICOM Conformance]] |
| − | To run the algorithm, complete the following steps:
| + | <li value="C">[[Supported Formats]] |
| − | <a name="wp1781625"></a></h5><ol type=" 1"><li value="1"> Open an input and reference image.<a name="wp1855835"></a><li value=" 2"> Call Algorithms > Registration > Optimized Automatic Registration.<a name="wp1855836"></a><li value="3">The Optimized Automatic Registration dialog box appears.<a name="wp1855837"></a></ol><p > | + | *[[Other formats supported by MIPAV ]] |
| − | In the dialog box,
| + | *[[Understanding MIPAV-related files]] |
| − | <a name="wp1856074"></a><ul><li >Select the image in the Register to box;<a name="wp1855839"></a><li >Complete the rest of the dialog box options. Refer to <a href="OptimizedAutomaticRegistration3D.html#wp1915255">"" </a>.<a name="wp1855840"></a><li value="4">Press OK.<a name="wp1855841"></a></ul><p >
| + | |
| − | The algorithm begins to run and the Registering Images window appears with the status. When the algorithm stops running the registered image appears in a new image frame. Refer to <a href="OptimizedAutomaticRegistration3D.html#wp1855969">Figure 27</a>.
| + | |
| − | <a name="wp1855842"></a>
| + | |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| + | |
| − | |-
| + | |
| − | | width="9%" valign="top" | <img src="images/recommendationicon.gif" align="right">
| + | |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" | '''Recommendation:''' First, use the default algorithm settings. And then, after completing the first registration run, modify them depending on the result you receive.
| + | |
| | | | |
| − | |}
| |
| − | </br><div >
| |
| − | {| border="1" cellpadding="5" cellspacing="0"
| |
| − | |+<div >'''Figure 27. The input image (A), reference image (B) and result image (C). Registration was performed using the default algorithm settings.'''<a name="wp1855969"></a></div>
| |
| − | |-
| |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: right; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/OARInputImage.jpg" border="0" hspace="0" vspace="0"></div>
| |
| − | <a name="wp1855977"></a></div>
| |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: right; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/OARReferenceImage.jpg" border="0" hspace="0" vspace="0"></div>
| |
| − | <a name="wp1855979"></a></div>
| |
| − | | <div style="color: #000000; font-size: 2pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 0pt; margin-right: 0pt; margin-top: 1pt; text-align: right; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"><div align="center"><img src="images/OARResultImage.jpg" border="0" hspace="0" vspace="0"></div>
| |
| − | <a name="wp1855981"></a></div>
| |
| − | |-
| |
| − | | <div >A<a name="wp1855963"></a></div>
| |
| − | | <div >B<a name="wp1855965"></a></div>
| |
| − | | <div >C<a name="wp1855967"></a></div>
| |
| − | |}<a name="wp1854408"></a></div><div >
| |
| − | <a name="wp1915252"></a></div><div >
| |
| − | <a name="wp1915255"></a></div><h3 >
| |
| − | Applying the Constrained AlgorithmOAR 3D
| |
| − | <a name="wp1915257"></a></h3><h5 >
| |
| − | To run the algorithm, complete the following steps:
| |
| − | <a name="wp1915061"></a></h5><ol type="1"><li value="1">Open an input and reference image.<a name="wp1915062"></a><li value="2">Call Algorithms > Registration > Constrained Optimized Automatic Registration.<a name="wp1915063"></a><li value="3">The Optimized Automatic Registration dialog box appears.<a name="wp1915064"></a></ol><p >
| |
| − | In the dialog box,
| |
| − | <a name="wp1915065"></a><ul><li >Select the image in the Register to box;<a name="wp1915066"></a><li >Complete the rest of the dialog box options. See <a href="OptimizedAutomaticRegistration3D.html#wp1915846">Figure 28</a>;<a name="wp1915069"></a><li >Press the Advanced Settngs button, and then complete the Advanced OAR settings dialog box. In this dialog box, you can limit the number of iterations and also the translation's range. Refer to <a href="OptimizedAutomaticRegistration3D.html#wp1914757">Figure 29</a> for the dialog box options.<a name="wp1915090"></a><li value="4">Press OK.<a name="wp1915071"></a></ul><p >
| |
| − | The algorithm begins to run and the Registering Images window appears with the status. When the algorithm stops running the registered image appears in a new image frame.
| |
| − | <a name="wp1915072"></a>
| |
| − | {| border="1" frame="hsides" width="90%" border="0" cellspacing="0"
| |
| − | |-
| |
| − | | width="9%" valign="top" | <img src="images/recommendationicon.gif" align="right">
| |
| − | | width="81%" bgcolor="#B0E0E6" font color="#000000" | '''Recommendation:''' For the Constrained Optimized Automatic Registration algorithm, first, use the default algorithm settings. And then, after completing the first registration run, modify them (including Advanced settings) depending on the result you receive.
| |
| | | | |
| − | |}
| + | <li value="D">[[Technical Information]] |
| − | </br><h4 > | + | <li value="E"> [[PlugIn Algorithm Median]] |
| − | Optimized Automatic Registration dialog box options
| + | *[[PlugIn Algorithm Median (con't, part 2)]] |
| − | <a name="wp1915044"></a></h4><div ><em>
| + | *[[PlugIn Algorithm Median (con't, part 3)]] |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | *[[PlugIn Algorithm Median (con't, part 4)]] |
| − | |+
| + | </ol> |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >'''Input options'''<a name="wp1915645"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Register'''<a name="wp1915651"></a></div>
| + | |
| − | | <div >Select the reference image from the list. The list alphabetically displays all opened images.<a name="wp1915653"></a></div>
| + | |
| − | | colspan="1" rowspan="3" | <div ><div align="center"><img src="images/OptimizedAutomaticRegistrationDialogBox.jpg" border="0" hspace="0" vspace="0"></div><a name="wp1915658"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Degrees of freedom'''<a name="wp1915660"></a></div>
| + | |
| − | | <div >Specifies the spatial model used to restrict the type of linear transformation being applied during registration. Select one of the following: rigid-6, global rescale-7, specific rescale-9, and affine-12.<a name="wp1915662"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Interpolation<a name="wp1915666"></a></div>
| + | |
| − | | <div >Specifies the type of interpolation the algorithm will use when resampling. The list includes: <a name="wp1915668"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · Trilinear
| + | |
| − | <a name="wp1915669"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · B-spline 3-rd order
| + | |
| − | <a name="wp1915670"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline"> | + | |
| − | · B-spline 4-th order
| + | |
| − | <a name="wp1915671"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · Cubic Lagrangian
| + | |
| − | <a name="wp1915672"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · Quintic Lagrangian
| + | |
| − | <a name="wp1915673"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · Heptic Lagrangian
| + | |
| − | <a name="wp1915674"></a></div><div style="color: #000000; font-size: 9pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 6pt; margin-top: 0pt; text-align: left; text-decoration: none; text-indent: 0pt; text-transform: none; vertical-align: baseline">
| + | |
| − | · Windowed sinc
| + | |
| − | <a name="wp1915675"></a></div><div >See also <a href="InterplationMethods.html#wp999048">"Interpolation methods used in MIPAV"</a>.<a name="wp1915678"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Cost function'''<a name="wp1915683"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Specify a cost function which will be used to evaluate registration. The list includes: Least squares, Correlation ratio, Normalized cross correlation, Normalized mutual information. For more information, refer to <a href="CostFunctions.html#wp999048">"Cost functions used in MIPAV algorithms" </a>.<a name="wp1915685"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Use the max of the min resolutions of the two datasets when resampling'''<a name="wp1915692"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >If this option is chosen in the dialog box, the method uses the maximum resolution of the two datasets. It throws away some image information, but works faster. If this option is not chosen the algorithms uses the minimum of the resolutions when resampling the images. This can be slower, but does not lose the information. <a name="wp1915694"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Initiate registration process by applying Least Squares<a name="wp1915698"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >This option provides a way to register images using the corresponding landmark points or VOIs placed in both images (you must put at least 4 landmarks). Note that this option works only for images that require rigid transformation e.g., rotation and/or translation, but it does not work on images that require scaling.<a name="wp1915700"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Rotations<a name="wp1915704"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Use this option if there is a need to make manual adjustments to registration. <a name="wp1915710"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Apply same rotations to all dimensions<a name="wp1915716"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >If checked, applies rotation that you specified to all dimensions. Otherwise, you should select the axis - X, Y, or Z - to which rotation should apply.<a name="wp1915718"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Coarse angle increment and Fine angle increment<a name="wp1915722"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >options can be used for two pass registration. The first pass would apply a coarse-level rotation to quickly arrive at an approximate registration, and then the second pass would apply a finer-level rotation to continue from the point where the first pass registration has finished. By default, the coarse angle increment is set to 15 degrees and fine angle increment is set to 6 degrees. <a name="wp1915728"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Weighted Images<a name="wp1915734"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >This gives weights for certain areas within the reference or transformed image. These areas can be specified using VOIs or reference files. Here, <em>higher weights mean a greater impact in that area on the registration.</em><a name="wp1915736"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >No weight<a name="wp1915740"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div > - the weighting factor is not used for registration.<a name="wp1915742"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="2" rowspan="1" | <div >Register area delineated by VOIs only<a name="wp1915746"></a></div>
| + | |
| − | | <div >uses VOI to register a selected area.<a name="wp1915750"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Weight registration<a name="wp1915752"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >If this option is checked, the following parameters become available:<a name="wp1915754"></a></div><div ><em>Choose ref. weight</em> opens the dialog where you can select the reference file which contains the information about the weighting factor.<a name="wp1915755"></a></div><div ><em>Choose input weight</em> opens the dialog where you can select the input file which contains the information about the weighting factor.<a name="wp1915756"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Output options<a name="wp1915760"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Display transformed image<a name="wp1915766"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >allows to view the transformed image in a separate window. By default, this option is active.<a name="wp1915768"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Interpolation: <a name="wp1915772"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >- select the calculation method, which is used to produce the output image from the array of transformation matrices and input image. You can choose among Trilinear, B-spline 3-rd order, B-spline 4-th order, Cubic lagrangian, Quintic lagrangian, Heptic lagrangian, or Windowed sinc. The default choice is trilinear.<a name="wp1915774"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Advanced options<a name="wp1915779"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Pressing the Advanced button opens the Advanced OAR Settings dialog box where you can specify the additional parameters which will be used to calculate the min cost function. Parameters are as follows:<a name="wp1915785"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Multiple of tolerance to bracket the minimum<a name="wp1915791"></a></div>
| + | |
| − | | <div >Recommended values are 10-60.<a name="wp1915793"></a></div>
| + | |
| − | | colspan="1" rowspan="3" | <div ><div align="center"><img src="images/AdvancedOARSettings.jpg" border="0" hspace="0" vspace="0"></div><a name="wp1915798"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Number of iterations<a name="wp1915800"></a></div>
| + | |
| − | | <div >Recommended are 1-5 iterations.<a name="wp1915802"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Number of minima from Level 8 to test at Level 4<a name="wp1915806"></a></div>
| + | |
| − | | <div >By default this is the best 20%, but you can enter your own number here.<a name="wp1915808"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="2" rowspan="1" | <div >Subsample image for speed <a name="wp1915812"></a></div>
| + | |
| − | | <div >subsamples the image.<a name="wp1915816"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Skip multilevel search. Assume images are close to alignment.<a name="wp1915818"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''OK'''<a name="wp1915824"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Applies the algorithm according to the specifications in this dialog box. <a name="wp1915826"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Cancel'''<a name="wp1915830"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Disregards any changes that you made in this dialog box and closes it.<a name="wp1915832"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Help'''<a name="wp1915836"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Displays online help for this dialog box.<a name="wp1915838"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >'''Figure 28. The Optimized Automatic Registration dialog box options '''<a name="wp1915846"></a></div>
| + | |
| − | |}</em><a name="wp1827490"></a></div><div >
| + | |
| − | <a name="wp1827349"></a></div><div >
| + | |
| − | <a name="wp1914546"></a></div><h4 >
| + | |
| − | Advanced OAR settings for Constrained Optimized Automatic Registration 3D
| + | |
| − | <a name="wp1914547"></a></h4><div ><em>
| + | |
| − | {| border="1" cellpadding="5" cellspacing="0"
| + | |
| − | |+
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Pressing the Advanced button in the main Optimized Automatic Image Registration 3D dialog box (see <a href="OptimizedAutomaticRegistration3D.html#wp1915846">Figure 28</a>) opens the Advanced OAR Settings dialog box where you can specify the additional parameters and (or) constrains (if Constrained AlgorithmOAR 3D was called) that will be used to calculate the min cost function. Parameters are as follows:<a name="wp1914697"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="2" rowspan="1" | <div >Multiple of tolerance to bracket the minimum<a name="wp1914703"></a></div>
| + | |
| − | | <div >Recommended values are 10-60.<a name="wp1914820"></a></div><div >
| + | |
| − | <a name="wp1914710"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Number of iterations<a name="wp1914712"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Recommended are 1-5 iterations.<a name="wp1914714"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="2" rowspan="1" | <div >Number of minima from Level 8 to test at Level 4<a name="wp1914718"></a></div>
| + | |
| − | | <div >By default, this is set to 5, but you can enter your own number here.<a name="wp1914722"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Translation search range<a name="wp1914828"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >These options allow the user to specify the limits in translations range (in the X, Y, and Z dimensions), and therefore, constrain the optimized registration. The options appear active only if the Constrained Optimized Registration algorithm is called.<a name="wp1914830"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Limit translation range<a name="wp1914834"></a></div>
| + | |
| − | | <div >If checked, this option limits the range of translations by the values (in mm) specified by the user.<a name="wp1914836"></a></div>
| + | |
| − | | colspan="1" rowspan="4" | <div ><div align="center"><img src="images/AdvancedOptionsForConstrainedOAR.jpg" border="0" hspace="0" vspace="0"></div><a name="wp1914838"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Apply same translation to all dimensions<a name="wp1914840"></a></div>
| + | |
| − | | <div >If checked, this option applies same limits to all dimensions.<a name="wp1914842"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Limits on X translation (or all translations)<a name="wp1914846"></a></div>
| + | |
| − | | <div >Specify the translation limit in X direction, or in all directions if the '''Apply same translation to all dimensions''' option is checked.<a name="wp1914848"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Limits on Y translation<a name="wp1914852"></a></div>
| + | |
| − | | <div >Specify the translation limit in Y direction.<a name="wp1914854"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Limits on Z translation<a name="wp1914937"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Specify the translation limit in Z direction.<a name="wp1914939"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="2" rowspan="1" | <div >Subsample image for speed <a name="wp1914724"></a></div>
| + | |
| − | | <div >subsamples the image.<a name="wp1914728"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >Skip multilevel search. Assume images are close to alignment.<a name="wp1914730"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >Initialize registration process by aligning COG's<a name="wp1915540"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >If checked, the algorithm, first, aligns centers of gravity (COG) or centers of mass (COM) of both images, and then proceed with registration. See also <a href="OptimizedAutomaticRegistration3D.html#wp1847495">"Calculating the center of mass" </a>.<a name="wp1915544"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >OK<a name="wp1914736"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Applies the algorithm according to the specifications in this dialog box. <a name="wp1914738"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Cancel'''<a name="wp1914742"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Disregards any changes that you made in this dialog box and closes it.<a name="wp1914744"></a></div>
| + | |
| − | |-
| + | |
| − | | <div >'''Help'''<a name="wp1914748"></a></div>
| + | |
| − | | colspan="2" rowspan="1" | <div >Displays online help for this dialog box.<a name="wp1914750"></a></div>
| + | |
| − | |-
| + | |
| − | | colspan="3" rowspan="1" | <div >'''Figure 29. The Advanced Optimized Automatic Registration (constrained) dialog box options '''<a name="wp1914757"></a></div>
| + | |
| − | |}</em><a name="wp1914765"></a></div><p ><a name="wp1914540"></a><p ><a name="wp1914541"></a><p ><a name="wp1914543"></a><p ><a name="wp1914544"></a><div style="color: #000000; font-size: 8pt; font-style: normal; font-weight: normal; margin-bottom: 0pt; margin-left: 12pt; margin-right: 0pt; margin-top: 2pt; text-align: left; text-decoration: none; text-indent: -12pt; text-transform: none; vertical-align: baseline"><a href="#wp1857635"><sup>1</sup></a>For 2D images. For 3D images, add the z coordinate also.
| + | |
| − | <a name="wp1857627"></a></div></blockquote>
| + | |
| | | | |
| − | ---- | + | == Talairach == |
| | + | *[[Introduction]] |
| | + | *[[Installing]] |
| | + | *[[Mapping Brains in Talairach Space]] |
| | + | **[[Task 1, Performing a semimanual transformation on datasets to the Talairach coordinate system]] |
| | + | **[[Task 2, Applying Talairach VOIs]] |
| | + | **[[Task 3, Segmenting the original image]] |
| | + | **[[Task 4, Transforming Talairach image and Talairach VOIs-1 to the original image]] |
| | + | **[[Task 5, Copying Talairach VOIs to segmented images]] |
| | + | **[[Task 6, Calculating statistics on VOIs]] |
| | + | **[[Viewing the Talairach grid on Talairach images]] |
| | | | |
| − | {| width="100%" border="0" cellspacing="0" cellpadding="0"
| + | == Frequently Asked Questions (FAQs) == |
| − | | valign="top" | <a href="NEIRegistrationPlugInBuildMPMaps.html"><img src="images/sprev.gif" border="0" alt="Previous"></a><a href="RegistrationAlighnPatientPosition(DICOM).html"><img src="images/snext.gif" border="0" alt="Next"></a>
| + | *[[FAQ: Understanding Memory]] |
| − | | align="right" | '''<font color="#800080" size="2" face="Verdana">Imaging Sciences
| + | *[[FAQ: Customizing MIPAV]] |
| − | Laboratory, CIT, NIH
| + | *[[FAQ: Understanding Image Basics]] |
| − |
| + | *[[FAQ: Understanding Changes in MIPAV]] |
| − | Matthew McAullife, Ph.D.
| + | *[[FAQ: Trimming VOIs]] |
| − |
| + | |
| − | mcmatt@exchange.nih.gov
| + | |
| − |
| + | |
| − | 301-594-2432
| + | |
| − |
| + | |
| | | | |
| − | Building 12A, Room 2041,
| + | == Videos == |
| − |
| + | |
| − | 9000 Rockville Pike, Bethesda,
| + | |
| − | MD 20892</font>'''
| + | |
| | | | |
| − | |}</body></html>
| + | *A video tutorial that explains how to load images into MIPAV and change the memory allocation - [http://www.youtube.com/embed/npSYWYJa530 play]. |
| | + | |
| | + | *A video tutorial that explains how to do basic image display options - [http://www.youtube.com/watch?v=02vUrfSaGWE&feature=relmfu play]. |
| | + | |
| | + | *A video tutorial that explains how to convert VOIs into binary masks that can be read by other programs such as Matlab or ImageJ. It also explains how to calculate statistics over VOIs - [http://www.youtube.com/watch?v=uCr-Zlxu454&feature=relmfu play]. |
| | + | |
| | + | * A video tutorial that explains how to save and capture PDFs of images - [http://www.youtube.com/watch?v=Zfh5nKITbkA&feature=relmfu play]. |
| | + | |
| | + | * A video tutorial that explains how to do basic image manipulations - [http://www.youtube.com/watch?v=uCr-Zlxu454&feature=relmfu play]. |
| | + | |
| | + | All tutorials are courtesy of [http://www.youtube.com/user/MIPAVvideos MIPAVvideos]. |
| | + | |
| | + | == Publications == |
| | + | * [[Publications]] |
This section is reflects the recent changes in documentation. It is mostly for Olga to help her tracking changes. Please don't pay any special attention to it.
No changes during the given period matching these criteria.
