next up previous contents
Next: Manual Coregistration Up: The Simple Coreg Tool Previous: The Simple Coreg Tool   Contents

Loading Data

Data for the coreg tool is loaded via the sequence tool. Two volumes are required: the target image volume i.e. the volume the user wishes to align to, and the source image volume i.e. the volume that will be manipulated with the parameters of the transformation model in order to align it to the target image volume. The Tina coreg tool calculates rigid coregistrations, where the transformation model consists of a translation, a rotation, and a scaling. Since this tool is used primarily to align medical data sets, which typically consist of $\approx 100000$ voxels, the maximum physically-meaningful accuracy can be achieved with only a small sample from the data. Therefore, in order to reduce the processor time required to perform a coregistration, only the data on three orthogonal planes projected through the image volumes are used. These three planes are specified by the centre point at which they intersect, which is entered by the user into the ``Centre x, y, z" fields in the simple coreg tool, and are displayed in the three coreg Tvs.

In order to load the data, first specify the target image volume and load it via. the sequence tool. Then specify the centre point in the simple coreg tool. Press the repaint button on any of the three coreg Tvs to display the three projected planes. The aim is to select three data-rich projections: alter the centre point and repeat until satisfactory projections are obtained. Setting the centre point to the actual centre of the image volume is usually a good starting point, although it should be remembered that the alignment will be most accurate at the centre point, and so if the intention is to examine only a particular structure, a centre point within that structure should be specified. Non-integer values can be entered as a centre-point: the projections displayed in the Tvs will be produced using the interpolation routine specified by the ``Reslice" choice fields (described below). Tri-linear interpolation (the ``linear" choce) is typically used as this represents a reasonable compromise between speed and accuracy.

Once satisfactory projections have been obtained, the data is latched. This stores the three projections displayed in the coreg tool Tvs, allowing a second image volume to be loaded. Latching of the data is achieved using the ``Latch" choice fields. These also specify how the data will be displayed when both volumes are loaded. The choices are

The ``anaglyph" choice is typically used. When the data is latched, the threshold specified in the AIR parameters dialog box is applied: voxels with grey-levels lower than the specified value will be ignored during the coregistration. This can be used to ignore the air space around MR images with certain modalities, or to eliminate certain image structures. However, in general the threshold should be set to a value lower than the lowest grey-level in the image volumes, ensuring that all of the image data is used in the coregistration. When the latch choice is selected, the number of voxels in the image volume that lie above the threshold will be displayed (as a percentage of the total number of voxels) in the text window of the top-level tinaTool. The displays in the three Tvs will also show the voxels that lie above the threshold in green, on top of the three projections in black-and-white, showing which structure have been selected.

Setting the latch choice back to ``image" at any point will discard the three projections through the target image volume stored in memory, and display the three projections through the image volume currently loaded into the sequence tool in the three coreg Tvs. This resets the coreg tool. If no sequence is currently loaded in the sequence tool, empty images will be displayed in the coreg Tvs.

Once the target images have been latched, the source image volume can be loaded via. the sequence tool. Pressing repaint in any of the three coreg Tvs will display the source image volume on top of the target image volume using the method specified by the latch choice. For example, if ``anaglyph" was selected, the target image volume will be displayed in green, and the source image volume in red.


next up previous contents
Next: Manual Coregistration Up: The Simple Coreg Tool Previous: The Simple Coreg Tool   Contents
root 2017-09-26