Post Processing Method

Post Processing Methods: Advanced Techniques

When the overlapping axial images from a helical CT acquisition are loaded onto a workstation, sophisticated displays can be produced. The most commonly utilized displays are multiplanar reformatting, maximum intensity projections, surface and volume rendering and virtual endoscopic viewing. These image displays can be used in a broad range of applications to aid in the interpretation of the study and to better understand complex anatomic relationships. (Figure 12)

Post Processing Methods (Figure 12)

The images displayed represent an overview of the reconstruction processes discussed in this section. Moving clockwise from upper left: Maximum Intensity Projection of the renal arteries. Note the large amount of calcified plaque at the level of the renal arteries and extending distally. 3D volume rendered image displaying a large abdominal aortic aneurysm. The spine and pelvis are also shown to provide reference to the vasculature. Curved Multi-planar Reformat of the renal arteries provides a look at the entire length of the arteries rather than shorter segments. Minimum Intensity Projection of the trachea. This technique is used to display the minimum intensity pixels unlike the Maximum Intensity Projection which displays the maximum intensity pixel values. This corresponds to the Hounsfield values in CT where air has a low value and bone has a high value. The final image is a virtual endoscopic view of the colon. This technique is being used not only in the colon but also in the airways, sinuses and blood vessels.

Multiplanar reformatting allows the user to control object reorientation in any desired plane. Axial, sagittal, coronal, oblique and curved reformats can be generated in seconds. The maximum intensity projection algorithm, or MIP image, is generated by depicting the highest pixel value along a ray or projection. Multiple projections are generated and displayed in a cine loop. 3D visualization can be enhanced using 3D Surface and Volume Rendering techniques. Shaded Surface Display (SSD) relies on a threshold value to eliminate pixels above or below the threshold. The remaining pixels are then given equal values and the computer generates a surface illuminated by an imaginary point source. SSD provides depth information and is quite useful in the depiction of overlapping and tortuous vessels. Additionally, it is useful to demonstrate relationships of vessels relative to other anatomy. SSD does not, however, retain attenuation information and calcified plaque cannot be differentiated from intraluminal contrast. Occlusions can be artificially rendered in areas of stenosis where partial volume averaging results in pixels being below the threshold. Volume Rendering, unlike Shaded Surface Display which reduces the original data to a surface model, incorporates all the data contained in the volume into the displayed image. Each pixel is assigned a transparency factor and an opacity factor based on its density value. This enables visualization of anatomic details "above and beneath the surface." The major drawback to Volume Rendering relates to the large computational requirements of this technique.

Shaded surface display and volume rendered images can be created from a perspective that simulates views obtained by inserting an endoscope into body structures. Applications of virtual endoscopy or Endo 3D are "virtual" views of the airways, sinuses, vessels, the colon and other anatomical cavities. There are examples of these advanced reconstruction techniques throughout this course. Take special note of the reconstruction methods and how they provide additional diagnostic and therapeutic information above and beyond the conventional axial images.