DOI: 10.1148/rg.255055044
RadioGraphics 2005;25:1409-1428
© RSNA, 2005
Informatics in Radiology (infoRAD)
Introduction to the Language of Three-dimensional Imaging with Multidetector CT1
Neal C. Dalrymple, MD,
Srinivasa R. Prasad, MD,
Michael W. Freckleton, MD and
Kedar N. Chintapalli, MD
1 From the Department of Radiology, University of Texas Health Science Center, 7703 Floyd Curl Dr, San Antonio, TX 78229-3900. Recipient of a Certificate of Merit award for an education exhibit at the 2004 RSNA Annual Meeting. Received March 7, 2005; revision requested May 23 and received June 22; accepted June 30. All authors have no financial relationships to disclose.
Address correspondence to N.C.D. (e-mail: dalrymplen{at}uthscsa.edu).
 |
Abstract
|
|---|
The recent proliferation of multi–detector row computed tomography (CT) has led to an increase in the creation and interpretation of images in planes other than the traditional axial plane. Powerful three-dimensional (3D) applications improve the utility of detailed CT data but also create confusion among radiologists, technologists, and referring clinicians when trying to describe a particular method or type of image. Designing examination protocols that optimize data quality and radiation dose to the patient requires familiarity with the concepts of beam collimation and section collimation as they apply to multi–detector row CT. A basic understanding of the time-limited nature of projection data and the need for thin-section axial reconstruction for 3D applications is necessary to use the available data effectively in clinical practice. The axial reconstruction data can be used to create nonaxial two-dimensional images by means of multiplanar reformation. Multiplanar images can be thickened into slabs with projectional techniques such as average, maximum, and minimum intensity projection; ray sum; and volume rendering. By assigning a full spectrum of opacity values and applying color to the tissue classification system, volume rendering provides a robust and versatile data set for advanced imaging applications.
© RSNA, 2005
|
Introduction
|
|---|
The recent proliferation of multi–detector row computed tomography (CT) has led to an increase in the creation and interpretation of images in planes other than the axial images traditionally viewed with CT. Powerful three-dimensional (3D) applications improve the utility of detailed CT data but also create confusion among radiologists, technologists, and referring clinicians when trying to describe a particular method or type of image. Parallel advances that have been made in the areas of CT acquisition and image processing software are of comparable importance, since postprocessing cannot improve on the finite constraints of the acquired CT data and innovative imaging paradigms are needed to optimize the use of exquisite and voluminous data.
The following examples are proposed as a guide to terminology commonly used when acquiring and manipulating CT data to create multiplanar and 3D images. Specific topics discussed are collimation; projection data; data reconstruction; section thickness and interval; nominal and effective section thickness; the volumetric data set; multiplanar reformation; curved planar reformation; average, maximum, and minimum intensity projection; shaded surface display; volume rendering; and segmentation. Although the technical aspects of data acquisition discussed are specific to CT, many of the postprocessing principles apply to magnetic resonance (MR) imaging as well.
|
Collimation
|
|---|
The concept of collimation is relatively straightforward with single–detector row CT. With the single–detector row technique, collimation refers to the act of controlling beam size with a metallic aperture near the tube, thereby determining the amount of tissue exposed to the x-ray beam as the tube rotates around the patient (1,2). Thus, in single–detector row CT, there is a direct relationship between collimation and section thickness. Because the term collimation may be used in several different ways in multi–detector row CT, it is important to distinguish between beam collimation and section collimation.
Beam Collimation
Beam collimation is the application of the same concept of collimation from single–detector row CT to multi–detector row CT. A collimator near the x-ray tube is adjusted to determine the size of the beam directed through the patient. Because multiple channels of data are acquired simultaneously, beam collimation is usually larger than reconstructed section thickness (3).
When a 16-channel scanner is used, for example, one of two settings is selected for most applications (Fig 1). Narrow collimation exposes only the central small detector elements. The data acquisition system controls the circuits that transmit data from the detector and collects data only from the intended elements (4,5). Wider collimation may expose the entire detector array. Unlike narrow collimation, in which the central elements are sampled individually, with wide collimation the 16 central elements are paired or binned, providing data as if they were eight larger elements (6). The four additional larger elements on each end of the detector array then complete the total of 16 channels of data. In this example, beam collimation would be 10 mm in the narrow setting or 20 mm in the wide setting.
View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1a. Beam collimation in 16-section CT. B = beam, C = collimator, DAS = data acquisition system, DE = detector elements, T = tube. (a) Narrow collimation exposes only the small central detector elements. (b) Wide collimation exposes all of the detector elements. The small central elements are paired or "binned" so that each pair acts as one larger element.
|
|
View larger version (64K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1b. Beam collimation in 16-section CT. B = beam, C = collimator, DAS = data acquisition system, DE = detector elements, T = tube. (a) Narrow collimation exposes only the small central detector elements. (b) Wide collimation exposes all of the detector elements. The small central elements are paired or "binned" so that each pair acts as one larger element.
|
|
Because beam collimation combined with table translocation determines the amount of z-axis coverage per rotation, it also helps determine the length of tissue or "volume coverage" that can be scanned within a given period (3). Larger beam collimation allows greater volume coverage within the time constraints of a given breath-hold or contrast material injection. An important point is that, as with single–detector row CT, narrow collimation in four- and 16-channel multi–detector row CT typically results in higher radiation dose to the patient compared with wide collimation (7,8).
Section Collimation
The concept of section collimation is more complex but vital to understanding the potential of multi–detector row CT. One of the key components of multi–detector row CT is a detector array that allows partition of the incident x-ray beam into multiple subdivided channels of data (3). Section collimation defines the acquisition according to the small axial sections that can be reconstructed from the data as determined by how the individual detector elements are used to channel data. As opposed to beam collimation, which determines volume coverage, section collimation determines the minimal section thickness that can be reconstructed from a given data acquisition.
Using the earlier example of a 16-channel scanner, let us assume that the small central detector elements are 0.625 mm and the large peripheral elements are 1.25 mm. The size of the elements exposed and the way in which data are sampled from them by the data acquisition system determine the physical properties of the projection data used to generate axial images (4,6,8). When narrow collimation is applied (in this example, an incident beam width of 10 mm), the central small detector elements are treated individually by the data acquisition system (Fig 2). This form of acquisition permits reconstruction of axial sections as small as the central detector elements, or a section collimation of 0.625 mm.
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2a. Section collimation in multi–detector row CT. (a) Narrow collimation is coordinated with the data acquisition system (DAS) to allow use of the small central detector elements (DE) individually, resulting in 16 sections with a thickness of 0.6 mm each. This setting allows data reconstruction down to a section thickness of 0.6 mm. (b) Wide collimation is coordinated with the data acquisition system (DAS) to pair the 16 small central detector elements (DE) and use the eight peripheral elements individually, resulting in 16 sections with a thickness of 1.2 mm each. This setting allows data reconstruction down to a section thickness of 1.2 mm.
|
|
View larger version (31K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2b. Section collimation in multi–detector row CT. (a) Narrow collimation is coordinated with the data acquisition system (DAS) to allow use of the small central detector elements (DE) individually, resulting in 16 sections with a thickness of 0.6 mm each. This setting allows data reconstruction down to a section thickness of 0.6 mm. (b) Wide collimation is coordinated with the data acquisition system (DAS) to pair the 16 small central detector elements (DE) and use the eight peripheral elements individually, resulting in 16 sections with a thickness of 1.2 mm each. This setting allows data reconstruction down to a section thickness of 1.2 mm.
|
|
When wide beam collimation (20 mm in this example) is used, the central elements are coupled so that two 0.625-mm elements are sampled as a single 1.25-mm element and the peripheral 1.25-mm elements are sampled individually, resulting in a section collimation of 1.25 mm. As a result, axial sections cannot be reconstructed smaller than 1.25 mm. Thus, section collimation is defined by the effective size of the channels of data sampled by the data acquisition system (the individual or coupled detector elements) and determines the minimum section thickness that can be reconstructed in a given acquisition mode. "Effective detector row thickness" is another term that has been used to describe section collimation (8).
If a routine abdominal examination interpreted at 5-mm section thickness reveals a finding and the radiologist or surgeon would like detailed coronal images, the section collimation determines whether the data can be reconstructed to 0.625-mm or 1.25-mm section thickness to provide a new data set for the reformatted images. Although it may be tempting to use the smallest section collimation available routinely, this may increase radiation dose to the patient (particularly with four- to 16-channel scanners) (7,8). Thus, section collimation is an important consideration in designing protocols with multi–detector row CT, as the anticipated need for isotropic data must be balanced with radiation dose considerations.
Section collimation and the quantity of data channels used during data acquisition are described by the term "detector configuration." For example, the detector configuration for a 16-channel scanner acquiring 16 channels of data, each 0.625 mm thick, is described as 16 x 0.625 mm. The same scanner could also acquire data by using different detector configurations, including 16 x 1.25 mm and 8 x 2.5 mm. The detector configuration also describes the relationship between section and beam collimation, since beam collimation can be calculated as the product of the section collimation and the number of data channels used (5,8).
Although section profiles for thin and thick collimation vary among different vendors, the general principles are applicable to all scanners. Correlation between beam collimation and section collimation on different types of 16-channel scanners is shown in the Table.
|
Projection Data
|
|---|
Projection data are the initial product of CT acquisition prior to filtered back projection and the longitudinal interpolation necessary to create axial reconstructed sections. Projection data consist of line integrals and are never viewed directly but are used to generate axial images. There are several reasons to recognize projection data in clinical practice: (a) Spatial properties of the projection data are defined by scan acquisition and cannot be altered subsequently. (b) Only the projection data are used to reconstruct axial images, so any retrospective data reconstruction requires access to the projection data. (c) Projection data are not used directly to create 3D images. (d) In most cases, it is not practical to archive these large data sets, so access to generate volumetric data sets is time limited.
The finite constraints of the projection data make it necessary to anticipate which applications are likely to be helpful in the interpretation of a particular type of examination before it is performed so that data with the requisite z-axis or "through-plane" spatial resolution are available (1). When 3D reformations are likely to be beneficial, appropriate thin-section reconstructions must be performed before the projection data are deleted. With this in mind, routine secondary data reconstruction may be performed for certain categories of examinations. Increasing the data storage capacity of the scanner can prolong accessibility to the data, decreasing the chances of frustration that may occur when additional image reconstruction is desired after the projection data are no longer available.
|
Data Reconstruction
|
|---|
Data or image reconstruction refers to the process of generating axial images from projection data (Fig 3). Axial data sets can be viewed for interpretation or used to create multiplanar or 3D images. This requires increasingly sophisticated interpolation algorithms that take into account redundancies in overlapping data, effects of table speed, and geometric variability of the cone beam tube output (5,9,10). Section thickness, reconstruction interval, field of view, and convolutional kernel (reconstruction algorithm) must be specified each time data are reconstructed. Multiple data reconstructions can be performed automatically for a variety of reasons, such as including both soft-tissue and lung kernels of the chest or providing a thin-section data set for 3D applications. Additional retrospective data reconstruction can be performed as long as the projection data remain available (2).
View larger version (38K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3. Reconstruction of axial images from projection data. Projection data are never viewed directly. Rather, they are used to generate axial images. In multi–detector row CT, images used for primary axial interpretation usually have a section thickness several times larger than the minimum thickness available and may be called "thick sections." However, axial images can also be generated with a smaller section thickness, as determined by the section collimation. These are usually called "thin sections" and are essential for creating multiplanar reformatted and 3D images.
|
|
|
Section Thickness and Interval
|
|---|
Section thickness is the length of each segment of data along the z axis used during data reconstruction to calculate the value of each pixel on the axial images through a combination of helical interpolation and z-filtering algorithms (3,4,10–12). This determines the volume of tissue that will be included in the calculation to generate the Hounsfield unit value assigned to each of the pixels that make up the image (13). Reconstruction interval or increment refers to the distance along the z axis between the center of one transverse (axial) reconstruction and the next. Interval is independent of section thickness and can be selected arbitrarily since it is not limited by scan acquisition (2,14). When section thickness and interval are identical, images are considered to be contiguous.
In some cases, such as high-resolution CT of the chest, a small section thickness is selected to provide high spatial resolution but may be sampled at large intervals through the lung to obtain a representative sample with a limited number of images (eg, 1-mm section thickness at a 10-mm interval). Such discontinuous images are appropriate for evaluating generalized parenchymal disease in the lungs, but lung nodules can easily be missed. For 3D imaging, an overlapping interval is usually selected, meaning that the interval is smaller than the section thickness, usually by 50% (Fig 4) (14–17). For example, 1.25-mm sections can be reconstructed every 0.625 mm so that the redundancy of data along the z axis results in smooth coronal or sagittal reformations. Although the section thickness is limited by the section collimation selected for scan acquisition, reconstruction interval is not limited by scan parameters (18). Even data reconstructed to the smallest section thickness available can be overlapped by using a smaller interval if necessary.
View larger version (71K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4a. Effects of an overlapping reconstruction interval. (a) Contiguous data set reconstructed with a section thickness and interval of 2.5 mm. Coronal reformatted image shows a jagged cortical contour due to stair-step artifact. (b) Overlapping data set reconstructed with a section thickness of 2.5 mm but with the interval decreased to 1.25 mm, an overlap of 50%. Such overlapping minimizes stair-step artifact and improves demonstration of a fracture of the right superior pubic ramus (arrowhead).
|
|
View larger version (70K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4b. Effects of an overlapping reconstruction interval. (a) Contiguous data set reconstructed with a section thickness and interval of 2.5 mm. Coronal reformatted image shows a jagged cortical contour due to stair-step artifact. (b) Overlapping data set reconstructed with a section thickness of 2.5 mm but with the interval decreased to 1.25 mm, an overlap of 50%. Such overlapping minimizes stair-step artifact and improves demonstration of a fracture of the right superior pubic ramus (arrowhead).
|
|
|
Nominal and Effective Section Thickness
|
|---|
As in single–detector row CT, table translation during scan acquisition and the interpolation algorithm used to generate axial sections have an effect on section thickness. Nominal section thickness is the section thickness specified by the collimation when a protocol is entered on the scanner. The actual section thickness of the reconstructed data is dependent not only on collimation but also on table speed and the method of z interpolation used (4,5,10,18–22). The term "effective section thickness" can be used to describe actual section thickness after broadening effects are taken into consideration (5). Some vendors provide this information on the image header or on the menu for image reconstruction (Philips Medical Systems, Siemens Medical Solutions, Toshiba Medical Systems); other vendors display only the nominal section thickness (GE Healthcare Technologies). Scan acquisition with a 16 x 1.25-mm detector configuration may result in effective section thickness of 1.3 mm with a low pitch and 1.5 mm with a higher pitch.
|
Volumetric Data Set
|
|---|
Although the diagnostic potential and sheer size of detailed CT data sets available with multi–detector row CT are likely to encourage integration of 3D imaging techniques into interpretation of even routine examinations (23), axial section interpretation remains an essential component of CT interpretation. While thin-section data sets may be reconstructed primarily when an examination is performed specifically for the purposes of CT angiography, colonography, or other advanced applications, 3D rendering techniques may also be useful for more routine examinations. To maintain acceptable contrast resolution on the primary axial interpretation sections, relatively thick sections are still reconstructed in most cases, typically ranging from 3 to 5 mm (8). Examinations performed with a field of view of 30–40 cm result in a pixel size of 0.5–0.8 mm on the axial sections, so a section thickness of 0.5–0.8 mm is required to generate a data set with similar spatial resolution in each dimension; such data are called isotropic data (Fig 5 ) (4,5,24,25).
View larger version (45K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5a. Anisotropic and isotropic data. (a) Single–detector row CT performed with a nominal section thickness of 5 mm and a 512 x 512 matrix results in reconstructed data that are anisotropic, consisting of voxels with a facing pixel size of approximately 0.625 mm but a depth of 5 mm. This data set provides satisfactory axial images but has limited potential for secondary data reconstruction. (b) Sixteen-channel CT performed with wide collimation results in reconstructed data that are anisotropic, with a z-axis dimension (1.25 mm) approximately twice the size of the facing pixel (0.625 mm). By overlapping the reconstruction interval (which is not limited by section collimation), this data set provides excellent reformatted and volume-rendered images for many applications. (c) Sixteen-channel CT performed with narrow collimation results in reconstructed data that are isotropic, consisting of voxels that are relatively symmetric in all dimensions (0.625 mm). This data set provides exquisite data for multiplanar and 3D applications.
|
|
View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5b. Anisotropic and isotropic data. (a) Single–detector row CT performed with a nominal section thickness of 5 mm and a 512 x 512 matrix results in reconstructed data that are anisotropic, consisting of voxels with a facing pixel size of approximately 0.625 mm but a depth of 5 mm. This data set provides satisfactory axial images but has limited potential for secondary data reconstruction. (b) Sixteen-channel CT performed with wide collimation results in reconstructed data that are anisotropic, with a z-axis dimension (1.25 mm) approximately twice the size of the facing pixel (0.625 mm). By overlapping the reconstruction interval (which is not limited by section collimation), this data set provides excellent reformatted and volume-rendered images for many applications. (c) Sixteen-channel CT performed with narrow collimation results in reconstructed data that are isotropic, consisting of voxels that are relatively symmetric in all dimensions (0.625 mm). This data set provides exquisite data for multiplanar and 3D applications.
|
|
View larger version (54K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5c. Anisotropic and isotropic data. (a) Single–detector row CT performed with a nominal section thickness of 5 mm and a 512 x 512 matrix results in reconstructed data that are anisotropic, consisting of voxels with a facing pixel size of approximately 0.625 mm but a depth of 5 mm. This data set provides satisfactory axial images but has limited potential for secondary data reconstruction. (b) Sixteen-channel CT performed with wide collimation results in reconstructed data that are anisotropic, with a z-axis dimension (1.25 mm) approximately twice the size of the facing pixel (0.625 mm). By overlapping the reconstruction interval (which is not limited by section collimation), this data set provides excellent reformatted and volume-rendered images for many applications. (c) Sixteen-channel CT performed with narrow collimation results in reconstructed data that are isotropic, consisting of voxels that are relatively symmetric in all dimensions (0.625 mm). This data set provides exquisite data for multiplanar and 3D applications.
|
|
Because only thin-section data with isotropic or near-isotropic properties provide diagnostic quality through-plane (long-axis) resolution, two separate data sets are often reconstructed: (a) a primary reconstruction consisting of relatively thick sections for axial interpretation and (b) a volumetric data set consisting of thin overlapping sections for 3D rendering (Fig 6). Optimal results are usually achieved by selecting the smallest section thickness available from the raw projection data (26). As discussed earlier, only section thickness is limited by scan parameters, so sections can be reconstructed at an interval smaller than the section thickness, resulting in overlap of data along the z axis (eg, reconstruction of 1.25-mm-thick sections every 0.625 mm) (1,14,18,27).
View larger version (44K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 6. Use of a volumetric data set. Projection data are typically used to reconstruct axial images of interpretive thickness for conventional review, which is performed by using printed film or with a picture archiving and communication system. Although it is occasionally useful to view thin axial images for osseous detail, axial viewing is usually performed with a section thickness of 3–5 mm. If necessary, a thin-section data set can be generated in addition to or in place of the traditional interpretive axial images. This may be called the volumetric data set because it is intended to be used not for primary axial interpretation but rather for generating high-quality multiplanar reformatted or volume-rendered images. This data set typically consists of axial images with a section thickness approaching 1 mm or even less, preferably with an overlapping interval.
|
|
Although projection data are stored on the scanner only for a limited time, a reconstructed thin-section data set can be archived on storage media or in a picture archiving and communication system, allowing access to high-quality image applications at a future date. Data reconstruction usually takes significantly longer than scan acquisition, and routine generation of large data sets can hinder scanner work flow at slow rates of reconstruction. If a scanner is purchased in anticipation of advanced 3D applications, rapid data reconstruction should be considered a priority.
|
Multiplanar Reformation
|
|---|
Multiplanar reformation (MPR) is the process of using the data from axial CT images to create nonaxial two-dimensional images (Fig 7). MPR images are coronal, sagittal, oblique, or curved plane images generated from a plane only 1 voxel in thickness transecting a set or "stack" of axial images (15,23,24,28). This technique is particularly useful for evaluating skeletal structures, since some fractures and joint alignment may not be readily apparent on axial sections.
View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7a. MPR. (a) Coronal reformatted image from routine abdominal-pelvic CT of a patient with bowel ischemia related to systemic lupus erythematosus vasculitis. Imaging in the coronal plane allowed visualization of bowel loop distribution throughout the abdomen and pelvis on a total of 28 images. Thickened distal loops of ileum are seen in the right lower quadrant with dilatation of more proximal small bowel loops. Arterial and venous patency was confirmed with this examination. (b) Sagittal reformatted image produced from CT data acquired with a trauma protocol. Examination of the chest, abdomen, and pelvis was performed with a detector configuration of 16 x 1.25 mm. Although a primary reconstruction thickness of 5 mm was used for axial interpretation, secondary data reconstruction to a section thickness of 1.25 mm at an interval of 0.625 mm allows a set of detailed full-spine sagittal images (approximately 20 1.5-mm-thick sections) to be created for every trauma case.
|
|
View larger version (67K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 7b. MPR. (a) Coronal reformatted image from routine abdominal-pelvic CT of a patient with bowel ischemia related to systemic lupus erythematosus vasculitis. Imaging in the coronal plane allowed visualization of bowel loop distribution throughout the abdomen and pelvis on a total of 28 images. Thickened distal loops of ileum are seen in the right lower quadrant with dilatation of more proximal small bowel loops. Arterial and venous patency was confirmed with this examination. (b) Sagittal reformatted image produced from CT data acquired with a trauma protocol. Examination of the chest, abdomen, and pelvis was performed with a detector configuration of 16 x 1.25 mm. Although a primary reconstruction thickness of 5 mm was used for axial interpretation, secondary data reconstruction to a section thickness of 1.25 mm at an interval of 0.625 mm allows a set of detailed full-spine sagittal images (approximately 20 1.5-mm-thick sections) to be created for every trauma case.
|
|
Multiplanar images can be "thickened" into slabs by tracing a projected ray through the image to the viewer’s eye, then processing the data encountered as that ray passes through the stack of reconstructed sections along the line of sight according to one of several algorithms (Fig 8) (24,29,30). Projectional techniques used in "thickening" of multiplanar images include maximum intensity projection (MIP), minimum intensity projection (MinIP), AIP, ray sum, and volume rendering and are sometimes called "multiplanar volume reformations" (31).
View larger version (53K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 8. Row of data encountered along a ray of projection. The data consist of attenuation information calculated in Hounsfield units. The value of the displayed two-dimensional pixel is determined by the amount of data included in the calculation (slab thickness) and the processing algorithm (maximum, minimum, or average intensity projection [AIP] or ray sum).
|
|
|
Curved Planar Reformation
|
|---|
Curved planar reformation is a type of MPR accomplished by aligning the long axis of the imaging plane with a specific anatomic structure, such as a blood vessel, rather than with an arbitrary imaging plane (15,16). Curved planar reformations can be created to include an entire structure on a single image. This is particularly useful in displaying an entire vessel, a ureter, or a long length of intestine, as these tubular structures are otherwise seen only by following them on consecutive images (Fig 9). Unlike surface- or volume-rendered 3D images, curved planar images display the cross-sectional profile of a vessel along its length, facilitating characterization of stenoses or other intraluminal abnormalities.
View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9a. Curved planar reformation. (a) Three-dimensional volume-rendered image shows the curved course of the right coronary artery. (b) Curved planar image of the right coronary artery shows a cross section of the vessel in its entirety. In this case, several points were selected along the course of the vessel on axial images; semiautomated software then defined an imaging plane that includes the entire length of the vessel. Because the imaging plane is defined by the vessel, other structures in the image are distorted.
|
|
View larger version (155K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 9b. Curved planar reformation. (a) Three-dimensional volume-rendered image shows the curved course of the right coronary artery. (b) Curved planar image of the right coronary artery shows a cross section of the vessel in its entirety. In this case, several points were selected along the course of the vessel on axial images; semiautomated software then defined an imaging plane that includes the entire length of the vessel. Because the imaging plane is defined by the vessel, other structures in the image are distorted.
|
|
However, manual derivation of the curved plane can be time-consuming and may result in artifactual "pseudolesions." The recent introduction of automated methods for generating curved planar reformations has been shown to decrease user interaction time by 86% while maintaining image quality and actually decreasing the number of artifacts (32). The concept of thickening MPRs into slabs may be applied to curved planar reformations as well, resulting in curved slab reformations (33).
|
Average Intensity Projection
|
|---|
AIP describes one type of algorithm used to thicken MPRs. The image represents the average of each component attenuation value encountered by a ray cast through an object toward the viewer’s eye (Fig 10). Starting with an MPR with a thickness of only 1 pixel (0.5–0.8 mm), thickening the multiplanar slab by using AIP may be used to produce images that have an appearance similar to traditional axial images with regard to low contrast resolution (Fig 11). This can be useful for characterizing the internal structures of a solid organ or the walls of hollow structures such as blood vessels or the intestine.
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 10. AIP of data encountered by a ray traced through the object of interest to the viewer. The included data contain attenuation information ranging from that of air (black) to that of contrast media and bone (white). AIP uses the mean attenuation of the data to calculate the projected value.
|
|
View larger version (169K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11a. Effects of AIP on an image of the liver. (a) Coronal reformatted image created with a default thickness of 1 pixel (approximately 0.8 mm). (b) Increasing the slab thickness to 4 mm by using AIP results in a smoother image with less noise and improved contrast resolution. The image quality is similar to that used in axial evaluation of the abdomen.
|
|
View larger version (160K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 11b. Effects of AIP on an image of the liver. (a) Coronal reformatted image created with a default thickness of 1 pixel (approximately 0.8 mm). (b) Increasing the slab thickness to 4 mm by using AIP results in a smoother image with less noise and improved contrast resolution. The image quality is similar to that used in axial evaluation of the abdomen.
|
|
A different processing algorithm, ray sum, is offered on some workstations in place of or in addition to AIP. Rather than averaging the data along each projected ray tracing, ray sum simply adds all values, as the name implies (30). Therefore, full-volume ray sum images may have an appearance similar to that of a conventional radiograph. However, thin-slab ray sum produces images that appear similar to AIP images.
|
Maximum Intensity Projection
|
|---|
MIP images are achieved by displaying only the highest attenuation value from the data encountered by a ray cast through an object to the viewer’s eye (Fig 12) (29,34). MIP is best used when the objects of interest are the brightest objects in the image (35) and is commonly used to evaluate contrast material–filled structures for CT angiography and CT urography. Large-volume MIP images have long been used to obtain 3D images from MR angiography data (30). Because only data with the highest value are used, MIP images usually contain 10% or less of the original data, a factor that was critical when computer processing power limited accessibility to advanced imaging techniques (35).
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 12. MIP of data encountered by a ray traced through the object of interest to the viewer. The included data contain attenuation information ranging from that of air (black) to that of contrast media and bone (white). MIP projects only the highest value encountered.
|
|
Thick-slab MIPs can also be applied to CT angiography data to include long segments of a vessel, but thin-slab MIP images (with section thickness less than 10 mm) viewed in sequence may provide more useful diagnostic information, as small structures are less likely to be obscured (Fig 13) (36,37). Although large-volume MIP images can demonstrate vessels in their entirety, the appreciation of 3D relationships between structures remains limited by a lack of visual cues that allow perception of depth relationships (16).
View larger version (125K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13a. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
View larger version (123K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13b. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
View larger version (131K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13c. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13d. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13e. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 13f. Effects of MIP slab thickness on a coronal image of the abdomen. (a, b) Changing from the AIP technique (a) to the MIP technique (b) at a fixed slab thickness of 2.5 mm results in increased conspicuity of vessels. (c–f) More vessels are included per image as the section thickness increases to 5 mm (c), 10 mm (d), 15 mm (e), and 20 mm (f). However, use of thick slabs also results in obscuration of the vessels by other high-attenuation structures (bones, other vessels).
|
|
|
Minimum Intensity Projection
|
|---|
Top
Abstract
Introduction
