Table of Contents
Cone Beam Radiography
Three dimensional radiography
The newest innovation in dental radiography is three dimensional radiography. The technology that produces these three dimensional images is called Cone Beam Computerized Tomography (CBCT). As the price of these machines come down, more and more of them will be showing up in dental offices. The technological innovation that made this technology possible was partly improvements in x-ray and sensor technology, but even more important were ever faster and cheaper computers with enormous memory capacity and some very innovative graphics programming.
And indeed, it is the capacity to store and quickly process huge amounts of data that can be computationally manipulated to reveal internal structures in the human body in virtually any orientation. Today it takes only a few minutes for a computer to create three dimensional images for the clinician. In contrast, the first CT scanner, unveiled in 1972 took two and a half hours of number crunching to produce a single two dimensional slice using the fastest computer available at that time.
Cone beam technology gives the clinician undreamed of information, and using all the computational power of a late generation computer, he or she can image bone and soft tissue in high resolution, see normal and pathological anatomy, and even measure structures accurately for procedures such as implants, root canals and reconstructive surgery.
How does the Cone Beam work?
The principles used in three-dimensional radiography are really the same ones that form the basis of CT scanning. Like linear array CT scanners, cone beam images are made on a machine with a rotating gantry, a continuously operating x-ray source, and a detector array. Unlike the fan shaped beam used in a CT scan, however, CBCT uses a cone, or pyramid shaped beam, and the linear array of detectors used in CT is replaced with a two dimensional array similar to a very large version of the CCD’s used in intraoral digital radiography. The patient is generally seated so that the area of interest (in the dental world, it is usually the mandible and/or maxilla) is centered in the beam. The patient is seated in such a way that the axis of rotation is centered in the area of interest. The x-ray source and the detector revolve around the patient in unison, the same as is done in a CT scan.
During the rotation of the device, between 150 and 600 images are captured by the detector. The images, again called views, are simple two dimensional arrays of pixels, and each one, by itself, resembles an ordinary AP or Lat (anterior-posterior or lateral) view of the field of view. A single rotation is all that is required to capture a three dimensional image of the structures in the field.
A three-dimensional image needs different terminology to describe its individual picture elements. We are used to speaking in terms of pixels (meaning picture elements) when dealing with ordinary digital images. However the term used when dealing with three dimensional picture elements is “voxels“. A voxel has not only height and width, as does a pixel, but it has depth as well.
Each view represents a coherent two-dimensional image recognizable by the human eye, so it is easy to visualize the motion picture analogy used when discussing the Clark shift on the CT scanning page. Just as important, however, is the fact that each individual pixel on any given view represents the total density of all objects that the x-ray beam had to traverse on its journey between the x-ray source and the pixel detector that registers it. Thus one of the algorithms used in creating the final image involves steps similar to the steps used in creating slices in the plane of the beam, just like the algorithm used in CT scans. (Click on the icon to the right to read this section if you haven’t already done so.)
Another algorithm used to manipulate the huge amount of data resulting from up to 600 very large two-dimensional digital images, plus the data from the third dimension accumulated in the first step discussed above, boils down to a really complex mathematical version of the much simpler mechanical algorithm a draftsman uses to draw a three-dimensional view of an object, using just a couple of two-dimensional orthographic drawings.
The orthographic drawings in the left hand image above are two-dimensional views of a “widget”. These simple drawings represent simplified versions of the two-dimensional views of the much more complex image captured by the cone beam detector. The letters stand for the lengths of each vertical and horizontal line in the drawing. In order to get from the multiple two-dimensional drawings on the left to the three dimensional drawing on the right, the draftsman uses a T-square and a 30 degree triangle.
The triangle allows the draftsman to draw lines at right angles to the T-square, as well as lines at 30 degrees in either direction. He begins by using a set of axes at 30 degrees, like the one on the left above. He then measures along the three axes using the same measurements taken from the orthographic drawings. He draws all horizontal lines along the thirty degree axes, and the vertical lines remain at ninety degrees to the t-square.
Finally, he adds the missing vertical and horizontal lines and ends up with the three dimensional drawing.
The mathematical transforms used in CBCT, while immensely complex, go about their business in nearly the same manner as the draftsman. The cone beam image is, of course, much more complex, with many layers of internal structure, but the principles are the same. Instead of three images of the object looked at from three different angles, the cone beam uses hundreds of images taken from hundreds of angles, as well as the depth “slice” images created using the first algorithm described on the CT scan page. Another advantage the computer enjoys is that once the initial three-dimensional image is constructed, it can rapidly rotate the image to any angle, slice any part of the image at any angle, and allow the practitioner to take virtually any measurement he or she wants. It also allows the practitioner to cut virtual windows through exterior layers to reveal the hidden interior anatomical structures in three dimensions.
Cone Beam technology provides a very significant dose reduction as compared to the conventional CT used for maxillofacial imaging. While older CT scans can expose a patient to an effective dose of up to 2000 mSV, a cone beam scan confined to the maxilofacial region can reduce this dose by up to 98.5%. It does this for a number of reasons:
- CBCT x-ray sources are tuned for seeing hard tissues which means that unlike CT scanners, cone beams are low intensity, high energy beams. This means fewer x-ray photons are needed per image, less absorption in soft tissues and less scatter.
- The sensors are exquisitely sensitive and need fewer photons to illuminate them.
- The collimation can be controlled by the operator to allow the beam to illuminate only the portion of the body under consideration.
- The scan is quite rapid, which not only reduces the patient’s exposure to ionizing radiation, but also reduces distortion due to patient movement.
For our purposes, resolution is most easily defined as the fuzziness of the image.
Cone beam sensors have a wide range of spatial definition, from 0.4 mm to as low as 0.076 mm. A machine with the 0.4 mm resolution would present images with fuzzier edges than one with the 0.076 mm resolution. On the other hand, the sensors are so sensitive that the highest resolution images are somewhat degraded due to noise and contrast artifacts caused by large amounts of scatter radiation, so there is probably a practical limit on the need to strive to buy the very highest resolution machines for normal clinical practice.
Other advantages to cone beam technology
Viewing hidden structures in three-dimensions gives the clinician the ability to see spatial relationships at a glance. Even without advanced software modules, once the three-dimensional image is rendered, the ability to see the bony structures and to rotate the image so that they can be seen from any angle makes diagnosis and treatment planning easier and more accurate. This includes assessment of bony and dental pathologies, recognition of structural maxilofacial deformities and fractures, preoperative assessment of impacted teeth, TMJ imaging, orthodontic evaluations, and assessment for the adequacy of bone available for implant placement.
Advanced software rendering techniques are limited only by the imagination of the software developer. These include, but are not limited to the following:
- The on-screen objects are neither distorted nor magnified. Thus the clinician can make accurate on-screen measurements of dental and bony structures using various on-screen measuring techniques and record them on a separate layer which may be overlaid on the original image
- slicing the image in virtually any plane, not limited to axial, sagittal and coronal planes, which creates images similar to those produced by a CT scanner.
- Magnification, or zooming in on specific areas of interest
- Cutting virtual windows through more superficial structures in order to view the underlying structures in anatomic relation to the overlying structures.
- The ability to add annotations including drawings and alphanumeric information to layers which are overlaid on top of the original image, but may be removed en masse for better visualization of the original image.
Alas! All is not a bed of roses in the world of cone beam imaging. With the improved technologies come increased responsibilities. A dentist might use his cone beam image for the reasons specific to his specialty, such as assessment for the adequacy of bone for implant placement, but if he or she does not recognize abnormal anatomical structures, he or she could be held legally responsible if the patient suffers future injuries relating to that missed observation. For example, if an adenocarcinoma has caused visible distortion or disintegration of any bony structure seen in the scan, the dentist is responsible for notifying the patient and referring the patient to an appropriate specialist, or baring that, for enlisting the help of a board certified radiologist to assess the images.
This also means that if the image includes the entire sinus region, the dentist is responsible for recognizing abnormalities in the sinus, even though this lies outside of his area of expertise and he has no training in that subject.
For this reason, it is unwise to begin using cone beam images until the clinician has had at least some training on the recognition of abnormalities in any part of the head or neck that he plans to image. Even if the scan is done by referral to another office, it remains the responsibility of the referring dentist to recognize pathology when he or she sees it. Conversely, any dentist who accepts referrals to use his or her cone beam for a patient from another office can be held legally responsible for not correctly reading possibly harmful pathological conditions on the scan, (and thus not notifying the referring dentist or the patient) even if the patient is not a patient of record in his own office.