| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
J Am Dent Assoc, Vol 136, No 10, 1379-1387.
© 2005 American Dental Association |  |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
COVER STORY |
The uses of digital radiography in dental practice
 |
ABSTRACT |
|---|
Overview. The two technologies now available are solid-state systems and phosphor plate systems. Each has its strong points, and the choice of which to use depends on the type of dental practice. Image processing improves the diagnostic quality of the radiographic information. Advanced image-processing techniques, such as subtraction radiography, are available for specialized clinics.
Conclusions and Clinical Implications. Digital radiography no longer is an experimental modality. It is a reliable and versatile technology that expands the diagnostic and image-sharing possibilities of radiography in dentistry. Optimization of brightness and contrast, task-specific image processing and sensor-independent archiving are important advantages that digital radiography has over conventional film-based imaging.
Key Words: Digital radiography; image processing; diagnosis
In 1895, German physicist Wilhelm Conrad Roentgen discovered the X-ray. Within two weeks after Roentgen made his discovery public, the first dental radiograph was made by German dentist Otto Walkoff, who placed in his own mouth small glass photographic plates wrapped in rubber dam and exposed them for 25 minutes. The characteristics and, more specially, the sensitivity of film have been improved considerably in the intervening 110 years. Patients and dental team members alike benefit from the reduction in necessary radiation dose that has resulted from the increased sensitivity of dental film. The primary principles of radiation protection in diagnostic radiology are justification (the patient will experience more benefit than harm from the exposure) and the "as low as reasonably achievable" (ALARA) principle (which also takes into account social and economic factors).
The first digital X-ray sensors for use in dentistry were introduced in the mid-1980s by Francis Mouyen (RVG, Trophy Radiologie, Croissy Beaubourg, France [now Trophy, A Kodak Company, Rochester, N.Y.]). The first dental digital system was capable only of acquiring a radiographic image; the image could not be stored on disk but had to be printed. However simple it appeared to be, it marked the start of a new era. Shortly thereafter, another system was developed by Per Nelvig and colleagues (Sens-A-Ray, Regam Medical Systems, Sundsvall, Sweden), and within a decade many more manufacturers entered the market. Digital systems have improved considerably since then and now are a well-accepted and useful technology in dental diagnosis.
Dose reduction often has been emphasized as one of the biggest advantages of digital radiography. It will be shown later in this article that this can be argued. There is another reason, however, why digital radiography adheres to the ALARA principle in an important way: information recorded in the radiographic image can be made available to the observer more easily when the image is in a digital format than when it is in an analog format. The most significant advantages of digital imaging, therefore, are computer-aided image interpretation and image enhancement, in addition to the obvious options of standardized image archiving and image retrieval. Figure 1Digital radiography is a reliable and versatile technology that expands the diagnostic and image-sharing possibilities of radiography in dentistry.
shows that dose reduction and improvement of access to information both work in the same way and result in a more favorable dose-information ratio.
|
|
THE TECHNOLOGY BEHIND DIGITAL RADIOGRAPHY |
|---|
There are two more advanced technologies that create digital images without an analog precursor: "direct" digital images and "semidirect" digital images. Direct digital images are acquired using a solid-state sensor. The solid-state sensors are based on charge-coupled device (CCD)- or complementary metal oxide semiconductor (CMOS)-based chips. Semidirect images are obtained using a phosphor plate system.
Solid-state systems. CCD technology. Most solid-state sensors in digital dental radiography are based on CCD technology. CCD refers to the design of the electronic chip that is used to capture the radiographic image. The chip converts into an electronic signal the energy of X-ray photons hitting the sensor. To increase the efficiency of this conversion, a scintillation layer is placed on top of the CCD. The scintillation layer converts X-ray photons into light photons, which then are absorbed by the CCD chip and converted into the electronic signal. This signal is sent to the computer by means of a cord between the sensor and the computer.
CMOS technology. Some solid-state sensors use CMOS technology. CMOS is not different from CCD technology in principle, but it does differ in terms of chip microarchitecture. In a CMOS chip, more of the electronic components controlling the conversion of photon energy into the electronic signal are incorporated into the chip itself. This simplifies the manufacturing process and, thus, reduces the costs of production. This architecture gave CMOS an advantage over CCD for some time. However, most digital cameras on the consumer market are based on CCD technology. The quality and production costs of CCD chips have benefited from this advantage of scale, and nowadays the image quality levels of CCD- and CMOS-based sensor systems are comparable.
Storage phosphor plate systems. Systems using photo-stimulable phosphor (PSP)—sometimes also referred to as "storage phosphor plate"—systems use a plate covered with phosphor crystals. This phosphor layer is able to store the energy of the X-ray photons for some time. A scanner is required to "read" the image information from the plate, which it does by scanning the plate with a laser beam of near-red wavelengths. The energy is released from the phosphor layer, detected by an image intensifier and subsequently converted into digital image information. The latent image will remain in the phosphor plate before the scanning phase for minutes to hours, depending on the environment in which the plates are stored. They should not be exposed to bright light or warmth because this will release the energy before it is read by the scanner. After the plates are scanned, they are exposed to bright light that erases all remaining energy; the plates then can be used again.
|
CHOOSING A SYSTEM |
|---|
CCD and CMOS systems. CCD and CMOS systems are connected to the computer by a cord. The connection can be via a wall box and the computer network. There are wireless systems on the market now, but these sensors are somewhat thicker than are systems with cords. The sensors are rigid and do not bend in the mouth of the patient and therefore could be more difficult for children to tolerate. The cord that connects the sensor to the computer may complicate the manipulations inside the patient’s mouth. The image produced by a solid-state sensor is available on the computer screen within a few seconds. As a result, such systems could be efficient aids in carrying out treatment such as an endodontic procedure, in which a second image easily can be made from a slightly different angle—for example, to make the second root canal better visible—with the sensor still in the same position. This could make a solid-state sensor the preferred sensor in a practice that focuses on endodontic procedures.
PSP systems. PSP systems are flexible to some extent, which will make it more comfortable for the patient. The plates require the extra step of the scanning procedure. The scanning time is somewhere between eight seconds for a single image and two minutes for a series of plates. Some systems take the same amount of time for the scanning of a series of up to eight plates, irrespective of the actual number of plates. The advantage of PSP is that a single scanner can be shared among several operatories, similar to arrangements for a conventional film processor.
Extraoral systems. Extraoral systems, including for panoramic and cephalographic radiograph machines, exist as well in digital versions. Both CCD-based and PSP-based systems are on the market. The advantage of CCD-based systems is the direct availability of the image. Some PSP-based systems feature the ability to scan both intraoral and extraoral plates; for others, a special scanner is required to do this. For computed tomography (CT), a PSP-system still is the only option because of the way in which the tomographic image is produced.
|
DOSE REDUCTION |
|---|
Neither is the dose reduction realized by digital radiography clear with respect to intraoral imaging.1 Phosphor plate systems have broad exposure latitude. This can be an advantage, because exposure conditions are not critical with respect to image quality. This implies that an overexposed image looks good even if the patient has received a higher dose than is necessary for a diagnostically useful image. Figure 2 shows that phosphor plate systems can provide good images even when the exposure time has been much greater than that required for film-based imaging. That means, however, that the user is not warned by an unsatisfactory image quality when the exposure time is too lengthy. CCD sensors, on the other hand, have a small exposure range. Overexposed images show completely saturated black areas, which give a clear indication to the user that the exposure time should be reduced. CMOS systems do not provide this clear indication but instead produce a darker image, comparable with what happens with film-based images, thus giving a signal that the exposure could be reduced.
|
|
WHAT IS A DIGITAL IMAGE? |
|---|
For a better understanding of the mechanism of image processing, it is good to know what a digital image actually is. Solid-state sensors and phosphor plate sensors in principle are not different with respect to the end result of the image acquisition process. The sensor system measures the photon intensity of the X-ray beam after it has passed through the object (the patient). These measurements are done in a two-dimensional array of small regions of 20 to 30 square micrometers, called "pixels" (the abbreviation of "picture elements") (Figure 3). The photon intensity is measured electronically on a scale of 256 gray values (0–255). Zero on this scale means that the maximum radiation is measured, which corresponds to black in the radiographic image, and 255 represents no radiation at all, or complete radiopacity (white). The measurements of the photon intensities for each pixel are sent to the computer and stored as an array of numbers representing the x and y coordinates and the photon intensity of each pixel. In fact, the digital image can be conceived as a table with columns and rows. The columns represent the x coordinates of the pixels and the rows the y coordinates. The values in each cell indicate the gray level of the pixel represented by that cell (Figure 3).
|
The table also can be used to perform image processing: a mathematical procedure is applied to the numerical representation of the digital image which results in a new set of pixel values. The resulting set of numbers then is used to display the processed image on the monitor screen (Figure 4). The algorithm between input and output image can be simple or complicated. A simple algorithm, for instance, is the reversal of the order of the gray values, resulting in a negative of the original image. The example of Figure 4 shows the enhancement of contours, which then are recognized more easily by the human visual system than are small contrasts. More advanced algorithms can be thought of as well—resulting, for instance, in a three-dimensional reconstruction of radiographic information or the automated recognition of image features.
|
|
EXAMPLES OF IMAGE PROCESSING |
|---|
). This is not an excuse for not adhering to correct exposure settings, but it can help prevent at least some retakes when the exposure settings turn out to be incorrect after the image has been taken.
|
On-screen measurements of digital images. Measurements of length, angle and area can be made on a digital image. The easiest way is to express the measurement as the number of pixels; however, a more convenient method is to use millimeters or inches as the unit of measurement. Calibration of the magnification factor of a particular sensor is needed to convert pixel measurements in real length measurements. This usually is done by measuring the known length of an endodontic file that is included in the image and calculating the magnification factor from the ratio of the known file length and the measured length.
Some software programs are able to recognize the sensor system that was used to acquire the image and, subsequently, the pixel size and inherent image magnification. This information then is used to convert pixel measurements directly into real distances, without the use of a calibration tool. One should be aware of the fact that the latter method does not compensate for magnification caused by projective distortion; the calibration method using, for instance, an endodontic file does this to some extent.
Subtraction radiography. More advanced image processing is available to dental practitioners as well. Subtraction radiography originally was described by Ziedses des Plantes4 in the 1930s. Webber and colleagues5 and Gröndahl and colleagues6 introduced digital subtraction into dental radiography. Subtraction radiography is able to show small differences between radiographs taken a specified time apart. Subtraction of the pixel values of two images with the same projection geometry shows the differences and suppresses the structures of both images that are similar, which also is called "anatomical noise."
The two images used for subtraction radiography must have identical projection geometry. Otherwise, the subtraction image shows the difference of the projection geometry instead of the difference between the first and the second exposure. Several reports have been published showing how to obtain identical images, based mainly on the use of film holders connected to individual bite-blocks. Recently, however, software solutions have been presented to reconstruct the second image according to the projection geometry of the first image, thus making the use of rigid film holders unnecessary; the software also can handle images obtained from different sensor systems, a situation that results in the images’ having different dimensional and magnification factors7 (Figure 6). These solutions facilitate subtraction radiography in general practice, even without dedicated film holders.
|
|
SECURITY ASPECTS OF DIGITAL RADIOGRAPHY |
|---|
Some companies have introduced the concept of using watermarks within their images when altered. This most likely would be a deterrent for those altering the images for fraudulent purposes. This security measure will not replace the principle that it is always the image in its original form that is stored in the database, but it will be an effective fraud protection when images are exported from the database and transferred to third parties.
It is interesting to see that most articles about fraudulent use of digital images are about how to change the image, not about how to detect an image that has been tampered with. The key issue is that insurance companies and dentists who receive images of referred patients should be aware of the possibilities of modifications in digital radiographs. In case of doubt, modifications can be recognized using standard image-processing procedures such as the measurement of the local noise level and irregularities of global contrast gradients. This, however, requires an understanding of the standard characteristics of a digital image. When the reliability of an image is at issue, the practitioner should consult an expert in digital radiography. One also should appreciate that it usually is not a single radiograph that is the basis of the suspicion; in most situations, there is more evidence in the patient history and the treatment records that makes the case questionable.
In conclusion, there is no reason to consider digital radiographs less reliable than their analog counterparts when the source of the image can be shown, and when the image can be traced back to the original database.9
|
IMPLEMENTATION OF DIGITAL RADIOGRAPHY IN DENTAL PRACTICE |
|---|
More importantly, the use of computers forces common procedures to be performed in a much more standardized way. Image acquisition takes place according to a strict protocol of consecutive steps, beginning with the selection of the name of the patient from whom the radiographs are going to be acquired from the imaging software’s database or from the patient management system. That means that no radiographic image can be taken unless a patient is selected and that a radiograph that is taken always is connected with a patient. This makes the work flow seemingly less flexible. However, the advantage is a more secure and reliable work flow, especially for the process of archiving and retrieving radiographic images.
An important aspect is the use of sensor systems from different manufacturers. Usually, the sensor is installed with the corresponding software and images are stored in the database provided by this software. This sounds logical, but in the long term it is not. Sooner or later sensors have to be replaced, or a sensor system from another manufacturer is added. It makes sense, therefore, to archive all images in a single database that is independent of a particular manufacturer. An essential requirement is that this kind of software is able to support sensor systems made by a large number of manufacturers. Especially in larger clinics and dental schools, sensor systems of different manufacturers will be used. When other people in the clinic or in the school need access to the image data, a single central database will provide this facility in a reliable and efficient way.
Diagnostic image quality. The quality of an image cannot be judged on the basis of a subjective assessment of what the image looks like. A "pretty" image is not always a good image when it relates to diagnostic imaging. The diagnostic image quality is the crucial parameter determining if an image is "good" or "not good." Defining image quality is a complicated process because the image as such is part of a longer chain of procedures and actions, beginning with the sensor system used to acquire the image and finally resulting in the clinician’s diagnostic decision. Therefore, in studies of the quality of images produced by a particular sensor system, researchers usually measure the quality by displaying the images to a group of observers who have to perform a specific diagnostic task. They then compare the observations statistically with those of observations of one or more other sensor systems and of conventional film-based imaging.
Many studies have been published about a variety of diagnostic tasks performed on digital radiographic images: the assessment of caries in different degrees, the estimation of periodontal bone loss, the recognition of the length of an endodontic file with respect to the root length or the presence of periapical radiolucencies.
The majority of the studies conclude that the diagnostic quality of digital images certainly is adequate; digital images perform at least as well as conventional radiographs and sometimes even better.11–14 This is valid for both intraoral and extraoral radiography.15 CMOS sensors produce images of the same quality as those acquired with CCD sensors.16 There often is no significant difference in image quality of the various sensor systems on the market today.17,18 The effect of image enhancement is still questionable, though sometimes observers feel more comfortable when interpreting digitally enhanced images as compared with unenhanced images.19
When digital images are printed on photographic paper, the diagnostic quality can be comparable with that of images on traditional film.20 It usually is recommended, however, that the clinician view images on a monitor screen, instead of printed on paper.
|
CONCLUSION AND FUTURE DIRECTIONS |
|---|
Three-dimensional reconstruction and rendering of radiographic image data has been introduced recently in the form of cone beam CT and local CT. Both cone beam and local CT have imaging geometry comparable with that of conventional CT, but offer a higher resolution with a much lower radiation dose to the patient.
Digital radiography no longer is an experimental modality. It is a reliable and versatile technique that will improve the diagnostic possibilities of radiography in dentistry.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
TOP
ABSTRACT
THE TECHNOLOGY BEHIND DIGITAL...
