Digital Imaging Modalities for Dental Practice

Bull. Tokyo dent. Coll., Vol. 42, No. 1, pp.1ϳ14, February, 2001 1

Review Article

DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE

MAMORU WAKOH and KINYA KUROYANAGI

Department of Oral and Maxillofacial Radiology, Tokyo Dental College, 1-2-2 Masago, Mihama-ku, Chiba 261-8502, Japan

Received 8 December, 2000/Accepted for Publication 9 February, 2001

Abstract The introduction of the computed tomograph in the 1970s revolutionized medical diagnosis by initiating the transition from analogue to digital imaging. During this period, more specialized equipment for image processing was developed, such as cathode- ray tubes for image display, special sensors for image acquisition, and storage devices for image archiving. Digital imaging systems designed exclusively for use in dentistry were developed in the latter half of the 1980s. Some are now being clinically applied under conditions of close scrutiny to determine diagnostic accuracy, image quality, and radiation exposure to patients. This article reviews the enabling technologies of digital systems used in dentistry, and focuses upon intraoral digital imaging systems, concepts for digital image acquisition, and variations in radiation dose and their effects on diagnostic accuracy of caries detection. Key words: Digital image—Imaging modalities—Intraoral radiography—Dentistry

TRANSFORMATION INTO tal imaging systems. In addition, these systems DIGITAL IMAGES SUITABLE FOR have the advantages of a lower radiation dose, DENTAL APPLICATIONS image processing, image reconstruction, and teleradiology. Conventional radiographic films that pass The turning point from film-based to digi- through all the functions of the imaging tal radiography dates back to 1972 when G.N. chain, namely image acquisition, chemical Hounsfield introduced his new invention processing, transportation, storage and image called computerized transverse axial scanning 21). display, have been important for image- The first unit accommodated only the based diagnosis. However, equipment such as patient’s head, and the scan time for one slice cathode-ray tubes for image display, special was approximately 4.5 minutes with an addi- sensors for image acquisition, and storage tional 1.5 minutes for reconstruction. This devices for image archiving has become more technique evolved into the computed tomo- specialized. Thus, the independent perform- graph, which requires only seconds to com- ance of these functions is a key feature of digi- plete a full body scan and uses a helical scan

1 2 M. WAKOH and K. KUROYANAGI

Table 1 Speciﬁcation of some intraoral digital imaging systems

Pixel size The number of gray levels Systems Sensors Pixel number ␮ 2 ( m ) Input data Output data 256 256 ,4 45ן45 385ןSens-A-Ray CCD 576 256 4,096 48ן48 400ןCDR CCD 720 256 1,024 39ן39 512ןCompuRay CCD 768 256 256 ,4 63ן63 288ןVIXA CCD 384 256 1,024 70ן70 416ןDigoRA IP 560 256 65,536 42ן42 733ןDenOptix IP 969 85ן85 367ן485

with a multi-detector. Digital imaging has with PSP applied to a CR (Computed Radio- been applied to dentistry since the 1980s graphy) System (Fuji Medical System, Co., to diagnose paranasal sinus and temporo- Tokyo, Japan) has been used for extraoral mandibular joint diseases, as well as oral and projections and image analysis including den- maxillofacial trauma. Advances in magnetic tal panoramic radiography26–28). In 1993, resonance imaging, ultrasonography, and McDavid et al.32,33) described a prototype digi- positron-emission tomography also propelled tal panoramic system that uses a solid-state radiology towards digital image interpreta- linear array of photodiodes. Arai et al.1) were tion and diagnosis. the earliest Japanese pioneers to investigate The first intraoral direct digital imaging digital panoramic dental radiography using system, RadioVisioGraphy (RVG) (Trophy a camera system. At least five commercially Radiologie, Vincennes, France), was intro- available systems for CCD-based panoramic duced into dental diagnosis in 1989 as a radiography are currently on the market. The replacement for film-based intraoral radio- DigiPan Universal (Trophy Radiology, Marne- graphy. It included a charge-coupled device la-Vallee, France) and the Orthophos DS (CCD) sensor and processing unit with (Sirona AG, Bensheim, Germany) were previ- cathode-ray tubes for image display2,20,36,50). ously discussed in relationship to radiation The first photostimulable phosphor (PSP)- dose reduction6,11). based intraoral system, DigoRA (Orion Co./ This article mainly reviews the enabling Soredex, Helsinki, Finland), became com- technologies of intraoral digital imaging sys- mercially available in 1994. These technolo- tems, namely concepts for digital image gies have undergone considerable improve- acquisition, variations in radiation dose, and ment, and many other systems have emerged the effects of dose on the diagnostic accuracy that include novel CCD9,34,35,41,49), PSP4,10,16) and of caries detection. complementary metal oxide semiconductor (CMOS)45 designs. Sensors with dimensions suitable for use in the mouth have been devel- CONCEPTS FOR oped due to progress in miniaturization of DIGITAL IMAGE ACQUISITION electronic circuitry. Extraoral digital imaging using analog sil- Dental digital images are acquired by sev- ver halide film has also progressed. Extraoral eral means. These include indirect imaging systems for dental purposes were originally from conventional radiographs using a flat- based mostly on PSP or CCD systems. In 1985, bed scanner, semidirect imaging using a PSP PSP-based panoramic radiography appeared detector and direct imaging based on a solid- with the advantages of reduced exposure state electric detector, such as CCD or CMOS and image enhancement25). This technology sensors. Table 1 shows some of the intraoral DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 3

Fig. 1 Examples of a CCD-based and a PSP-based system a) CCD sensor (CDR ) b) Imaging plate and laser scanner (DigoRA) c) Personal computer accommodating the processing unit, the digital interface card, and the image processing software, and CRT display

digital imaging systems that are currently extraoral types of radiography for dental available. Fig. 1 shows examples of systems applications are PSP-based. In 1975, a general with a CCD sensor or with an imaging plate method of using scanning optics to release (IP) and laser scanner. energy from a storage phosphor and of digi- tizing the information was patented, and the 1. Indirect digital image acquisition first technical paper describing this method- Dental diseases were diagnosed before the ology was published in 198353). In this system, advent of PSP and CCD sensors using digi- image acquisition, chemical processing, trans- tized images of conventional, intraoral or portation and storage, as well as image display panoramic films obtained from flatbed scans are performed independently of each other. of transparencies10,36). The resolution of digi- A storage phosphor material has to satisfy tized radiographs is generally from 150 to the following conditions. First, the compound 900 dpi. When the resolution is too high, the must create and store the latent image with- individual silver grains of the X-ray film out appreciable degradation until it is ready become visible, and the quality of such digital to be scanned by laser. Second, the phosphor images is limited10). must be sufficiently simulated by light that the stored energy in the latent image can be 2. Semidirect digital image acquisition released. Third, the released energy must Semidirect digital systems including intra- have a wavelength that can be readily oral, panoramic, cephalometric and other detected in the presence of the stimulating 4 M. WAKOH and K. KUROYANAGI

Fig. 2 Scheme of direct digital image acquisition using two types of CCD-based systems

light. The most popular type is europium- tion of 12 lp/mm16). This value is almost equal doped barium fluorohalide, which acts as an to that of CCD-based systems. activator to create luminescence centers. The fluorohalide is mixed with a polymer that 3. Direct digital image acquisition binds the storage phosphor crystals to a base. Direct digital imaging can be based on The phosphor is covered with a protective solid-state electric detectors such as CCD or coating. When the phosphor is exposed to X- CMOS sensors. Direct sensor systems include radiation, an electron of the europium ion is a CCD or CMOS sensor, a processing unit, a excited to the conduction band and trapped digital interface card, computer, and soft- in the crystalline lattice of the storage phos- ware. Some systems are supplied with a dedi- phor, where it causes luminescence. When cated timer so that X-ray production and the excited fluorohalide complex is stimu- image acquisition can be synchronized. These lated by red and green laser lights, fluo- systems are hard-wired to the X-ray equip- rescent light proportional to the X-ray dose ment, and the sensor automatically starts absorbed is generated and converted into an image acquisition when it detects an increase amplified electric signal. This analog signal in the radiation level. There are two major is converted to digital form by an analog-to- types of CCD used in intraoral radiography: digital (A/D) converter. In other words, ana- those that use a scintillation screen to expose log signals are converted into a fixed number the CCD chip, and those that use an image of values, depending on the number of bits sensor that is directly exposed to X-radiation that are used for the purpose. Intraoral sys- (Fig. 2). The first group of these systems uses tems usually store images as 8-bit data. In gen- an optical coupler, either a fiber- optic or a eral, thin phosphor layers achieve better spa- series of lenses, to transfer light from a scintil- tial resolution and sharpness than thick lay- lation screen to a CCD that is not resistant ers, but current intraoral systems have a reso- to frequent direct exposure to X-radiation, lution of about 6 to 8 lp/mm, except for the resulting in image distortion and image DenOptix system, which has a spatial resolu- density2,36,41,49). The second group of these sys- DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 5 tems uses “hardened” CCDs which are more displayed with 256 shades. If 256 shades of sensitive and resistant to X-radiation. This gray are to be represented, the signal from imaging process results in decreased optic each pixel is converted to the appropriate distortion, but ground haze or noise presents pixel value from 0 to 255. a problem20,50). Another method of direct digital imaging The CCD sensor was first developed at the uses CMOS-based sensors. Compared with Bell Laboratories of AT&T for video applica- CCD-based sensors, one advantage of CMOS tions in the late 1970s. These sensors consist technology is the integration of control cir- of a thin wafer of silicon crystals arranged cuitry, including the A/D converter, directly in a grid. The limited number of pixels that into the sensor. The image output is from can be grouped together in the CCD sensor individual pixels and is analog, but it is con- restricts the image resolution. The smaller verted to digital for display on a monitor. the pixel size, the better the resolution. The Another advantage is that blooming, which sensors of the intraoral digital system have resembles allowing too much light through a been improved to pixel sizes of 20␮m from viewbox and which occurs in CCD sensors up to 70␮m. This development has greatly due to excess charge leakage into adjacent improved the spatial resolution of CCD sen- pixels, can be avoided. This type of sensor has sors, because the resolution range, which gen- been applied to the most recent Computed erates images equivalent to conventional film- Dental Radiography (CDR) (Schick Technol- based images, is similar to that of the human ogies, Inc., NY, USA) digital intraoral digital eye (6 to 10 lp/mm). As a result, the diag- system, but it has not yet been incorporated nostic image quality is acceptable for clinical into panoramic and cephalometric imaging practice. Incidentally, the pixel size of CCD- equipment. based panoramic system is generally around 100␮m. Direct digital systems with CCD sensors can RADIATION DOSE REDUCTION perform real time imaging; an image is displayed on a monitor within a few seconds. Radiation safety is an important issue in Charge-coupled devices consisting of silicon dental radiography. The desired amount of atoms covalently bonded to adjacent atoms information must be obtained by exposing constitute arrays of X-ray- or light-sensitive patients to a minimal amount of radiation. pixels. In other words, CCDs consist of large Digital intraoral and panoramic radiography numbers of photoelectric cells. Exposure to generally requires a lower dose per exposure X-rays or light breaks the bonds and the photo- than does conventional dental radiography. electric cells generate voltage in proportion Many investigators have compared exposure to the amount of emission (radiation) energy. doses among digital intraoral, conventional The amount of energy required to break the intraoral film and digital panoramic radio- bond is approximately 1.1 V. The CCD charge graphy in vitro6,11). Table 2 shows the results is read out by transferring the collected from some papers on intraoral radiography. charge in each pixel, then it is destroyed, Task-dependent dose reductions of up to which results in a latent electric image being 94% with enhancement for individual expo- distributed as a pattern of charges on a sures8) and 80–90% for standard exposures8,52) matrix. This analog radiographic information have been reported from studies of early RVG is converted to a digital form to be stored in systems that used CCD sensors. In our study of a computer. An analog-to-digital converter the RVG-S system, the entry exposure dose for transforms the original continuous density optimal imaging required from one-half to range into a set of discrete gray levels. Images one-third the exposure time of E-speed film, generally acquired as 10 to 16 bits in most or one-quarter to one-sixth of that needed systems are stored as 8 bits (per pixel) and for D-speed film. The ratios of dose reduc- 6 M. WAKOH and K. KUROYANAGI

Table 2 Exposure reduction in intraoral digital imaging systems compared with dental X-ray ﬁlms Table 2 (The percentage number in the digital system indicates the exposure rate required when Table 2 X-ray ﬁlm required 100%)

Dental film Digital system Authors D-speed E-speed RVG series VIX series SAR Flash Dent CDR DIG Soh52) 100% — 22.30% — — — — — Sanderink48) — 100% 40% 30% 40% 40%–80% — — Wakoh62) — 100% 35–50% — — — — — Farman10) 100% 20% 40% 50% 50% 20% — Brettle4) — 100% — — — — — 10–20% Huysmans22) — 100% — — — — — 6% Wakoh63) — 100% — — — — 19% — RVG: RadioVisioGraphy CCD system, VIX: Visualix/VIXA CCD system, SAR: Sens-A-Ray CCD system, FD: Flash Dent CCD system, CDR: Computed Dental Radiography CCD system, DIG: Digora PSP system magnified :ן tions in various regions of the mouth ranged an appropriate image quality can be main- from 50–65% with respect to E-speed film and tained at only 6% of the exposure time 73–76% with respect to D-speed film62). These needed for E-speed film3,22). findings proved that CCD sensor systems CCD-based systems with a narrower dynamic reduced doses by 50–75% of those required range also have to contend with deterio- for conventional film based methods. The task- rated radiation reduction. The smaller sensi- dependent dose reduction in digital imaging tive area of most intraoral CCD sensors may system strongly relates to the dynamic range require more exposure per investigation, of the system. increasing the total dose20,48,56). Because the The dynamic range is narrower in CCD- smaller sensitive area and possible discomfort than in PSP-based systems with a high calibra- associated with the CCD sensor make it diffi- tion exposure and conventional film. Accord- cult to position, retakes accompanied by ing to some authors, that of CCD-based digital added exposure and re-exposure for the intraoral radiographic systems reaches a max- patient may be needed. According to a study imum of 12␮C/kg, although it is slightly by Versteeg et al.60), 28% of the sensor images changed by the tube voltage setting15,43,59,64). obtained from the incisal to the molar region Our study of the RVG-S system found that the were unacceptable. Similar results were found dynamic range was 8.6 times narrower than in a initial study of the RVG system in which that for conventional films62). The smaller, but the repeat exposure rate was 25%20). However, acceptable exposure range results from the assuming that all 25–28% of the original fact that CCD sensors are more sensitive to X- images that are retaken are acceptable, this is radiation and reach full saturation at a lower still a dose reduction compared with film. A exposure than conventional X-ray film. Addi- wider dynamic range in PSP-based systems tionally, considering reports that the informa- means a large selection range of exposure to tion required to make a clinical diagnosis us- X-radiation; conventional film or timely PSP ing a CCD system is similar to that of film, may represent more information but induce these systems satisfy the ALARA (as low as overexposures. PSP-based systems accommo- reasonably achievable) concept12). In contrast, date an automatic range control (ARC) mech- PSP-based systems including DigoRA and anism that displays constant image density DenOptix offered a maximal responsible expo- without being relative to X-ray exposure. The sure of over 40␮C/kg under high calibration ability to optimize storage-phosphor images exposure44). The wide dynamic range of PSP- means that exposure levels can be low relative based system has been proven by the fact that to the film. Only a small portion of this wide DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 7 range, however, is used to acquire an optimal diagnosis with digital imaging were con- image for a clinical diagnosis. From the view- ducted using the indirect method65). This point of radiation protection, whether an report proved that the sensitivity of caries ARC should be utilized when a dental clini- detection was higher in digitized film images cian gives priority to ALARA concept over than in either xeroradiographs or film radio- image quality is doubtful. graphs. In addition, quantitative measure- Perhaps, in the near future, a specific dose ments of caries depth assessed on digitized within a particular dynamic range will be stan- radiographs were correlated with the histo- dardized for the diagnosis of specific struc- logic depth. Despite these close correlations, tures and locations. In addition, the area of no evidence indicated that digital imaging sensitivity and the thickness of the receptor underestimated lesion depth when compared may more closely resemble those of tradi- with whitish decalcified histologic zones. tional film. If these improvements come Since the first intraoral direct digital imaging to pass, then intraoral digital imaging will system, RadioVisioGraphy, was introduced in be established as the modality, because radia- 198937), several investigators have studied car- tion dose will be reduced, image enhance- ies diagnosis using this type of system. Most of ment tools will change contrast and density, these studies examined the detectability of and wet processing using chemical will be these systems compared with that of conven- unnecessary. tional film. When a new diagnostic modality Some investigators have studied dose reduc- to evaluate dental caries is released, labora- tion using digital panoramic radiography with tory experiments should initially confirm that a human phantom. The DigiPan and the the new system is suitable for application to Orthophos DS receptors based on a solid-state clinical practice. Studies of the first genera- sensor produce satisfactory images with entry tion of CCD sensors did not identify any sig- exposure savings of approximately 70% and nificant differences between direct digital up to 43%, respectively, when compared with systems and digitized film66), but paper-print a conventional film/rare earth screen combi- images of this first direct digital system were nation6,11). A 60% reduction in dose is also less accurate than film47). Since then, CCD claimed by the manufacturer of the DIMAX sensors have been substantially improved, system. and the first dental PSP-based system, DigoRA, became commercially available in 1994. No statistically significant differences were identi- DIAGNOSTIC STUDY OF CARIES fied among three CCD-based systems and DigoRA in the detection of occlusal and proxi- In previous studies, the accuracy for caries mal caries lesions69). The digital systems did diagnoses for proximal5,17,18,23,63,69), occlusal car- not perform differently from the various film ies17,18,69) using intraoral digital systems has types evaluated by the same observers using been assessed. The root canal length has been the same teeth17,19). A more recent CCD system endodontically assessed7,48), and periodontal has also proven comparable to film for inter- bone changes14,31,61) have been examined. In preting proximal caries72). In primary teeth, particular, conventional film radiography, the PSP-based system storage did not perform xeroradiography, indirect and direct (semi- differently from film in the detection of cavi- direct) digital imaging have promoted radio- tated proximal surfaces42). In 1997, we also graphic caries diagnosis. Establishment of the compared the diagnostic accuracy of proxi- validity of digital imaging for caries diagnosis mal caries using a first generation CDR system required systematic research into caries. with that of conventional dental X-ray film images using receiver operating characteris- 1. Laboratory experiments tics (ROC) analysis. The ROC curve gives the The first studies on the accuracy of caries overall performance of a diagnostic system 8 M. WAKOH and K. KUROYANAGI

Caries defect

Fig. 3a) Experimental dentition with simulated proximal carious defect and the obtained image with the defect at distal surfaces of the second premolar tooth (see arrow) using a CDR system.

et al. criticized the relevance to diagnostic image precision of natural dental caries when pseudo- or mechanical caries were generated by using a round bur on extracted teeth. Kang et al.23,24) reported that mechanically-produced proximal defects are more easily detected than natural proximal caries because the bound- aries between the mechanically-produced defect and the tooth surface are more dis- tinct. The defects were therefore uniform because of this clear boundary. This situation results in a large difference in image contrast, increased detection rate, and improved diagnostic accuracy. Consequently, their study strengthened the support for Wenzels’ view- point. Wenzel et al. stated that many other Fig. 3b) ROC curves for intraoral digital imaging system (CDR) (original image), Ultraspeed and experimental designs have been suggested, EKTASPEED Plus dental X-ray film images. ranging from the use of another observer or the clinical state of the tooth surface as validation, but such methods should be aban- and sensitivity at any specificity. We mechani- doned. In addition, they stated that the out- cally drilled the proximal surfaces of an ex- come of a new diagnostic method must be tracted tooth and then plugged the defects held against the true diagnosis that is obtained with dentine powder to simulate proximal by a validation or reference method, also caries with low contrast (Fig. 3a). The ROC called a “gold standard”, to test accuracy68). curve for the CDR image was similar to those That is, only one method can fully resolve this of conventional dental X-ray films; the diag- problem: cutting the tooth in sections to visu- nostic accuracy of proximal caries was similar, alize demineralization as a discolored area in and this result was also similar to those of the dental tissues under a microscope. This is previous studies (Fig. 3b, Table 3). However, the only validation method that is currently a problem with the theoretical method of reliable enough for evaluating the efficacy of detecting caries was pointed out24,46). Wenzel a new caries diagnostic procedure71). DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 9

Table 3 Diagnostic accuracy of proximal caries detection: comparison of X-ray film, digitized film, CCD and PSP systems including image magnification processing (The numbers with no units indicate the mean areas under the ROC curves for each modality. When the number is close to 1, the accuracy of caries detection is high)

Dental ﬁlm Digitized ﬁlm Digital system

Authors Sample Cause CDR DIG Results ן ן ן -D- E speed speed 3 6 18 RVG VIX SAR 4ן .orig 4ן 2ן .orig

Hintze17) Extracted teeth natural 0.61 0.61 0.59 0.5 ᕃ Møystad37) Extracted teeth natural 0.7 0.71 0.59 ᕄ Wenzel69) Extracted teeth natural 0.57 0.6 0.6 0.61 ᕅ Extracted teeth natural 0.74 Hintze18) ᕆ in vivo natural 0.7 Extracted teeth natural 67% Kang24) ᕇ Extracted teeth mechanical 74% Møystad38) Extracted teeth natural ᕈ Svanaes54) Extracted teeth natural 0.73 0.72 ᕉ Wakoh63) Extracted teeth mechanical 0.77 0.75 0.75 0.77 0.66 ᕊ RVG: RadioVisioGraphy CCD system, VIX: Visualix/VIXA CCD system, SAR: Sens-A-Ray CCD system, CDR: Computed Dental Radiography CCD system, DIG: Digora PSP system magnified :ן Ͻ Results ᕃ : CCDsϭDental film ᕇ : mechanicalϾnatural Results ᕄ : excessive magnifiedϽmagnified ᕈ : Enhanced PSPϾDental film Dental filmסCCDs ᕉ : magnified PSPסResults ᕅ : PSP in vitro ᕊ : excessive magnified CCDϽDental filmסResults ᕆ : in vivo

Validation of the methodology of data 2. Clinical study analysis may also be required. When no valid The preceding validation method cannot expression of the true state of the disease can be applied to clinical studies of caries diag- be obtained, a study of precision become nosis, because teeth in situ cannot be sec- appropriate. Several methodologies can evalu- tioned. However, a simulation of the in vivo ate diagnostic tests13,30,73). Currently, Kappa sta- situation can often be achieved with an in vitro tistics that determine the sensitivity, specific- model. A recent study has compared labora- ity, and predictive values for the test and ROC tory and clinical results of occlusal and approx- analysis can evaluate precision and accuracy, imal caries diagnosis in the same teeth, com- respectively. Naitoh et al.40) reported that the paring pre- with post-extraction teeth. That precision of caries detection in terms of k val- study found no significant difference in diag- ues did not differ between one new digital nostic accuracy between the two approaches18). system and X-ray film. ROC analysis has dem- Therefore, laboratory and clinical diagnosis onstrated that the accuracy of caries diagnosis may closely correspond. Nevertheless, no is not significantly different when a discrete diagnosis method for caries is yet sufficiently rating scale is used57,58). accurate to act as a gold standard for the clini- In general, intraoral digital dental imaging cal evaluation of new tests. Clinical studies systems seem to be as accurate as current should focus rather on parameters other dental X-ray films for detecting caries except than accuracy, such as diagnostic precision for only in a very few in vitro studies. Intra- or reproducibility, intraobserver and inter- oral radiographic modalities including direct observer variation, the consequence to the (semidirect) digital systems and conventional patient of excluding or establishing a diagno- X-ray film can detect occlusal and approximal sis, impact on treatment strategy and progno- dentinal caries lesions, but they seem unable sis, and the radiation dose. Considering that to efficiently detect initial enamel caries these factors constitute the ultimate evalua- lesions. tion of any new diagnostic method, no studies 10 M. WAKOH and K. KUROYANAGI

Fig. 4 Effect for caries diagnosis by the combination of the brightness and image contrast on CRT display 0 at horizontal axis means a default setting of image contrast; the background on monitor is visually a higher density at a minus setting than at the default setting. The bright number also influ- ences the density; the background on the monitor is visually a higher density at a plus setting than at a minus setting. appear to have investigated the effects of digi- ment of digitized images of low density radio- tal imaging on overall clinical procedures. graphs, sensitivity increased by approximately 20% without an increase in the number of 3. Application of image processing tools false positive scores67). Other studies devel- Digital imaging systems have several advan- oped a task-dependent algorithm for approx- tages in terms of image processing tools com- imal caries; accuracy was higher in enhanced pared with the film-based imaging. Digital PSP-based digital images than in unenhanced radiography is often considered as a replace- digital or film radiographs38). One study con- ment for conventional radiography, and this cluded that enhancement may be beneficial is true to some extent. However, digital radio- for some observers but that selecting the graphy offers options that are not attainable proper procedure is time consuming51). We with film-based radiography. In this sense, also identified an acceptable range of image digital radiography is a new method of diag- contrast for the detection of enamel defects nostic imaging. The lower dose with restric- by manipulating the contrast and brightness tions on X-ray exposure and the gain of time of the CRT display; detection was highly for image acquisition are key advantages. dependent on contrast (Fig. 4)29). Therefore, Another unique advantage of digital imaging if observers find the initial images to be of is imaging processing. Caries have been diag- suboptimal density and contrast, they should nosed using image enhancement inclusive of be enhanced. The effect of zooming to contrast29,38,51,67), edge enhancement38,54), zoom- display various images sizes on the monitor ing37,54,63), image compression70) and image syn- has also been investigated. Møystad et al.37) thesis39,55). Contrast enhancement seems to be reported that the enlargement of digitized most important prerequisite for predicting film radiographs to examine bite-wing radiog- caries visibility, although the spatial resolution raphy has an upper limit beyond which diag- is lower in most digital systems than in the nostic accuracy may be reduced. Our study conventional dental X-ray film. In one study using a CCD-based system also suggested that addressing the effects of contrast enhance- magnification has limits in terms of detecting DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 11 small holes on proximal surfaces and that preventing of inferiority of image quality with excessive magnification reduces the reliability the development of storage devices. Never- of diagnosing early caries on proximal sur- theless, the gold standard remains analog faces (Table 3)63). However, the diagnostic film. Because of competition from newer digi- accuracy of the images from which informa- tal technologies, manufacturers of X-ray film tion had been lost by reducing the image size continue to improve image quality on film at using PSP-based system was comparable with lower radiation doses. Film still has a higher that of conventional film54). These results can- spatial resolution than digital imaging sys- not necessarily be explained by the spatial tems, but studies have not demonstrated that resolution of the digital sensors and conven- this is clinically important. When newer digi- tional X-ray film. They may be influenced by tal receptors are developed in terms of physi- the observation environment, especially the cal factors such as spatial resolution, signal-to- luminosity of the display-monitor, and by noise ratio, and contrast resolution, when changes in contrast resolution at the time of more advanced radiographic methods such image magnification and illumination of the as creating three-dimensional displays of viewbox. In other words, spatial resolution dental structure are established, and when may not be a significant factor in diagnosing the cost of plant and equipment is reduced, caries by radiographic means. One study diagnostic dentistry will truly enter a totally found no relationship between accuracy and digital era. image compression for occlusal surfaces; for approximal surfaces, the accuracy was pro- gressively reduced with a difference of 14% between the original and the images com- REFERENCES pressed by 20%70). Investigations regarding the use of Tuned-Aperture Computed Tomography 1) Arai, Y., Shinoda, K., Hashimoto, K. and based on the principle of tomosynthesis gen- Sairenji, E. (1992). Clinical trial of new digi- 39,55) tal panoramic tomography. Dentomaxillofac erated less explicit results . Radiol 21, 223–224. Regardless of the more limited resolution, 2) Benz, C. and Mouyen, F. (1991). Evaluation a digital imaging system seems to diagnose of a new RadioVisioGraphy system image qual- caries as well as conventional film radiogra- ity. Oral Surg Oral Med Oral Pathol 72, 627–631. phy, but its diagnostic ability may have several 3) Borg, E. and Gröndahl, H-G. (1996). On the dynamic range of different X-ray photon advantages in the future. These are primarily detectors in intra-oral radiography. Dento- the possibility for image enhancement and a maxillofac Radiol 25, 82–88. significant reduction in radiation dose to the 4) Brettle, D.S., Workman, A. and Ellwood, R.P. patient. (1996). The imaging performance of a storage phosphor system for dental radiography. Br J Radiol 69, 256–261. 5) Dove, S.B. and McDavid, W.D. (1992). A com- CONCLUSION parison of conventional intra-oral radiography and computer imaging techniques for the The movement from analog film towards detection of proximal surface dental caries. digital imaging systems is based on a number Dentomaxillofac Radiol 21,127–134. 6) Dula, K., Sanderink, G.C.H., van der Stelt, P.F., of advantages of digital imaging systems com- Mini, R. and Buser, D. (1998). Effects of dose pared with conventional film radiography: reduction on the detectability of standardized the reduction of the number of retakes radiolucent lesions in digital panoramic radi- because of the possibilities of image process- ography. Oral Surg Oral Med Oral Pathol Oral ing changing the image contrast and density, Radiol Endod 86, 227–233. 7) Ellingsen, M.A., Harrington, G.W. and Hol- the shortening of working time from expo- lender, L.G. (1995). Radiovisiography versus sure to image display due to avoiding chemi- conventional radiography for detection of cal processing and wet processing errors, the small instruments in endodontic length deter- 12 M. WAKOH and K. KUROYANAGI

mination. Part 1. In vitro evaluation. Journal of 19) Hintze, H., Christofferson, L. and Wenzel, A. Endod 21, 326–331. (1996). In vitro comparison of Kodak Ultra- 8) Farman, A.G., Mouyen, F. and Razzano, M. speed, Ektaspeed, and Ektaspeed Plus, and (1993). RadioVisioGraphy: Concept and Agfa M2 Comfort dental X-ray films for the applications. Computers in Clinical Dentistry detection of caries. Oral Surg Oral Med Oral (Preston, J.D. ed.), pp.125–137. Quintessence Pathol Oral Radiol Endod 81, 240–244. Publishing Co., Inc., Carol Stream, I1. 20) Horner, K., Shearer, A.C., Walker, A. and 9) Farman, A.G., Scarfe, W.C., Schick, D.B. and Wilson, N.H.F. (1990). Radiovisiography: An Rumack, P.M. (1995). Computed dental radi- initial evaluation. Br Dent J 168, 244–248. ography: Evaluation of a new charge-coupled 21) Hounsfield, G.N. (1972). Computerized device-based intraoral radiographic system. transverse axial scanning (tomography), Part Quintessence Int 26, 399–404. 1. Description of system. Br J Radiol 46, 1016– 10) Farman, A.G. (1995). Computed maxillofa- 1022. cial imaging: Perspectives and prospectives. 22) Huysmans, M-CDNJM, Hintze, H. and Wenzel, Computer Assisted Radiology (Lemke, H.U., A. (1997). Effect of exposure times on in vitro Inamura, K., Jaffe, C.C. and Vannier, M.W. diagnosis using the Digora system. Eur J Oral eds.), pp.949–957. Springer, Berlin, Germany. Sci 105, 15–20. 11) Farman, T.T., Farman, A.G., Kelly, M.S., 23) Kang, B.-C., Farman, A.G., Scarfe, W.C. and Firriolo, F.J., Yancey, J.M. and Stewart, A.V. Goldsmith, L.J. (1995). Mechanical defects (1998). Charge-coupled device panoramic in dental enamel vs. natural dental caries: radiography: Effect of beam energy on radia- Observer differentiation using Ektaspeed Plus tion exposure. Dentomaxillofac Radiol 27, 36– film. Caries Res 30, 156–162. 40. 24) Kang, B.-C., Farman, A.G., Scarfe, W.C. and 12) Gelsky, D.E. and Baker, C.G. (1984). The Goldsmith, L.J. (1996). Observer differentia- ALARA concept. Population exposure from tion of proximal enamel mechanical defects X-rays in dentistry—as low as reasonably versus natural proximal dental caries with achievable? J Can Dent Assoc 50, 402–403. Computed Dental Radiography. Oral Surg Oral 13) Gröndahl, H-G. (1994). Radiologic diagnosis Med Oral Pathol Oral Radiol Endod 82, 459–465. in caries management. Textbook of Clinical 25) Kashima, I., Kanno, M., Higashi, T. and Cariology 2nd ed. (Thylstrup, A. and Takano, M. (1985). Computed panoramic Fejerskov, O. eds.), pp.367–382. Munksgaard, tomography with scanning laser-stimulated Copenhagen. luminescence. Oral Surg Oral Med Oral Pathol 14) Gurgan, C., Gröndahl, K. and Wennstrom, J.L. 60, 448–453. (1995). Observer variation in the radio- 26) Kashima, I., Brando, S. and Konishi, D. graphic assessment of the bone level on the (1990). Bone trabecular pattern analysis in buccal and lingual surface of mandibular Down syndrome with use of computed pan- molars. Dentomaxillofac Radiol 24, 165–168. oramic radiography. Oral Surg Oral Med Oral 15) Hayakawa, Y., Farman, A.G., Kelly, M.S. and Pathol 70, 360–364. Kuroyanagi, K. (1995). Signal-to-noise ratio: 27) Kashima, I., Tajima, K. and Nishimura, K. Computed Dental Radiography versus Sens-A- (1990). Diagnostic imaging of diseases affect- Ray. Oral Radiol 11, 109–113. ing the mandible with the use of computed 16) Hildebolt, C.F., Vannier, M.V., and Conover, panoramic radiography. Oral Surg Oral Med G.L. (1996). Preliminary evaluation of a Oral Pathol 70, 110–116. new intraoral photostimulable phosphor plate 28) Kashima, I. (1995). Computed radiography system. Computer Assisted Radiology, (Lemke, with photostimulable phosphor in oral and H.U., Inamura, K., Jaffe, C.C. and Vannier, maxillofacial radiology. Oral Surg Oral Med Oral M.W. eds.), pp.969–974. Elsevier, Amsterdam, Pathol Oral Radiol Endod 80, 577–598. Netherlands. 29) Kitagawa, H., Wakoh, M. and Kuroyanagi, K. 17) Hintze, H., Wenzel, A. and Jones, C. (1994). (1999). Image contrast range for detection of In vitro comparison of D- and E-speed film enamel defects using a digital dental imaging radiography, RVG, and Visualix digital radi- system. Oral Radiol 15, 95–104. ography for the detection of enamel 30) Kraemer, H.C. (1992). Evaluating Medical approximal and dentinal occlusal caries Tests—Objective and Quantitative Guidelines. lesions. Caries Res 28, 363–367. Sage Publications, London. 18) Hintze, H. and Wenzel, A. (1996). Clinical 31) Matsuda, Y., Shonome, M., Seki, K., Hasegawa, and laboratory radiographic caries diagnosis: K. and Okano, T. (1996). Accuracy of intra- A study of the same teeth. Dentomaxillofac oral radiographic system (Digora) in detect- Radiol 25, 115–118. ing artificial periodontal bone defects. Dental DIGITAL IMAGING MODALITIES FOR DENTAL PRACTICE 13

Radiol 3, 22–26. (in Japanese) digital intra-oral radiographic systems on tube 32) McDavid, W.D., Dove, S.B., Welander, U. and voltage. Dentomaxillofac Radiol 28, 364–367. Tronje, G. (1992). Direct digital extraoral 44) Nishikawa, K., Mori, T., Shibuya, H., radiography of the head and neck with a solid Hayakawa, Y., Wakoh, M. and Kuroyanagi, K. state linear array. Oral Surg Oral Med Oral Pathol (1999). Dose response of two storage- 74, 811–817. phosphor systems for intra-oral radiography. 33) McDavid, W.D., Dove, S.B., Welander, U. and Oral Radiol 15, 125–133. Tronje, G. (1993). Dimensional reproduc- 45) Paurazas, S.B., Geist, J.R., Pink, F.E., Hoen, tion in direct digital rotational panoramic M.M. and Steiman, H.R. (2000). Comparison radiography. Oral Surg Oral Med Oral Pathol of diagnostic accuracy of digital imaging by 75, 523–527. using CCD and CMOS-APS sensors with E- 34) Molteni, R. (1992). Visualix, a new system for speed film in the detection of periapical bony direct dental X-ray imaging: A preliminary lesions. Oral Surg Oral Med Oral Pathol Oral report. Dentomaxillofac Radiol 21, 222–223. Radiol Endod 89, 356–362. 35) Molteni, R. (1993). Direct digital dental X- 46) Price, C. and Ergul, N. (1998). A comparison ray imaging with Visualix/VIXA. Oral Surg of a film-based and a direct digital dental Oral Med Oral Pathol 76, 235–243. radiographic system using a proximal caries 36) Mouyen, F., Benz, C., Sonnabend, E. and model. Oral Surg Oral Med Oral Pathol Oral Lodter, J.P. (1989). Presentation and physical Radiol Endod 85, 113–118. evaluation of RadioVisioGraphy. Oral Surg Oral 47) Russell, M. and Pitts, N.B. (1993). Radiovisio- Med Oral Pathol 68, 238–242. graphic diagnosis of dental caries: Initial com- 37) Møystad, A., Svanaes, D.B., Risnes, S., parison of basic mode videoprints with bite- Larheim, T.A. and Gröndahl, H-G. (1995). wing radiography. Caries Res 27, 65–70. Effect of image magnification of digitizes 48) Sanderink, G.C.H., Huiskens, R., van der Stelt, bitewing radiographs on approximal caries P.F., Welander, U.S. and Stheeman, S.E. detection: An in vitro study. Dentomaxillofac (1994). Image quality of direct digital intra- Radiol 24, 255–259. oral X-ray sensors in assessing root canal 38) Møystad, A., Svanaes, D.B., Risnes, S., length. The RadioVisioGraphy, Visualix/ Larheim, T.A. and Gröndahl, H-G. (1996). VIXA, Sens-A-Ray, and Flash Dent systems Detection of approximal caries with a storage compared with Ektaspeed films. Oral Surg Oral phosphor system. A comparison of enhanced Med Oral Pathol 78, 125–132. digital images with dental X-ray film. Dento- 49) Scarfe, W.C., Farman, A.G. and Kelly, M.S. maxillofac Radiol 25, 202–206. (1994). Flash Dent: An alternative charge- 39) Nair, M.K., Tyndall, D.A., Ludlow, J.B. and coupled device/scintillator-based direct digi- May, K. (1998). Tuned aperture computed tal intraoral radiographic system. Dento- tomography and detection of recurrent caries. maxillofac Radiol 23, 11–17. Caries Res 32, 23–30. 50) Shearer, A.C., Horner, K. and Wilson, N.H.F. 40) Naitoh, M., Yuasa, H., Toyama, M., Shiojima, (1990). RadioVisioGraphy for imaging root M., Nakamura, M., Ushida, M., Iida, H., canals: An in vitro comparison with conven- Hayashi, M. and Ariji, E. (1998). Observer tional radiography. Quintessence Int 21, 789– agreement in the detection of proximal caries 794. with direct digital intraoral radiography. Oral 51) Shrout, M.K., Russel, C.M., Potter, B.J., Powell, Surg Oral Med Oral Pathol Oral Radiol Endod 85, B.J. and Hildebot, C.F. (1996). Digital 107–112. enhancement of radiographs: Can it improve 41) Nevig, P., Wing, K. and Welander, U. (1992). caries diagnosis? J Am Dent Assoc 127, 469–473. Sens-A-Ray: A new system for direct digital 52) Soh, G., Loh, F-C. and Chong, Y-H. (1993). intraoral radiography. Oral Surg Oral Med Oral Radiation dosage of a dental imaging system. Pathol 74, 818–823. Quintessence Int 24, 189–191. 42) Nielson, L-L., Hoernoe, M. and Wenzel, A. 53) Sonoda, M., Takano, M., Miyahara, J. and (1996). Radiographic detection of cavitation Kato, H. (1983): Computed radiography uti- in approximal surfaces of primary teeth using lizing scanning laser stimulated luminescence. a digital storage phosphor system and conven- Radiology 148, 833–838. tional film, and the relationship between cavi- 54) Svanaes, D.B., Møystad, A., Risnes, S., tation and radiographic lesion depth: An in Larheim, T.A. and Gröndahl, H-G. (1996). vitro study. Int J Paediatr Dent 6, 167–172. Intra-oral storage phosphor radiography for 43) Nishikawa, K., Shibuya, H., Wakoh, M. and approximal caries detection and effect of Kuroyanagi, K. (1999). Dependency of dose image magnification—Comparison with con- response of five charge-coupled device-based ventional radiography. Oral Surg Oral Med Oral 14 M. WAKOH and K. KUROYANAGI

Pathol Oral Radiol Endod 82, 94–100. Depth of occlusal caries assessed clinically, by 55) Tyndall, D.A., Clifton, T.L., Webber, R.L., conventional film radiographs, and by digi- Ludlow, J.B. and Horton, R.A. (1997). TACT tized, processed radiographs. Caries Res 24, imaging primary caries. Oral Surg Oral Med 327–333. Oral Pathol Oral Radiol Endod 84, 214–225. 66) Wenzel, A., Hintze, H., Mikkelsen, L. and 56) Velders, X.L., Sanderink, G.C.H. and van der Mouyen, F. (1991). Radiographic detection Stelt, P.F. (1996). Dose reduction of two digi- of occlusal caries in noncavitated teeth: A tal sensor systems measuring file lengths. Oral comparison of conventional film radiographs, Surg Oral Med Oral Pathol Oral Radiol Endod 81, digitized film radiographs, and RadioVisio- 607–612. Graphy. Oral Surg Oral Med Oral Pathol 72, 621– 57) Verdonschot, E.H., Wenzel, A. and Bronk- 626. horst, E.M. (1993). Applicability of receiver 67) Wenzel, A. and Fejerskov, O. (1992). Validity operating characteristics (ROC) analysis on of diagnosis of questionable caries lesions in discrete caries depth ratings. Community Dent occlusal surfaces of extracted third molars. Oral Epidemiol 21, 269–272. Caries Res 26, 188–194. 58) Verdonschot, E.H., Wenzel, A. and Bronk- 68) Wenzel, A. and Verdonschot, E.H. (1994). horst, E.M. (1993). Assessment of diagnostic Some considerations in the evaluation of diag- accuracy in caries detection: An analysis of two nostic tests in dentistry. Dentomaxillofac Radiol methods. Community Dent Oral Epidemiol 21, 23, 179–182. 203–208. 69) Wenzel, A., Borg, E., Hintze, H. and Gröndahl, 59) Versteeg, C.H., Sanderink, G.C.H. and van der H.-G. (1995). Accuracy of caries diagnosis in Stelt, P.F. (1997). Efficacy of digital intra-oral digital images from charge-coupled device radiography in clinical dentistry. J Dent 25: and storage phosphor systems: An in vitro 215–224. study. Dentomaxillofac Radiol 24, 250–254. 60) Versteeg, C.H., Sanderink, G.C.H., van Ginkel, 70) Wenzel, A., Gotfredsen, E., Borg, E. and F.C. and van der Stelt, P.F. (1998). An evalua- Gröndahl, H-G. (1996). Impact of glossy tion of peri-apical radiography with a charge- image compression ( JPEG) on accuracy of coupled device. Dentomaxillofac Radiol 27, 97– caries detection in digital images taken with a 101. storage phosphor system. Oral Surg Oral Med 61) Visser, H., Hermann, K.-P., Pfeiffer, T., Umlauf, Oral Pathol Oral Radiol Endod 81, 351–355. J., Kruger, W. and Harder, D. (1995). An 71) Wenzel, A. and Hintze, H. (1999). Compari- anthropomorphic phantom for periodontal son of microscopy and radiography as gold radiography. Dentomaxillofac Radiol 24, 195– standards in radiographic caries diagnosis. 197. Dentomaxillofac Radiol 28, 132–136. 62) Wakoh, M., Farman, A.G., Scarfe, W.C., Kelly, 72) White, S.C. and Yoon, D.C. (1997). Compara- M.S. and Kuroyanagi, K. (1994). Radiation tive performance of digital and conventional exposures with the RVG-S and conventional images for detecting proximal surface caries. intraoral X-ray film. Oral Radiol 10, 33–40. Dentomaxillofac Radiol 26, 32–38. 63) Wakoh, M., Kitagawa, H., Harada, T., Shibuya, 73) Wulff, H.R. (1981). Rational diagnosis and H. and Kuroyanagi, K. (1997). Computed treatment. An Introduction to Clinical Decision Dental Radiography system versus conven- Making, 2nd ed., Blackwell Scientific Publica- tional dental X-ray films for detection of simu- tions, Oxford. lated proximal caries. Oral Radiol 13, 73–82. 64) Welander, U., Nelvig, P., Tronje, G., McDavid, W.D., Dove, S.B., Morner, A-C. and Reprint requests to: Cederlund, T. (1993). Basic technical proper- Dr. Mamoru Wakoh ties of a system for direct acquisition of digital Department of Oral and intraoral radiographs. Oral Surg Oral Med Oral Maxillofacial Radiology, Pathol 75, 506–516. Tokyo Dental College, 65) Wenzel, A., Fejerskov, O., Kidd, E. Joyston- 1-2-2 Masago, Mihama-ku, Bechael, S. and Groeneveld, A. (1990). Chiba 261-8502, Japan