The impact of image compression on diagnostic quality of digital images for detection of chemically-induced periapical lesions.

OBJECTIVES: To test the hypothesis that there is no significant difference in the detectability of chemically-induced periapical lesions between a non-compressed image and one subjected to a Joint Photographic Experts Group (JPEG) lossy compression technique at a ratio of 23:1 or less. METHOD: Chemically-induced periapical lesions were created by placing a solution of 70% perchloric acid at the apex of extracted teeth in 13 human jaw cadaver specimens. Acid was applied in seven incremental time periods from 0-32 h. Extracted teeth were replaced in the socket and images were made using the Schick CDR digital sensor. Using a JPEG lossy compression algorithm, five compression ratios of 2:1, 14:1, 23:1, 28:1 and 47:1 were applied to the images. Images were viewed three times by three observers who ranked the presence or absence of a lesion at three sites, the mesial area, distal area and apex of the tooth, on a 5-point confidence scale. Intraobserver and interobserver agreement and agreement between the compressed and the original images were assessed with intraclass correlation coefficients (ICCs). RESULTS: Overall ICCs for measuring intraobserver agreement using all images were 0.77, 0.84, and 0.50 for the three observers, respectively. The overall ICC for assessing agreement between observers was 0.57. There was no significant difference (P>0.05) between compressed and original images for any site at compression ratios of 2:1, 23:1 and 28:1. There were significant differences for a compression ratio of 47:1. CONCLUSION: JPEG compression does not impact detectability of artificial periapical lesions at low and moderate compression ratios up to and including 28:1.

Radiographic digital image enhancement. Part II: transform domain techniques.

Digital image enhancement techniques can have a significant impact on the diagnostic quality of a radiographic image. Selection of an image enhancement technique is a function of image content and its attributes of interest. The diagnostic value of a radiographic image is commonly reduced by its intrinsic limitations as well as interference from unwanted signals. In this second article on radiographic image enhancement the Fourier transform, the most commonly used transformation method, is used to examine examples of methods for image enhancement, with particular emphasis on the removal or reduction of noise.

Radiographic image enhancement. Part I: spatial domain techniques.

Digital image enhancement techniques provide a multitude of choices for improving the visual quality of diagnostic images. Appropriate choice of such techniques is greatly influenced by the imaging modality, task at hand and viewing conditions. This sequence of two articles will provide an overview of underlying concepts, along with algorithms commonly used for radiographic image enhancement. The first article focuses on spatial domain techniques for radiographic image enhancement, with particular reference to point processing methods, histogram modification and unsharp masking.

Direct digital radiography for caries detection and analysis.

Recent developments in the field of electronic imaging have provided a new set of imaging tools for intra-oral imaging and clinical diagnosis. In this chapter, we review the general structure and characteristics of image sensors used in digital radiography and their application for clinical caries detection and analysis. An overview of the literature, comparing the diagnostic accuracy of digital and film-based radiography and the impact of image-processing methods, is provided. We also describe several approaches for quantitative assessment of radiographic images for caries detection and assessment. Examples include digital subtraction radiography, and 2-D and 3-D density profiling. To increase the accuracy and reduce observer variability the use of 3-D imaging and computer-aided diagnosis is presented as future direction for clinical caries diagnosis.

Detection of natural white spot caries lesions by an ultrasonic system.

Ultrasound has been used in industrial business as one of the nondestructive measurement methods. It was hypothesized that nondestructiveness of the ultrasonics could be useful in determination of demineralization of noncavitated carious lesions on human enamel. This investigation was designed to determine the presence of natural carious lesions on proximal surfaces of human molar teeth using an ultrasonic system compared to radiography and histology as the gold standards. Measurements were made directly from proximal surfaces of 20 mandible molar teeth with white spot carious lesion by 2 examiners independently with the ultrasonic system. Ultrasonic evaluation of each natural white spot lesion had a sensitivity of 88%, specificity of 86%, positive predictive value of 88% and a negative predictive value of 86%, and the chance-corrected agreement was also satisfactory (kappa=0.74) compared to histology. The radiograph demonstrated chance-corrected agreement of 0.41:0.38 for the first and second examiners, respectively. Duncan test analysis of the numerical values was significantly different for the intact and the noncavitated carious surfaces (p<0.05). The results indicated that the ultrasonic evaluation is a sensitive method for the detection of the natural white spot carious lesions and can differentiate the changes in elastic properties of enamel numerically.

Digital radiographic image archival, retrieval, and management.

Data in the health care environment can encompass a wide range of contents, each representing one aspect of health care delivery. Examples of such a diverse and yet related content are diagnostic images, laboratory results, patient history, hospital administration, and insurance and payer coverage. The focus of this article is the handling of diagnostic imaging in general and radiographic imaging in particular, therefore fundamental concepts and issues are presented and discussed.

Dental caries diagnosis.

As with any disease process, the early initiation of therapy for dental caries is often the most effective means to ensure resolution. However, for any therapy to be effective, early diagnosis is paramount to success. Unfortunately, current tools used in dental caries detection are not sensitive enough to diagnose the disease process in its early stages and, frequently, once a diagnosis is made restoration is the only effective means of treatment. Quantitative light-induced fluorescence, electrical conductance measurements, direct digital radiography, and direct fiber optic transillumination are methods that may offer promise as effective tools in diagnosing early dental caries. This article describes some of the research that has been performed to make these methods a reality.

Detection of early interproximal caries in vitro using laser fluorescence, dye-enhanced laser fluorescence and direct visual examination.

This in vitro study evaluated the use of laser fluorescence (LF) for the detection of early interproximal carious lesions and whether the detection could be enhanced using a fluorescent dye (DELF). Direct visual examination (DV) was used for comparison. Eighty extracted teeth were used, arranged in 20 blocks, each block having 2 premolars and 2 molars, lined up in a simulated sextant situation. After cleaning with a microabrasion kit, a subcontact window on half of the surfaces (60) was exposed to Carbopol white-spot solution for 5 days. The teeth were remounted in stone and examined by three independent examiners. For LF and DELF an argon laser was used (mixed wavelength of 488 and 514 nm) viewed through glasses (excluding wavelength <520 nm). For DELF a sodium fluorescein dye (0. 075%) was applied before examination. A clinical examination light was used for DV. The approximal surfaces were scored for lesion presence or absence. To verify lesion presence, the subcontact area was cut perpendicularly to the surface, stained with rhodamine B, and images were taken using a confocal microscope. The images were analyzed using a histogram program for lesion depth and image area. Lesions were present in 62 out of 120 approximal surfaces, with an average depth of 60 microm (range 17-190 microm). Sensitivity ranges for LF, DELF and DV were 56-74, 61-79 and 58-74%, and specificity ranges 67-78, 86-98 and 83-97%, respectively. With this model DELF compared favorably with DV and LF in sensitivity, but specificity was better for DELF and DV than for LF.

An in vitro comparison between laser fluorescence and visual examination for detection of demineralization in occlusal pits and fissures.

It has been demonstrated that when excited by laser light carious enamel appears dark compared to luminescent sound enamel. The aim of this study was to compare the sensitivity and specificity of visual exams (V), laser fluorescence (LF) and dye-enhanced LF (DELF) for detecting demineralization in occlusal pits and fissures. The actual presence of lesions was determined by subsequent confocal laser microscopy (CM), which was compared to histology (H). Independent clinical examiners visually graded three sites on occlusal surfaces of extracted, human premolars as sound or carious and also rated the color of each graded site as: 0 = same as surrounding enamel; 1 = white; 2 = light brown, or 3 = brown/dark brown. An argon laser was used to illuminate the teeth for LF and DELF; the images were captured with a CCD camera and then analyzed. DELF images were captured after the teeth had been exposed to 0.075% sodium fluorescein. Sections were then cut from each specimen and analyzed by CM and H for the presence or absence of caries. Results showed that DELF (0.72) was significantly more sensitive (p<0.05) than LF (0.49) and V (0.03) for detecting caries, but there were no significant differences among the methods in specificity (V 1.00; LF 0.67; DELF 0.60). When color was used as an indication of caries in V (VC, sensitivity 0.47; specificity 0.70), V exams were not different from LF. The area under the ROC curve, using H as the gold standard and CM as the test, was 0.78. Results indicated that DELF was the best diagnostic tool and that VC and LF were equally effective as diagnostic methods, when color of fissures was included as an indication of demineralization in the visual exam.

Laser fluorescence detection of demineralization in artificial occlusal fissures.

Laser fluorescence (LF) has been used previously to detect early smooth-surface lesions. Although its use for detection of occlusal demineralization has been implicated, it has not been demonstrated. The aim of this study was to determine whether LF could detect demineralization in the base of artificial fissures. To employ LF for detection of occlusal demineralization an apparatus was devised to direct laser light into fissures and simultaneously detect fluorescence from the base of the fissures. Three groups (n = 40/group) of differing fissure types were prepared (straight wall, converging and diverging wall) with either a sound or lesioned base. One half of each group was examined with LF and dye-enhanced LF (DELF); the other half was examined with LF, exposed to plaque, examined with LF and DELF, air-polished and examined with DELF. All images were scored twice as either (1) carious; (2) sound, or (3) undetermined, by a group of 3 examiners. For fissures without plaque, the average sensitivity was higher for DELF (0.76) than for LF (0.54) (p < 0.05). Likewise, the average specificity was higher for DELF (0.64) than for LF (0.29) (p < 0.05). In the presence of plaque, sensitivity was higher for DELF (0.91) compared to LF (0.43); however, specificity was lower for DELF (0.05) compared to LF (0.55). When the fissures were air-polished and then examined with DELF, sensitivity averaged 0.82 (p < 0.05), and specificity increased consistently (average 0.51, p < 0.05). It was concluded that, in the absence of plaque, DELF was a better diagnostic tool than LF for detection of demineralization in artificial fissures.

Effect of two fluorescent dyes on color of restorative materials.

PURPOSE: The objective of these two studies was to determine whether the use of dyes in vitro would affect the shade of tooth-colored restorative materials. MATERIALS AND METHODS: Two separate studies were performed. In Study 1, four groups (N = 36) of restorative materials; (1) resin composite (RSN), (2) resin-modified glass ionomer (RMGI), (3) conventional glass ionomer (CGI), and (4) porcelain (PCN), were subjected to a rinse in either tap water (control) or Pyrromethene 556 dye at a concentration of 0.01M for 30 seconds or 0.0015 M for 60 seconds. They were then rinsed with 25% ethanol as would be done in clinical usage. In Study 2, the same restorative materials (except PCN) were subjected to a rinse in either tap water (control) (N = 8) or 0.0018 M sodium fluorescein (N = 8) for 60 seconds followed by a 5-second rinse with tap water. For both studies the specimens were analyzed for overall color changes (delta E*) comparing baseline to immediately post dye exposure and then again after 1, 4, 24, and 48-hour storage under running tap water. In addition, an image of each specimen, illuminated under an argon laser light (HGM), was captured with a miniature charged-coupled device (CCD) color camera at the same time intervals when the color was measured. These images were analyzed for fluorescence using computer assisted methods. A software program then computed the mean and standard deviation of the fluorescence values for each image. The data were analyzed using repeated measures ANOVA. RESULTS: delta E* values from baseline on specimens treated with either Pyrromethene 556 or sodium fluorescein were not significantly different from water at any time. The only exceptions were CGI specimens exposed to 0.01 M Pyrromethene 556 and RMGI exposed to sodium fluorescein which had a higher delta E* immediately after treatment, however this effect was reversed after 1-hour rinse. Mean delta E* was less than 2.69 for specimens treated with Pyrromethene 556 and less than 3.20 for specimens treated with sodium fluorescein after 1-hour rinse in water. When the mean fluorescence level, as determined by the computer, was averaged there was no difference between the Pyrromethene 556 treatments and control for RMGI, RSN and PCN after 4-hour rinse in water and, as time in storage increased, the fluorescence level decreased. CGI treated with 0.01 M Pyrromethene 556 was significantly more fluorescent even after 48-hour rinse in water. All specimens treated with sodium fluorescein were also significantly more fluorescent than control even after 48-hour rinse in water. It was concluded that exposure of these restorative materials to Pyrromethene 556 at 0.01 M or 0.0015 M or to 0.0018 M sodium fluorescein will not lead to any prolonged visually detectable color changes.

Relative ability of laser fluorescence techniques to quantitate early mineral loss in vitro.

This in vitro investigation was undertaken to explore two different nondestructive methods to detect very early demineralization. These methods were based on the premise that the clinical detection of caries at a very early stage of formation might permit more efficient reversal of the caries process than may occur when lesions are detected at a more advanced stage, such as a so-called 'white spot'. The methods evaluated in this study were quantitative laser fluorescence (QLF) and an experimental dye-enhanced laser fluorescence (DELF) technique. Prepared and polished bovine enamel specimens were demineralized in a conventional lactic acid-Carbopol solution for varying periods of time between 0 and 24 h with an area of sound enamel retained on each specimen. The coded and randomized specimens were then analyzed for the presence and severity of enamel demineralization using QLF after which they were exposed to a selected dye (Pyrromethene 556) and similarly examined using DELF. The specimens were then sectioned and examined by conventional transverse microradiography and by confocal laser-scanning microscopy. Results were analyzed statistically with sensitivity and specificity determined using sound enamel as the reference. The results indicated that QLF could detect demineralization which occurred as a result of 8 h exposure to the decalcification solution and was able to quantify changes in lesion severity associated with longer demineralization. While DELF was capable of detecting enamel demineralization after only 2 h exposure to the decalcification solution, it was unable to quantify increasing amounts of demineralization associated with longer periods of exposure to the decalcification solution. In summary, while DELF was able to detect very early demineralization, only QLF was capable of detecting and quantifying changes in the extent of the decalcification occurring with demineralization periods up to 24 h.

Interpretation of chemically created periapical lesions using direct digital imaging.

Perchloric acid (70%) was used to create simulated periapical lesions in tooth sockets of 15 dentate cadaver jaw specimens. Using the Trophy USA direct digital radiographic system, linear images were captured at selected time intervals after initial acid application and altered by contrast reversal, pseudocolor enhancement, and two forms of histogram equalization. The 525 total images were randomized for display on a computer monitor for evaluation by five endodontists. Images were evaluated twice by each rater, with viewings 1 to 2 wk apart. Statistical analysis determined interrater variability, intrarater reproducibility, and the relative merits of each enhancement technique. At 8, 12, 16, and 24 h after acid application, both techniques of histogram equalization yielded a statistically significant improvement over reverse contrast in perception of periapical patholais. Linear and pseudocolor-enhanced images were also significantly more diagnostic than reverse contrast at 12, 16, and 24 h. Intrarater reproducibility showed moderate agreement, but analysis showed only a fair level of interrater agreement.

Rate of mesial translation of mandibular molars using implant-anchored mechanics.

Retromolar dental implants served as anchorage to close first molar extraction sites in five adult patients. Rates of unidirectional space closure for mandibular second molars were assessed with periapical radiographs superimposed on anatomical landmarks and retromolar anchorage implants. Regression analysis revealed that mesial displacement of the midroot area of the leading root was less variable (r = 0.97) than for other landmarks on the same teeth: crown (r = 0.83), alveolar crest (r = 0.82) or apex (r = 0.90). When mesial root movement (uprighting) was the principal feature of the initial mechanics (4 of the 5 patients), mesial movement of the apex was about 0.60 mm/mo for the first 8 months and then decreased to about 0.34 mm/mo as the trailing (distal) root of the second molar engaged the relatively dense bone formed by the leading (mesial) root. During the last year of space closure, radiolucent foci were noted 1-2 mm ahead of the distal root. These data suggest: (1) sustained orthodontic translation is a physiological manifestation of bone modeling and remodeling throughout the adjacent alveolar process, and (2) rate of mandibular molar translation is inversely related to the apparent radiographic density of the resisting alveolar bone.