|Year : 2015 | Volume
| Issue : 3 | Page : 79-82
Assessment of digital enhancement filters in the radiographic determination of alveolar bone level
Matheus Lima Oliveira1, Luciana Moraes2, Jonathan Nicholas Santos Pereira2, Guilherme Monteiro Tosoni2
1 Department of Oral Diagnosis, Piracicaba Dental School, University of Campinas, Piracicaba, Brazil
2 Department of Oral Diagnosis and Surgery, Araraquara Dental School, São Paulo State University-UNESP, Araraquara, SP, Brazil
|Date of Web Publication||27-Nov-2015|
Dr. Matheus Lima Oliveira
901, Limeira Avenue, 13414-903 Piracicaba, SP
Source of Support: None, Conflict of Interest: None
Context: Recent studies have emphasized the advantages of digital radiographic imaging systems for periodontal evaluation. Because digital enhancement filters are designed to improve image quality, the real impact of such improvement on diagnostic accuracy needs to be evaluated for different clinical tasks. Aims: To assess the efficacy of digital enhancement filters in the radiographic determination of the alveolar bone level. Materials and Methods: Twenty-four digital periapical radiographs were obtained from mandibular sites with alveolar bone loss. Two enhancement filters-perio and invert-were applied separately and jointly on all radiographic images. The alveolar bone level was measured by consensus agreement of two examiners in 55 sites on the standard and enhanced radiographic images. A single examiner obtained the actual measurements on the dry mandible by using a periodontal probe and a digital caliper. For each assessment site, three measurements were obtained and averaged. One-way analysis of variance for differences between the actual measurements and radiographic image groups was performed (α = 0.05). Results: The digital radiographic measurements overestimated the actual by approximately 0.2 mm. There was no statistically significant difference between radiographic image groups. Furthermore, none of them differed from the actual measurements. Conclusion: Under the tested conditions, the perio and invert enhancement filters did not interfere with the efficacy of the radiographic images, which were accurate in determining the alveolar bone level.
Keywords: Alveolar bone loss, digital dental accuracy, dimensional measurement accuracy, radiographic image enhancement
|How to cite this article:|
Oliveira ML, Moraes L, Santos Pereira JN, Tosoni GM. Assessment of digital enhancement filters in the radiographic determination of alveolar bone level. J Oral Maxillofac Radiol 2015;3:79-82
|How to cite this URL:|
Oliveira ML, Moraes L, Santos Pereira JN, Tosoni GM. Assessment of digital enhancement filters in the radiographic determination of alveolar bone level. J Oral Maxillofac Radiol [serial online] 2015 [cited 2019 Nov 15];3:79-82. Available from: http://www.joomr.org/text.asp?2015/3/3/79/170610
| Introduction|| |
Periodontal disease is a multifactorial and chronic infectious process that begins with an inflammation of the gingival tissues surrounding the teeth and becomes more severe by resorbing the underlying bone. Although the clinical examination reveals characteristics of the periodontal condition, it has to be complemented by radiographic exams. , The radiographic image provides information about the periodontal support and alveolar bone level along the progression of the disease, and contributes significantly to the diagnosis and treatment plan.  The determination of the alveolar bone height is an important step in such assessment, which consists of obtaining the distance between the cement-enamel junction (CEJ) and the alveolar crest (AC). 
Technological advances have brought positive alternatives to the radiographic examination, by means of digital radiography. Importantly, radiographic digital imaging eliminates chemical processing, allows for image postprocessing, can be electronically transmitted and may require less radiation than analogue film.  Digital image processing corresponds to any computer-aided editing with the purpose of enhancing aspects of the image not readily apparent in its original form and providing images with diagnostically important information.  Each digital radiographic imaging system works with specific accompanying software that provides a range of enhancement filters.
The VistaScan storage phosphor plate system (Dürr Dental, Beitigheim-Bissingen, Germany) is a digital radiographic imaging equipment that is capable of capturing images with high spatial and contrast resolutions: Twenty line pairs per millimeter (L p/mm) and 16 bits, respectively. The software DBSWIN (Dürr Dental) acquires and manages the radiographic images, and presents several enhancement filters, in addition to some specifically intended for periodontal, endodontic and caries evaluation. Among the image enhancement possibilities, there are two filters known as perio and Grey Scale Inversion (henceforth, invert). The former is a high-pass filter and was developed to enhance periodontal structures by accentuating the transition in density levels after mathematical recomputation of pixels, , and the latter inverts the image grey scale, changing the pixels of low value (dark) to high value (bright) and vice-versa.  Caution is needed in the application of such filters, since their efficacy has not been scientifically proven for all diagnostic tasks. Thus, the aim of this paper is to assess the efficacy of the perio and invert filters in the radiographic determination of the alveolar bone level.
| Materials and Methods|| |
This study was designed according to the local Institutional Research Ethics Committee and carried on after its approval (protocol # 01387512.4.0000.5416). Thirteen partially edentulous dry human mandibles presenting one or more sites of alveolar bone loss <2 mm were selected. Exclusion criteria were cervical restorations, and any damage on AC and/or CEJ. Twenty-four periapical radiographs were obtained from these sites by paralleling technique using VistaScan size 2 storage phosphor plates and a Gendex 765DC Intraoral X-ray Unit (Gendex, Des Plaines, IL, USA) adjusted at 65 kVp, 7 mA, and 0.125 s. An acrylic plate (20-mm thick) was placed between the X-ray tube and the mandible as soft-tissue equivalent material, and the focal spot-to-image receptor distance was fixed at 32.5 cm. The VistaScan Combi Plus Scanner (Dürr Dental, Beitigheim-Bissingen, Germany) captured the latent image on the phosphor plates using the computer software DBSWIN 5.1.1. All plates were scanned with a spatial resolution of 20 L p/mm and contrast resolution of 16 bits, and stored as TIFF files. In addition, two enhancement filters-perio and invert-were applied separately and jointly on all radiographic images, resulting in a total of four groups of image: "No filter" (henceforth, Standard), perio, invert, perio + invert [Figure 1].
|Figure 1: Digital radiographic images (a: Standard, b: Perio, c: Invert, d: Perio + invert)|
Click here to view
The CEJ and the AC were determined by consensus agreement of two observers, who measured the distance between those anatomical landmarks in 55 sites on the radiographic images. The observers were senior dental students who had successfully accomplished oral radiology and periodontology classes and were previously trained and calibrated to make the measurements. The radiographic measurements were done in a low-light room with the aid of a digital ruler and a 23-inch widescreen LED monitor with a spatial resolution of 1920 × 1080 pixels (Philips 236VL, Amsterdam, Netherlands). At least 1-week interval elapsed between the assessments of each image group and the observers were allowed to adjust the contrast and brightness in all of them. A single and independent examiner obtained the actual measurements by using a digital caliper (Mitutoyo CD-12" PSX, Kawasaki, Japan) to measure the distance between the tip of a Williams periodontal probe and a rubber stop, after being placed on the physical AC and CEJ, respectively. For each assessment site, three measurements were obtained and averaged both on the radiographic image and on the dry mandible.
One-way analysis of variance (ANOVA) for differences between the actual measurements and radiographic image groups was performed. The significance level was set at 5% (α = 0.05).
| Results|| |
As shown in [Figure 2], the mean values, in millimeter, (± standard deviation) of the actual measurements were 3.00 ± 0.84 and the measurements obtained from the Standard, perio, invert and perio + invert radiographic images were, respectively, 3.26 ± 1.32, 3.25 ± 1.33, 3.27 ± 1.34, 3.24 ± 1.34. The radiographic measurements overestimated the actual by approximately 0.2 mm.
|Figure 2: Mean values (black squares) and ± standard deviation (horizontal dashes), in millimeter, of the actual and radiographic measurements (standard, perio, invert and perio + invert images)|
Click here to view
[Table 1] shows that the ANOVA did not detect statistically significant difference between the radiographic image groups. Furthermore, none of the groups was significantly different from the actual measurements (P = 0.771).
|Table 1: One-way ANOVA for CEJ-AC distance data according to radiographic image groups and actual measurements|
Click here to view
| Discussion|| |
Digital image receptors have shown to be more accurate for periodontal diagnosis than analogue film. Vandenberghe et al.  showed that both the photostimulable phosphor (PSP) and charge coupled device (CCD) digital imaging systems are adequate for periodontal diagnosis, with notable dose reduction. However, caution is needed when using CCD sensors at high exposure times to avoid image deterioration by blooming effects.
The efficacy of digital enhancement filters on radiographic images has been studied in the scientific literature since the introduction of digital imaging systems.  Digital enhancement filters are developed with the aim of allowing for better accuracy in the diagnostic process, by supposedly improving the image quality. Some filters are only available for specific software applications, but the majority of them are more commonly widespread. Despite the fact that the perio filter, assessed in the present study, is available exclusively at the DBSWIN software, it is a sharpening filter that contrasts the edges of the structures.  This can also be verified on a subjective evaluation of the [Figure 1]b. On the other hand, the invert filter (also known as negative or positive) can be easily found in several digital image editing software applications.
Studies have shown that enhancement filters have great acceptance by examiners, who demonstrate significant satisfaction with their use in the evaluation of periodontal structures.  Remarkably, that is a subjective analysis and not always related to the accuracy of diagnostics outcomes.  The results of our ex vivo study revealed a nonstatistically significant overestimation of the distance between the CEJ and AC, since all digital radiographic measurements were about 0.2 mm greater than the actual ones, irrespective of the enhancement mode. It can be justified by the geometry-related inherent magnification observed on the periapical paralleling technique. The distance between the object and the image receptor coupled with the divergent cone-shaped X-ray beam enlarges the projected image. Considering the fact that this study aimed to reproduce a clinical situation and made use of human mandibles, it is important to highlight the thickness of the alveolar and cortical bone on the lingual aspect of the mandible, when studying the results. Though numerically noteworthy, the observed overestimation is presumably clinically irrelevant.
Studies evaluating some task-specific enhancement filters from the VistaScan system are controversial. Haiter-Neto et al.  evaluated the caries 1 and caries 2 filters (intended for caries lesion detection) and concluded that both cannot be recommended for detection of proximal caries lesions. de Azevedo Vaz et al.  demonstrated that the perio, endo, caries 1 and caries 2 filters were not accurate for measuring peri-implant bone level, and only the Standard images, the fine and emboss filters provided satisfactory results. Conversely, Oliveira et al.  stated that the endo filter (intended for endodontic purposes), associated with high spatial and contrast resolutions is recommended for the determination of file lengths.
The invert is a very popular filter and its usefulness has been evaluated for different diagnostic tasks. Similarly to our study, Scaf et al.  and Oliveira et al.  evaluated periodontal bone loss and endodontic measurements, respectively, in Standard and invert digital radiographic images and concluded that both images were reliable, with no significant difference in the diagnostic accuracy. In contrast, Castro et al.  demonstrated that the diagnostic accuracy of the invert in the detection of proximal caries did not differ from Standard images and both had a very poor performance. Additionally, de Molon et al.  showed that measurements of periodontal bone defects in grey scale-inverted digital images obtained with CCD sensors were inferior to analogue film.
In the present study, the separate and joint use of the perio and invert filters did not contribute for better accuracy in the measurement of the alveolar bone level, which is in disagreement with Vandenberghe et al.,  who achieved better results after applying the perio filter on 12-bit images. It can possibly be explained by the higher contrast resolution of our images (16 bits) that may have favored the radiographic image of the alveolar bone crest to be even subtler. Nevertheless, perio, invert and perio + invert images were not significantly different from the standard images and actual values, what demonstrates that any of them can provide reliable measurements. The wide availability of digital enhancement filters does not exempt the radiographic technique from having the technical principles carefully observed.
| Conclusion|| |
The perio and invert enhancement filters did not interfere with the efficacy of PSP-based radiographic images, which were accurate in determining the alveolar bone level.
Financial support and sponsorship
The study was funded by FAPESP, a research foundation of the State of São Paulo, Brazil (process no: 2012/08744-6).
Conflicts of interest
There are no conflicts of interest.
| References|| |
Esmaeli F, Shirmohammadi A, Faramarzie M, Abolfazli N, Rasouli H, Fallahi S. Determination of vertical interproximal bone loss topography: Correlation between indirect digital radiographic measurement and clinical measurement. Iran J Radiol 2012;9:83-7.
Safi Y, Kadkhodazadeh M, Safai P, Esmaeelinejad M, Shamloo N. Evaluation of alveolar crest bone loss via premolar bitewing radiographs: Presentation of a new method. J Periodontal Implant Sci 2014;44:222-6.
Corbet EF, Ho DK, Lai SM. Radiographs in periodontal disease diagnosis and management. Aust Dent J 2009;54 Suppl 1:S27-43.
Tihanyi D, Gera I, Eickholz P. Influence of individual brightness and contrast adjustment on accuracy of radiographic measurements of infrabony defects. Dentomaxillofac Radiol 2011;40:177-83.
Parks ET, Williamson GF. Digital radiography: An overview. J Contemp Dent Pract 2002;3:23-39.
Borg E, Gröndahl K, Persson LG, Gröndahl HG. Marginal bone level around implants assessed in digital and film radiographs: In vivo
study in the dog. Clin Implant Dent Relat Res 2000;2:10-7.
de Azevedo Vaz SL, Neves FS, Figueirêdo EP, Haiter-Neto F, Campos PS. Accuracy of enhancement filters in measuring in vitro
peri-implant bone level. Clin Oral Implants Res 2013;24:1074-7.
Wenzel A, Haiter-Neto F, Frydenberg M, Kirkevang LL. Variable-resolution cone-beam computerized tomography with enhancement filtration compared with intraoral photostimulable phosphor radiography in detection of transverse root fractures in an in vitro
model. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;108: 939-45.
Oliveira ML, Vieira ML, Cruz AD, Bóscolo FN, de Almeida SM. Gray scale inversion in digital image for measurement of tooth length. Braz Dent J 2012;23:703-6.
Vandenberghe B, Corpas L, Bosmans H, Yang J, Jacobs R. A comprehensive in vitro
study of image accuracy and quality for periodontal diagnosis. Part 1: The influence of X-ray generator on periodontal measurements using conventional and digital receptors. Clin Oral Investig 2011;15:537-49.
Kokubo Y, Hardy WH. Digital image processing: A path to better pictures. Ultramicroscopy 1982;8:277-86.
Yalcinkaya S, Künzel A, Willers R, Thoms M, Becker J. Subjective image quality of digitally filtered radiographs acquired by the Dürr VistaScan system compared with conventional radiographs. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;101:643-51.
Oliveira ML, Ambrosano GM, Almeida SM, Haiter-Neto F, Tosoni GM. Efficacy of several digital radiographic imaging systems for laboratory determination of endodontic file length. Int Endod J 2011;44:469-73.
Haiter-Neto F, Casanova MS, Frydenberg M, Wenzel A. Task-specific enhancement filters in storage phosphor images from the VistaScan system for detection of proximal caries lesions of known size. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2009;107:116-21.
de Oliveira ML, Pinto GC, Ambrosano GM, Tosoni GM. Effect of combined digital imaging parameters on endodontic file measurements. J Endod 2012;38:1404-7.
Scaf G, Morihisa O, Loffredo Lde C. Comparison between inverted and unprocessed digitized radiographic imaging in periodontal bone loss measurements. J Appl Oral Sci 2007;15:492-4.
Castro VM, Katz JO, Hardman PK, Glaros AG, Spencer P. In vitro
comparison of conventional film and direct digital imaging in the detection of approximal caries. Dentomaxillofac Radiol 2007;36:138-42.
de Molon RS, Morais-Camillo JA, Sakakura CE, Ferreira MG, Loffredo LC, Scaf G. Measurements of simulated periodontal bone defects in inverted digital image and film-based radiograph: An in vitro
study. Imaging Sci Dent 2012;42:243-7.
Vandenberghe B, Bosmans H, Yang J, Jacobs R. A comprehensive in vitro
study of image accuracy and quality for periodontal diagnosis. Part 2: The influence of intra-oral image receptor on periodontal measurements. Clin Oral Investig 2011;15:551-62.
[Figure 1], [Figure 2]