Abstract
Objectives
Digital radiography has become an integral part of dentistry. Digital radiography does not require film or dark rooms, reduces X-ray doses, and instantly generates images. The aim of our study was to compare the subjective image quality of two digital dental radiographic systems with conventional dental film.
Materials & methods
A direct digital (DD) ‘Digital’ system by Sirona, a semi-direct (SD) digital system by Vista-scan, and Kodak ‘E’ speed dental X-ray films were selected for the study. Endodontically-treated extracted teeth (n = 25) were used in the study. Details of enamel, dentin, dentino-enamel junction, root canal filling (gutta percha), and simulated apical pathology were investigated with the three radiographic systems. The data were subjected to statistical analyzes to reveal differences in subjective image quality.
Results
Conventional dental X-ray film was superior to the digital systems. For digital systems, DD imaging was superior to SD imaging.
Conclusion
Conventional film yielded superior image quality that was statistically significant in almost all aspects of comparison. Conventional film was followed in image quality by DD, and SD provided the lowest quality images. Conventional film is still considered the gold standard to diagnose diseases affecting the jawbone.
Recommendations
Improved software and hardware for digital imaging systems are now available and these improvements may now yield images that are comparable in quality to conventional film. However, we recommend that studies still use more observers and other statistical methods to produce ideal results.
Keywords: Oral radiology, Digital radiography, Comparative radiographic study, Direct digital imaging, Semi-direct digital radiography
1. Introduction
Digital radiography is a recent advance in the field of maxillofacial radiology and does not require the use of a conventional radiological film. The first digital dental radiographic system was the RVG (Radio-Visio-Graphy) system introduced in 1989 by Trophy (White and Pharoah, 2004). Film is replaced by an electrostatic device that sends images to a computer. Since then, several companies have introduced digital systems specifically for dental imaging. There are currently more than 15 companies who make digital systems for both dental and panoramic imaging. Digital systems have many advantages including lower radiation doses, real time imaging, no requirement for dark rooms, and image manipulation can be easily performed (Dental Association Council on Scientific Affairs et al., 2006). Several studies have investigated the efficacy of digital systems to diagnose dental pathologies (Wakoh and et al., 1997).
Currently, there are two digital image acquisition platforms available. They are direct digital (DD) and semi-direct (SD) digital imaging systems. DD uses charge couple device (CCD) or complimentary metal oxide semi conductor (CMOS) sensors. SD digital imaging uses a photostimulable phosphor (PSP) as the sensor. The sensors used in DD systems are bulky and connected to computers by cables or wireless networks. Sensors for SD digital systems are thin like film, so they offer portability and flexibility, but need to be digitized using special scanners.
Many studies have been conducted on the diagnostic accuracy of these systems. Most conclude that film and digital systems are nearly comparable to diagnose diseases affecting tooth-bearing areas of the jawbone (Dental Association Council on Scientific Affairs et al., 2006). Since few comparative studies using all three systems exist, this study evaluated the subjective image quality of DD and SD digital systems with conventional film for dental radiographic assessments.
2. Material and methods
This study was approved by the Scientific and Ethics Committee of the College of Dentistry at King Khalid University and we obtained informed consent for use of samples. A total of 25 extracted teeth were selected for the study. The teeth were previously used by pre-clinical endodontic students for training on root canal procedures. All teeth had simulated apical pathologies and wax root canal fillings. All teeth were mounted on a plaster base and radiographs were taken with DD, SD digital, and film radiography. X-rays were generated by a Sirona digital dental unit operating at 70 kVp and 0.8 mA. X-ray doses were given according to manufacturer protocols. A Digital system (Sirona, Germany) was used for DD acquisitions, a Vista-scan (Durr, Germany) was used for SD digital acquisitions, and E speed film (Kodak, Rochester, NY) was used for conventional radiography. Both sensors and the film were adult size 2. The teeth were placed over the sensors or film and the X-ray tube head remained at a fixed distance from the teeth. A method to ensure all images were acquired in parallel was employed to acquire images (Fig. 1). Films were processed using automatic processing machine (Durr X-24). Kodak Dental Readymatic processing solutions were used and the processing time was 6 min as recommended by the manufacturer. Test radiographs were taken daily and their image quality was compared to a reference film. DD images were saved immediately after performing image enhancement by both the observers at the same time. SD digital images were generated using a Vista-scan drum scanner (Fig. 5) and were saved after appropriate image adjustments were made by both observers together using software filters. Both observers evaluated digital images on computer monitors and films using a DENTSPLY X-ray view box with a 2× magnifying lens (Fig. 2). Digital images were also magnified 2× to standardize the evaluation (Figs. 3 and 4). Imaging evaluated enamel clarity, dentin clarity, dentino-enamel junction (DEJ) clarity, root canal filling (gutta-percha) clarity, and details of the simulated apical pathology. These were evaluated by using the 5-point scale provided. The scale graded criteria as poor, average, good, very good, and excellent. Scale values analysis was carried out using SPSS software (Version 21). The means and standard deviations of readings for both observers were calculated (Table 1). One-way ANOVAs were used to determine significance between groups (Table 2) and Fisher’s Least Significant Difference (LSD) tests were used for pairwise comparisons (Table 3). This was done for readings of both observers individually. The Spearman–Brown correlation test was used for readings from both observers (Table 4).
Figure 1.
Paralleling method.
Figure 5.
Vista-scan drum scanner.
Figure 2.
DENTSPLY X-ray view box and 2× magnifying lens.
Figure 3.
Direct digital image.
Figure 4.
Semi direct digital image.
Table 1.
Mean and standard deviation for readings of observers 1 and 2.
| N | Mean Obs-1 | Mean Obs-2 | Std. deviation Obs-1 | Std. deviation Obs-2 | Std. Error Obs-1 | Std. Error Obs-2 | Minimum Obs-1 | Minimum Obs-2 | Maximum Obs-1 | Maximum Obs-2 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Enamel clarity | DD | 25 | 1.88 | 1.68 | .900 | .476 | 0 | 1 | 3 | 2 | .184 | .095 |
| SD | 25 | 1.24 | .44 | .597 | .583 | 0 | 0 | 2 | 2 | .119 | .117 | |
| FL | 25 | 2.20 | 2.40 | .816 | .816 | 0 | 0 | 3 | 3 | .163 | .163 | |
| Total | 75 | 1.77 | 1.51 | .869 | 1.032 | 0 | 0 | 3 | 3 | .101 | .119 | |
| DEJ clarity | DD | 25 | 1.68 | 1.48 | .748 | .714 | 0 | 0 | 3 | 3 | .150 | .143 |
| SD | 25 | 1.28 | .48 | .614 | .586 | 0 | 0 | 2 | 2 | .123 | .117 | |
| FL | 25 | 1.88 | 2.32 | .726 | .852 | 0 | 0 | 3 | 3 | .145 | .170 | |
| Total | 75 | 1.61 | 1.43 | .733 | 1.042 | 0 | 0 | 3 | 3 | .085 | .120 | |
| Dentin clarity | DD | 25 | 1.40 | 1.32 | .577 | .627 | 1 | 0 | 3 | 3 | .115 | .125 |
| SD | 25 | 1.04 | .40 | .351 | .500 | 0 | 0 | 2 | 1 | .070 | .100 | |
| FL | 25 | 1.60 | 2.12 | .645 | .927 | 1 | 0 | 3 | 3 | .129 | .185 | |
| Total | 75 | 1.35 | 1.28 | .581 | .994 | 0 | 0 | 3 | 3 | .067 | .115 | |
| Apical pathology clarity | DD | 25 | 2.08 | 1.96 | .812 | 1.172 | 0 | 0 | 3 | 3 | .162 | .234 |
| SD | 25 | 1.44 | 1.00 | .917 | .913 | 0 | 0 | 3 | 3 | .183 | .183 | |
| FL | 25 | 1.44 | 1.56 | 1.003 | 1.044 | 0 | 0 | 3 | 3 | .201 | .209 | |
| Total | 75 | 1.65 | 1.51 | .951 | 1.107 | 0 | 0 | 3 | 3 | .110 | .128 | |
| GP | DD | 25 | 2.24 | 2.32 | .723 | .476 | 1 | 2 | 3 | 3 | .145 | .095 |
| SD | 25 | 1.44 | .52 | .961 | .714 | 0 | 0 | 3 | 2 | .192 | .143 | |
| FL | 25 | 2.28 | 2.56 | .678 | .651 | 1 | 1 | 3 | 3 | .136 | .130 | |
| Total | 75 | 1.99 | 1.80 | .878 | 1.103 | 0 | 0 | 3 | 3 | .101 | .127 | |
Table 2.
ANOVA for readings of observers 1 and 2.
| Sum of squares Obs1 | Sum of squares Obs2 | Df Obs1 | Df Obs2 | Mean square bs1 | Mean square Obs2 | F Obs1 | F Obs2 | Sig. Obs1 | Sig. Obs2 | ||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Enamel clarity | Between groups | 11.910 | 49.147 | 2 | 2 | 5.955 | 24.573 | 9.790 | 59.773 | .000 | .000 |
| Within groups | 43.185 | 29.600 | 71 | 72 | .608 | .411 | |||||
| Total | 55.095 | 78.747 | 73 | 74 | |||||||
| DEJ clarity | Between groups | 4.667 | 42.427 | 2 | 2 | 2.333 | 21.213 | 4.784 | 40.278 | .011 | .000 |
| Within groups | 35.120 | 37.920 | 72 | 72 | .488 | .527 | |||||
| Total | 39.787 | 80.347 | 74 | 74 | |||||||
| Dentin clarity | Between groups | 4.027 | 37.040 | 2 | 2 | 2.013 | 18.520 | 6.916 | 36.958 | .002 | .000 |
| Within groups | 20.960 | 36.080 | 72 | 72 | .291 | .501 | |||||
| Total | 24.987 | 73.120 | 74 | 74 | |||||||
| Apical pathology clarity | Between groups | 6.827 | 11.627 | 2 | 2 | 3.413 | 5.813 | 4.085 | 5.290 | .021 | .007 |
| Within groups | 60.160 | 79.120 | 72 | 72 | .836 | 1.099 | |||||
| Total | 66.987 | 90.747 | 74 | 74 | |||||||
| GP | Between groups | 11.227 | 62.160 | 2 | 2 | 5.613 | 31.080 | 8.832 | 80.379 | .000 | .000 |
| Within groups | 45.760 | 27.840 | 72 | 72 | .636 | .387 | |||||
| Total | 56.987 | 90.000 | 74 | 74 | |||||||
Table 3.
Multiple comparisons Fisher’s Least Significant Difference (LSD) test for readings of observers 1 and 2.
| Dependent variable | (I) Type | (J) Type | Mean difference (I − J) Obs1 | Mean difference (I − J) Obs2 | Std. error Obs1 | Std. error Obs2 | Sig. Obs1 | Sig. Obs2 |
|---|---|---|---|---|---|---|---|---|
| Enamel clarity | DD | SD | .635∗ | 1.240∗ | .223 | .181 | .006 | .000 |
| FL | .325 | .720∗ | .223 | .181 | .149 | .000 | ||
| SD | FL | .960∗ | 1.960∗ | .221 | .181 | .000 | .000 | |
| DEJ clarity | DD | SD | .400∗ | 1.000∗ | .198 | .205 | .047 | .000 |
| FL | .200 | .840∗ | .198 | .205 | .315 | .000 | ||
| SD | FL | .600∗ | 1.840∗ | .198 | .205 | .003 | .000 | |
| Dentin clarity | DD | SD | .360∗ | .920∗ | .153 | .200 | .021 | .000 |
| FL | .200 | .800∗ | .153 | .200 | .194 | .000 | ||
| SD | FL | .560∗ | 1.720∗ | .153 | .200 | .000 | .000 | |
| Apical pathology clarity | DD | SD | .640∗ | .960∗ | .259 | .296 | .016 | .002 |
| FL | .640∗ | .400 | .259 | .296 | .016 | .182 | ||
| SD | FL | .000 | .560 | .259 | .296 | 1.000 | .063 | |
| GP | DD | SD | .800∗ | 1.800∗ | .225 | .176 | .001 | .000 |
| FL | .040 | .240 | .225 | .176 | .860 | .177 | ||
| SD | FL | .840∗ | 2.040∗ | .225 | .176 | .000 | .000 | |
The correlations between 2 observers is significant at the 0.05 level.
Table 4.
Correlations between the two observers (Spearman Brown correlation test).
| Obs. 2 |
|||||
|---|---|---|---|---|---|
| Obs. 1 | Enamel clarity | DEJ clarity | Dentin clarity | Apical pathology clarity | GP |
| Enamel clarity | 0.525∗∗ | ||||
| DEJ clarity | 0.537∗∗ | ||||
| Dentin clarity | 0.532∗∗ | ||||
| Apical pathology clarity | 0.502∗∗ | ||||
| GP | 0.542∗∗ | ||||
The correlations between 2 observers is significant at the 0.01 level.
3. Results
Means and standard deviations for readings from Observers 1 and 2 are shown in Table 1. ANOVA analyses revealed significant differences between image types for Observer 1 in all the areas of comparison (Table 2). Post-hoc least squared difference (LSD) comparisons (Table 3) revealed that the DD system was significantly better than the SD digital system in all aspects. Additionally, film was found to be better than the SD digital system in all areas of comparison except for the clarity of apical pathology, where the results were similar (Table 3). No significant differences were found between film and DD imaging, but film did show trends of higher mean values (i.e., better results) in all areas except for the clarity of apical pathology.
For Observer 2, ANOVA revealed significant differences between groups for all aspects of comparison (Table 2). Post-hoc LSD comparisons (Table 3) showed that film was significantly better than both DD and SD digital imaging in all the areas of comparison except for the clarity of apical pathology and clarity of the gutta-percha filling, which showed trends to be better by film, but were not statistically significant. Film was significantly better than the SD digital imaging in all areas of comparison except the clarity of apical pathology, which showed a statistically insignificant trend to be better by film (Table 3).
The results of both observers were found to be better with film than either digital system (Table 4). DD imaging was almost as good as film with Observer 1. Spearman–Brown correlations showed a significant correlation between the readings of the two observers, indicating the reliability of both data sets for comparison (Table 4).
4. Discussion and conclusion
Digital radiography has been a major advancement in dentistry over the past 10 years. Intra-oral and panoramic radiographs can be acquired without darkrooms or chemicals for film processing (Wenzel, 1998). However, the main advantage has been the substantial reduction in X-ray doses compared to film (Dunn and Kantor, 1993; Wenzel and Grondahl, 1995). After its introduction in 1989, many companies have developed digital systems with advancements such as wireless technology.
The two main types of digital radiography are DD and SD digital. DD displays images immediately after acquisition (Brennan, 2002) while SD digital images can be displayed on a computer after scanning the PSP sensor. Some studies use the term indirect in place of SD, while others consider both digital platforms as direct digital and scanning of film to be the indirect method. Films can be scanned with a transparency scanner to digitize it and this does represent an entirely indirect digital technique (Brennan, 2002).
Our results showed that both observers had statistically significant correlations with each other. Results from both observers indicated that conventional film was superior to both digital methods. However, results from Observer 1 using the DD system were nearly as good as with film, except for detection of apical lesions. Therefore, film is superior for diagnostic accuracy. Between the digital systems, our study demonstrates that DD is superior to SD. While some studies agree with our findings, others show SD to be superior to DD (Yalcinkaya et al., 2006). Other studies also demonstrate digital systems to be as good as film (Farrier et al., 2009). The digital systems we used were older models and the latest software releases from the manufacturers dramatically enhance their performance compared to film (Yalcinkaya et al., 2006). The new filter functions (caries filters, perio filters, etc.) available with the Vista-scan compared favorably to conventional film (Yalcinkaya et al., 2006). The one used in this study was an older version and this might have negatively influenced the performance SD digital radiography.
Image details of enamel, dentin, DEJ, root canal filling, and simulated apical pathologies are the best possible criteria to compare subjective image quality of dental radiographs. One important criterion not included in our study was the bone trabecular pattern. This was not included because we used plaster models with teeth for the study. The only criterion where film was inferior to the digital systems was the clarity of apical pathology. DD was found to better visualize this for Observer 1. For Observer 2, film was still superior to DD, but the result was not statistically significant. This suggests that DD may be as good as film for apical disease detection.
Digital systems with greater resolution are constantly being released. This may ultimately make digital systems equal to or superior to film (Molander et al., 2004; Akdeniz and Soqur, 2005; Syriopoulos et al., 2000). A previous study compared image quality of digital panoramic images with traditional films using a 5-point scale (Molander et al., 2004). This study concluded that both were comparable. Another study found that subjective image quality of enhanced images from SD digital imaging was superior to both the original SD digital images and conventional film (Akdeniz and Soqur, 2005). Substantial reduction in X-ray doses and the possibility for image manipulation make digital dental radiography the optimal method for most dentists and patients. The latest digital image software packages (e.g., the Kodak RVG 6500 systems) also have improved diagnostic functions due to their increased image resolution (http://www3.carestreamdental.co.uk/ddi/en-GB/rvgimaging/6500#Features). Lastly, since most digital imaging software logs records of image manipulations, they are now accepted in many countries as a proof of insurance.
The selection of equipment is dependent upon the needs of the dentist. DD sensors are bulky, but image formation is instant. SD digital imaging sensors are thin and flexible like film, but they need to be scanned for digitization. In general, if both digital systems are available, the appropriate system can be chosen for each patient (e.g., patients with shallow palates or vestibules, or uncooperative patients can receive SD digital imaging, while other patients can undergo DD imaging). In the future, most dentists may opt for digital systems if they can immediately display images that can be manipulated at later time points while maintaining comparable subjective image qualities to film. When purchasing a digital radiographic system, it is advisable to purchase the latest software and hardware. We recommend that studies with greater number of observers, larger sample sizes, and other statistical methods could help clarify comparisons between different acquisition systems.
Conflict of interest
We confirm that there are no known conflicts of interest associated with this publication and there has been no financial support for this work that could have influenced its outcome.
Footnotes
Peer review under responsibility of King Saud University.
References
- Akdeniz B.G., Soqur E. An ex vivo comparison of conventional and digital radiography for perceived image quality of root fillings. Int. Endod. J. 2005;38(6):397–401. doi: 10.1111/j.1365-2591.2005.00958.x. [DOI] [PubMed] [Google Scholar]
- American Dental Association Council on Scientific Affairs The use of dental radiographs-update and recommendations. J. Am. Dent. Assoc. 2006;vol. 137:1304–1312. doi: 10.14219/jada.archive.2006.0393. [DOI] [PubMed] [Google Scholar]
- Brennan J. An introduction to digital radiography in dentistry. J. Orthod. 2002;29:66–69. doi: 10.1093/ortho/29.1.66. [DOI] [PubMed] [Google Scholar]
- Dunn S.M., Kantor M.L. Digital radiology: facts and fictions. JADA. 1993;124(12):38–47. [PubMed] [Google Scholar]
- Farrier S.L. A comparative study of image quality and radiation exposure for dental radiographs produced using a charge couple device and a phosphor plate system. Int. Endod. J. 2009;42:900–907. doi: 10.1111/j.1365-2591.2009.01593.x. [DOI] [PubMed] [Google Scholar]
- Molander B. Quality of film-based and digital panoramic radiography. Dentomaxillofac. Radiol. 2004;33(1):32–36. doi: 10.1259/dmfr/17777906. [DOI] [PubMed] [Google Scholar]
- Syriopoulos K. Radiographic detection of approximal caries: a comparison of dental films and digital imaging systems. Dentomaxillofac. Radiol. 2000;29(5):312–318. doi: 10.1038/sj/dmfr/4600553. [DOI] [PubMed] [Google Scholar]
- Wakoh Mamoru. Computed dental radiography system versus conventional dental X-ray films for detection of simulated proximal caries. Oral Radiol. 1997;13(2):73–82. [Google Scholar]
- Wenzel A. Digital radiography and caries diagnosis. Dentomaxillofac. Radiol. 1998;27(1):3–11. doi: 10.1038/sj.dmfr.4600321. [DOI] [PubMed] [Google Scholar]
- Wenzel A., Grondahl H.G. Direct digital radiography in the dental office. Int. Dent. J. 1995;45(1):27–34. [PubMed] [Google Scholar]
- White S.C., Pharoah M.J. fifth ed. Mosby; 2004. Oral Radiology: Principles and Interpretation. [Google Scholar]
- <http://www3.carestreamdental.co.uk/ddi/en-GB/rvgimaging/6500#Features> and Benefits.
- Yalcinkaya S. 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;101:643–651. doi: 10.1016/j.tripleo.2005.08.003. [DOI] [PubMed] [Google Scholar]