Influence of VistaScan image enhancement filters on diagnosis of simulated periapical lesions on intraoral radiographs

The effect of different spatial resolutions and enhancement filters on radiographic detection of simulated furcation defects with intraoral digital radiography

OBJECTIVE

The objective of this study was to evaluate the effects of spatial resolution and the application of enhancement filters in the diagnosis of simulated furcation defects and image quality with intraoral radiographs.

STUDY DESIGN

Periapical images were acquired with photostimulable phosphor plates of molars in dry skulls (n = 8) and mandibles (n = 10) with 4 stages of furcation defect simulation. The plates were scanned in fast scan and high resolution modes to produce different spatial resolutions. Four image filters were applied. Six observers scored the detection of furcation defects. Mean values of area under the curve in receiver operating characteristic evaluation, accuracy, sensitivity, and specificity were calculated. Mean gray value (brightness), noise, and contrast-to-noise ratio (CNR) were calculated for enamel, dentin, and alveolar bone to evaluate image quality. Analysis of variance compared the values between the different spatial resolutions and filters. The significance of difference was established at P < .05.

RESULTS

There were no significant differences in overall diagnostic values comparing image spatial resolutions and filters. Diagnostic outcomes were significantly better for the largest defects than the smallest lesions but were generally poor in detecting lesions. All structures showed greater brightness in high resolution. . Noise was greater in all structures with all enhancement filters except inversion. Dentin and alveolar bone exhibited more noise and lower CNR in high resolution.

CONCLUSIONS

Varying spatial resolution and applying enhancement filters did not significantly affect the diagnosis of furcation defects.

Does enhancement filter application increase the diagnostic accuracy of misfit detection at the implant-prosthesis interface?

STATEMENT OF PROBLEM

Misfits at the implant-prosthesis interface may increase complications in dental implants and affect peri-implant tissue health. Periapical radiographs are the most used imaging examinations for detecting misfits at the implant-prosthesis interface, although digital systems have largely replaced film-based radiographs. Whether postprocessing tools such as enhancement filters assist diagnosis by highlighting misfits is unclear.

PURPOSE

The purpose of this in vitro study was to assess the influence of enhancement filter application in the diagnostic accuracy of misfit detection at the implant-prosthesis interface.

MATERIAL AND METHODS

A total of 32 dental implants were placed in dry human mandibles. A polyester strip was inserted at the implant-prosthesis interface to simulate a 50-μm misfit; prosthetic crowns installed directly on the implant platforms were used as controls. Standard paralleling periapical images were acquired by using a semidirect system (photostimulable phosphor plate) with the application of Highlight, Invert, and Colorization filters, as well as a direct system (metal oxide complementary semiconductor sensor) with filters Sharpness 3, Invert, and Pseudocolorization. Oral radiologists evaluated the images with and without the application of filters. The areas under the receiver operating characteristics curves (Az values), sensitivity, specificity, accuracy, positive predictive value, and negative predictive values were calculated. The Az values were compared with the receiver operating characteristic (ROC) curves comparison test of the Epidat 3.1 software (α=.05).

RESULTS

Although images without filter application presented descriptively higher diagnostic values than those with filter application, the Az values for images with and without filter application in both semidirect and direct systems showed no significant differences (P>.05).

CONCLUSIONS

Enhancement filter application did not significantly influence the diagnostic accuracy of misfit detection at the implant-prosthesis interface.

The impact of digital filters on the diagnosis of simulated root resorptions in digital radiographic systems

OBJECTIVES

To evaluate the influence of digital filters of intraoral radiographic systems on the diagnosis of simulated internal and external root resorptions and image quality.

MATERIALS AND METHODS

Internal root resorption (IRR) and external root resorption (ERR) were simulated in 34 teeth. For image acquisition, two radiographic systems were used: Digora Toto and VistaScan. All filters available in these systems were applied. Three observers scored the detection of root resorptions in a 5-point scale. The noise and the contrast-to-noise ratio (CNR) were calculated. The area under ROC curve, sensitivity, specificity, and accuracy were obtained. One-way ANOVA with Tukey's post hoc tests compared the diagnostic values, noise, and CNR between the filters (α = 0.05).

RESULTS

For ERR, there were no significant differences in diagnostic values between the filters tested for both systems. For IRR, Original and Noise Reduction filters presented higher sensitivity than the Sharpen2 filter for images from Digora Toto, with no differences between the other groups. For VistaScan, there were no significant differences of diagnostic values between the groups studied. Noise values differed among the filters of both systems. The CNR of the filters differed only for the bone region for Digora Toto, while for VistaScan, both tooth and bone regions differed.

CONCLUSIONS

Despite promoting changes in pixel intensities and affecting the noise level of the radiographic images, the digital filters of Digora Toto and VistaScan systems do not affect the diagnosis of internal or external root resorptions.

CLINICAL RELEVANCE

Digital filters are common tools in digital radiographic systems and may be used by the professional without impairment in root resorptions diagnosis.

Exploring digital filters for internal root resorption: how can we improve the diagnosis of small lesions?

OBJECTIVES

This study aimed to evaluate the impact of enhancement filters in detecting small simulated internal root resorptions (IRR).

METHODS:

Forty-two extracted human teeth were sectioned, connected, and stored in a dry human jaw and X-rayed with photostimulable phosphor plates (PSPs), composing the control group (CG). In the middle-third of the root canals, IRR lesions were simulated using Da Silveira protocol. Later, the specimens were X-rayed to create the test group (TG). All images acquired were exported with seven enhancement filters plus the original image. Three examiners used a five-point Likert scale to evaluate the images regarding the presence/absence of IRR. Diagnostic efficacy was assessed from sensitivity and specificity results. Comparison among filters was performed by using receiver operating characteristic (ROC) curve analysis.

RESULTS:

Moderate values of Kappa interexaminer (0.403-0.620) and high values of Kappa intraexaminer (0.757-0.915) were observed. The best performance occurred in the CG (p < 0.05). Original images presented the greatest sensitivity and area under the ROC curve (0.595-0.750), while the Endo filter presented the greatest specificity (0.952). Inversion and Pseudo-3D images produced the greatest doubt in the diagnosis, significant for CG with the Pseudo-3D filter (p < 0.05).

CONCLUSIONS:

The Original and 'Endo' filters should be chosen as it offers greater diagnostic ability and allows more confidence during the evaluation.

Artificial intelligence for detection of periapical lesions on intraoral radiographs: Comparison between convolutional neural networks and human observers

OBJECTIVE

The aim of this study was to compare the diagnostic performance of convolutional neural networks (CNNs) with the performance of human observers for the detection of simulated periapical lesions on periapical radiographs.

STUDY DESIGN

Ten sockets were prepared in bovine ribs. Periapical defects of 3 sizes were sequentially created. Periapical radiographs were acquired of each socket with no lesion and with each lesion size with a photostimulable storage phosphor system. Radiographs were evaluated with no filter and with 6 image filter settings. A CNN architecture was set up using Keras-TensorFlow. Separate CNNs were evaluated for randomly sampled training/validation data and for data split up by socket (5-fold cross-validation) and filter (7-fold cross-validation). CNN performance on validation data was compared with that of 3 oral radiologists for sensitivity, specificity, and area under the receiver operating characteristic curve (ROC-AUC).

RESULTS

Using random sampling, the CNN showed perfect accuracy for the validation data. When data were split up by socket, the mean sensitivity, specificity, and ROC-AUC values were 0.79, 0.88, and 0.86, respectively; when split up by filter, they were 0.87, 0.98, and 0.93, respectively. For radiologists, the values were 0.58, 0.83, and 0.75, respectively.

CONCLUSIONS

CNNs show promise in periapical lesion detection. The pretrained CNN model yielded in this study can be used for further training on larger samples and/or clinical radiographs.

Does dose optimisation in digital panoramic radiography affect diagnostic performance?

OBJECTIVES

To compare the overall diagnostic performance of digital panoramic radiographs obtained with low-dose protocols and to estimate the absorbed dose in the head and neck.

MATERIALS AND METHODS

Forty-eight panoramic radiographs were obtained from eight imaging phantoms using six exposure protocols of progressively lower tube voltages (kVp) and currents (mA), as follows: (1) 70 kVp and 12.5 mA, (2) 66 kVp and 10 mA, (3) 66 kVp and 8 mA, (4) 66 kVp and 5 mA, (5) 66 kVp and 4 mA and (6) 66 kVp and 3.2 mA. Five oral radiologists independently evaluated the images and reported all detectable radiographic findings. Intra-examiner reproducibility was assessed by re-evaluation of 25% of the images. The data were analysed using the McNemar and weighted Kappa tests. Absorbed doses of the six protocols were obtained from thermoluminescent dosimeters placed inside a Rando phantom and compared using one-way ANOVA with post hoc Tukey (α = 0.05).

RESULTS

The overall diagnostic performance of panoramic radiographs obtained with low-dose protocols did not differ from that of panoramic radiographs obtained with the highest dose (p > 0.05). Moreover, substantial agreement was observed between all protocols. Protocol 1 resulted in the highest absorbed dose and protocols 4, 5 and 6 in the lowest absorbed doses, with the difference being significant (p ≤ 0.05).

CONCLUSION

Although digital panoramic radiography is considered a relatively low-dose examination, the radiation dose can be further reduced without negatively affecting its overall diagnostic performance.

CLINICAL RELEVANCE

Considering the risks associated with X-rays, digital panoramic radiographs can be obtained at even lower exposure levels.

Effect of digital enhancement on the radiographic assessment of vertical root fractures in the presence of different intracanal materials: an in vitro study

OBJECTIVES

To evaluate the effect of enhancement tools of intraoral digital radiographs on the assessment of vertical root fracture (VRF) and to quantify the resultant image noise.

MATERIALS AND METHODS

Thirty single-rooted human teeth (15 control and 15 fractured) were each radiographed in four intracanal conditions: no filling, gutta-percha, metal post, and fiberglass post, totaling 120 original images. Two filters were applied to the original images-Sharpen filter (SF) and Edge Enhancement filter (EE), and brightness and contrast were adjusted in four combinations (B&C1 to 4), resulting in 840 images. Five oral radiologists analyzed the images for VRF detection. Pixel intensity was obtained in two regions from the radiographs. Diagnostic values were calculated and compared by two-way ANOVA, and the SD values of pixel intensity values were compared by one-way ANOVA (α = 0.05).

RESULTS

There were no significant differences in accuracy for VRF detection between the experimental groups (p > 0.05). Teeth with metal post presented the lowest sensitivity (p < 0.05) for all experimental conditions, except for SF and EE (p > 0.05). B&C2, B&C3, and B&C4 had higher specificity than SF (p ≤ 0.05) for all intracanal conditions. Analysis of pixel intensity showed that all enhanced images presented statistically significant higher noise compared to those of the original images (p ≤ 0.05).

CONCLUSION

Digital enhancement tools in digital radiography increase image noise; however, they can be used without compromising VRF detection.

CLINICAL RELEVANCE

The use of digital enhancement does not impair the detection of VRF and, therefore, can be applied for this purpose according to the observer preference.

Quantitative analysis of metal artifact reduction using the auto-edge counting method in cone-beam computed tomography

The metal artifact reduction (MAR) algorithm is used in most CBCT unit to reduce artifact from various dental materials. The performance of MAR program of a CBCT unit according to the dental material type under different imaging mode was evaluated as introducing automatic quantification of the amount of artifact reduced. Four customized phantoms with different dental prostheses (amalgam, gold, porcelain-fused-metal, zirconia) underwent CBCT scanning with and without the MAR option. The imaging was performed under varied scanning conditions; 0.2 and 0.3 mm³ voxel sizes; 70 and 100 kVp. The amount of artifacts reduced by each prosthesis and scanning mode automatically counted using canny edge detection in MATLAB, and statistical analysis was performed. The overall artifact reduction ratio was ranged from 17.3% to 55.4%. The artifact caused by the gold crown was most effectively reduced compared to the other prostheses (p < 0.05, Welch’s ANOVA analysis). MAR showed higher performance in smaller voxel size mode for all prostheses (p < 0.05, independent t-test). Automatic quantification efficiently evaluated MAR performance in CBCT image. The impact of MAR was different according to the prostheses type and imaging mode, suggesting that thoughtful consideration is required when selecting the imaging mode of CBCT.

Detection of Simulated Periapical Lesion in Intraoral Digital Radiography with Different Brightness and Contrast

Objective:

To assess the detection of simulated periapical lesions in digital intraoral radiography with different levels of brightness and contrast combinations, and to investigate the observers’ preference of image quality for this diagnostic task.

Methods:

Digital radiographs were acquired prior to periapical lesion simulation and after each one of four defects enlargement. Original images were adjusted in 4 brightness and contrast combinations. Five observers evaluated the images according to the presence of periapical lesion on a 5-point scale. In a second moment, the observers ordinated the images subjectively, according to quality, from the best to the worst to detect the bone defect. The area under the receiver operating characteristic curve was calculated for the diagnostic values and compared by two-way ANOVA. The significance level was set at 5% (P<0.05).

Results:

No differences were found between the diagnostic values of the five combinations of brightness and contrast (P>0.05). The overall results showed low values of area under the Receiver Operating Characteristic (ROC) curve and sensitivity of the periapical radiography in the detection of periapical lesions of sizes from 1 to 3, which rose substantially in size 4. For image quality, combinations with the lowest brightness and highest contrast were preferred by the observers in 58% of the cases.

Conclusion:

Brightness and contrast adjustments do not influence the detection of simulated periapical lesions in digital intraoral radiography. Lower brightness and higher contrast images were preferred for this diagnostic task.