Effect of brightness and contrast variation for detectability of root resorption lesions in digital intraoral radiographs

Assessment of landmark detection in cephalometric radiographs with different conditions of brightness and contrast using the an artificial intelligence software

OBJECTIVES

To evaluate the reliability and reproducibility of an artificial intelligence (AI) software in identifying cephalometric points on lateral cephalometric radiographs considering four settings of brightness and contrast.

METHODS AND MATERIALS

Brightness and contrast of 30 lateral cephalometric radiographs were adjusted into four different settings. Then, the control examiner (ECont), the calibrated examiner (ECal), and the CEFBOT AI software (AIs) each marked 19 cephalometric points on all radiographs. Reliability was assessed with a second analysis of the radiographs 15 days after the first one. Statistical significance was set at p < 0.05.

RESULTS

Reliability of landmark identification was excellent for the human examiners and the AIs regardless of the type of brightness and contrast setting (mean intraclass correlation coefficient >0.89). When ECont and ECal were compared for reproducibility, there were more cephalometric points with significant differences on the x-axis of the image with the highest contrast and the lowest brightness, namely N(p = 0.033), S(p = 0.030), Po(p < 0.001), and Pog'(p = 0.012). Between ECont and AIs, there were more cephalometric points with significant differences on the image with the highest contrast and the lowest brightness, namely N(p = 0.034), Or(p = 0.048), Po(p < 0.001), A(p = 0.042), Pog'(p = 0.004), Ll(p = 0.005), Ul(p < 0.001), and Sn(p = 0.001).

CONCLUSIONS

While the reliability of the AIs for cephalometric landmark identification was rated as excellent, low brightness and high contrast seemed to affect its reproducibility. The experienced human examiner, on the other hand, did not show such faulty reproducibility; therefore, the AIs used in this study is an excellent auxiliary tool for cephalometric analysis, but still depends on human supervision to be clinically reliable.

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.

Influence of the image file format of digital periapical radiographs on the diagnosis of external and internal root resorptions

OBJECTIVES

To evaluate the influence of different image file formats of digital radiographic images on the diagnosis of external (ERR) and internal root resorption (IRR).

MATERIALS AND METHODS

Thirty-four human teeth were selected. For ERR, 20 teeth were used (10 control and 10 with simulated ERR), and for IRR, 14 teeth were used (before and after IRR simulation). Digital periapical radiographs were acquired using the Digora Toto system and exported in four different image file formats: TIFF, BMP, PNG, and JPEG, totaling 192 radiographs. Five examiners evaluated the images using the JPEGView software and scored the detection of ERR or IRR on a 5-point scale. Sensitivity, specificity, accuracy, and the area under the ROC curve were obtained for the diagnosis of ERR and IRR in the different image file formats. Two-way ANOVA compared the diagnostic values between the file formats and the Kappa test assessed intra- and inter-examiner agreement. The significance level was set at 5% (α = 0.05).

RESULTS

The mean values of intra-examiner agreement were substantial (0.740) for ERR and almost perfect (0.836) for IRR and, inter-examiner was fair (0.263) and moderate (0.421), respectively. No statistically significant differences were found among the different file formats for the diagnostic values of ERR and IRR.

CONCLUSION

The file format of digital radiographs does not influence the diagnosis of ERR and IRR.

CLINICAL RELEVANCE

Digital radiographic images may be susceptible to computational factors; however, they can be stored in multiple file formats without affecting the diagnosis of dental root resorptions.

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.