TY - JOUR
T1 - Beam hardening correction tool improves the diagnosis of incipient caries lesions in Micro-ct images
AU - Rovaris, Karla
AU - Melo, Saulo Leonardo Sousa
AU - Queiroz, Polyane Mazucatto
AU - Loch, Carolina
AU - Schwass, Donald R.
AU - Haiter-Neto, Francisco
N1 - Publisher Copyright:
© 2019 The Authors.
PY - 2019
Y1 - 2019
N2 - Objective: To evaluate the effect of Beam Hardening Correction tool (BHC) on Micro-CT images used in the diagnosis of incipient caries lesions. Methods: Six posterior teeth, with 6 sound proximal surfaces and 6 natural white spot lesions, were scanned using a SkyScan1174 device. Each tooth was imaged 16 times in various acquisition parameters, and reconstructed 3 times applying different levels of BHC (0, 15 and 30%). The 288 datasets were analyzed by three examiners. Gold-standard was obtained by microscopic analyses under polarized light. Area under the ROC curve was obtained for each protocol and the BHC parameter was isolated to individually observe the influence of such parameters of reconstruction. Results: Microtomographic images were accurate for the diagnosis of incipient caries lesions even when the BHC tool was not applied. Accuracy was perfect in 50 % of the protocols without BHC. When adding the artefact correction, the number of protocols that reached perfect accuracy increased to 81.25%. No difference was found between the protocols that used 15 or 30% BHC. Statistically significant differences were found only in four comparisons, and on those the images that had BHC were more accurate. Two non-BHC protocols reached accuracy scores of 0.653 and 0.881, and were improved to perfect accuracy with BHC application. conclusion: The use of the BHC tool should be encouraged when using Micro-CT to detect incipient caries lesions, especially when low-quality parameters aiming less acquisition and processing time are used.
AB - Objective: To evaluate the effect of Beam Hardening Correction tool (BHC) on Micro-CT images used in the diagnosis of incipient caries lesions. Methods: Six posterior teeth, with 6 sound proximal surfaces and 6 natural white spot lesions, were scanned using a SkyScan1174 device. Each tooth was imaged 16 times in various acquisition parameters, and reconstructed 3 times applying different levels of BHC (0, 15 and 30%). The 288 datasets were analyzed by three examiners. Gold-standard was obtained by microscopic analyses under polarized light. Area under the ROC curve was obtained for each protocol and the BHC parameter was isolated to individually observe the influence of such parameters of reconstruction. Results: Microtomographic images were accurate for the diagnosis of incipient caries lesions even when the BHC tool was not applied. Accuracy was perfect in 50 % of the protocols without BHC. When adding the artefact correction, the number of protocols that reached perfect accuracy increased to 81.25%. No difference was found between the protocols that used 15 or 30% BHC. Statistically significant differences were found only in four comparisons, and on those the images that had BHC were more accurate. Two non-BHC protocols reached accuracy scores of 0.653 and 0.881, and were improved to perfect accuracy with BHC application. conclusion: The use of the BHC tool should be encouraged when using Micro-CT to detect incipient caries lesions, especially when low-quality parameters aiming less acquisition and processing time are used.
KW - Dental caries
KW - Dentistry
KW - Diagnostic imaging
KW - Radiology
KW - X-ray microtomography
UR - http://www.scopus.com/inward/record.url?scp=85058941163&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85058941163&partnerID=8YFLogxK
U2 - 10.1259/dmfr.20180123
DO - 10.1259/dmfr.20180123
M3 - Article
C2 - 30084261
AN - SCOPUS:85058941163
SN - 0250-832X
VL - 48
JO - Dentomaxillofacial Radiology
JF - Dentomaxillofacial Radiology
IS - 1
M1 - 20180123
ER -