Influence of dental fillings and tooth type on the performance of a novel artificial intelligence-driven tool for automatic tooth segmentation on CBCT images – A validation study

Fontenele, Rocharles Cavalcante; Gerhardt, Maurício do Nascimento; Pinto, Jáder Camilo; Gerven, Adriaan Van; Willems, Holger; Jacobs, Reinhilde; Freitas, Deborah Queiroz

doi:10.1016/j.jdent.2022.104069

Cited by 24 publications

(10 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 3 compares the segmentation performance of our proposed methods to those proposed by other researchers. The DSC in our study is relatively lower than some previous studies 20,21,27,[29][30][31][32] , in which the DSC ranges from 0.934 31 to 0.97 30 . In our study, we calculated the DSC slice-by-slice and then averaged the DSC of all slices rather than calculated the DSC for the whole CBCT volume as other studies 20,21,23,27,[29][30][31][33][34][35][36][37] .…”

Section: Comparisons Of Ppv Among U-netscontrasting

confidence: 83%

“…Nevertheless, the highest DSC achieved by our 3.5Dv5 U-Net is consistent with other previous studies 23,[33][34][35] , in which the DSC ranges from 0.9 23 to 0.921 33 . Our study achieved an accuracy ranging from 0.997 to 0.999 which is higher than that reported in previous studies 30,36,37 . Our 2D U-Nets achieved a sensitivity ranging from 0.934 to 0.943 which is similar to that (0.91 to 0.94 and 0.932) of Fontenele's study 30 and Lee's study 34 , respectively, and higher than that (0.83) of Shaheen's study 23 .…”

Section: Case Demonstrationcontrasting

confidence: 71%

“…Our study achieved an accuracy ranging from 0.997 to 0.999 which is higher than that reported in previous studies 30,36,37 . Our 2D U-Nets achieved a sensitivity ranging from 0.934 to 0.943 which is similar to that (0.91 to 0.94 and 0.932) of Fontenele's study 30 and Lee's study 34 , respectively, and higher than that (0.83) of Shaheen's study 23 . In addition, our U-Nets with majority voting achieve a PPV ranging from 0.978 to 0.996 which is similar to that (0.98) of Shaheen's study 23 and higher than that (0.904) of Lee's study 34 .…”

Section: Case Demonstrationcontrasting

confidence: 71%

See 2 more Smart Citations

Improving performance of deep learning models using 3.5D U-Net via majority voting for tooth segmentation on cone beam computed tomography

Hsu

Yuh

Lin

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Deep learning allows automatic segmentation of teeth on cone beam computed tomography (CBCT). However, the segmentation performance of deep learning varies among different training strategies. Our aim was to propose a 3.5D U-Net to improve the performance of the U-Net in segmenting teeth on CBCT. This study retrospectively enrolled 24 patients who received CBCT. Five U-Nets, including 2Da U-Net, 2Dc U-Net, 2Ds U-Net, 2.5Da U-Net, 3D U-Net, were trained to segment the teeth. Four additional U-Nets, including 2.5Dv U-Net, 3.5Dv5 U-Net, 3.5Dv4 U-Net, and 3.5Dv3 U-Net, were obtained using majority voting. Mathematical morphology operations including erosion and dilation (E&D) were applied to remove diminutive noise speckles. Segmentation performance was evaluated by fourfold cross validation using Dice similarity coefficient (DSC), accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV). Kruskal–Wallis test with post hoc analysis using Bonferroni correction was used for group comparison. P < 0.05 was considered statistically significant. Performance of U-Nets significantly varies among different training strategies for teeth segmentation on CBCT (P < 0.05). The 3.5Dv5 U-Net and 2.5Dv U-Net showed DSC and PPV significantly higher than any of five originally trained U-Nets (all P < 0.05). E&D significantly improved the DSC, accuracy, specificity, and PPV (all P < 0.005). The 3.5Dv5 U-Net achieved highest DSC and accuracy among all U-Nets. The segmentation performance of the U-Net can be improved by majority voting and E&D. Overall speaking, the 3.5Dv5 U-Net achieved the best segmentation performance among all U-Nets.

show abstract

Section: Comparisons Of Ppv Among U-netscontrasting

confidence: 83%

Section: Case Demonstrationcontrasting

confidence: 71%

Section: Case Demonstrationcontrasting

confidence: 71%

See 1 more Smart Citation

Improving performance of deep learning models using 3.5D U-Net via majority voting for tooth segmentation on cone beam computed tomography

Hsu

Yuh

Lin

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…For instance, Bayraktar et al conducted caries detection on bitewing radiographs, and thanks to the bitewing radiographs they excluded the possibility of any superimpositions as the modality is superior in interproximal caries detection [ 94 ]. Moreover, Fontenele et al conducted a similar study, excluding the cases that have metal/motion artefacts by CBCT for the detection of caries [ 95 ]. Zhu et al conducted tooth segmentation for ectopic eruptions and excluded any cases that had an extraction history, periapical periodontitis, or the presence of cystic lesions;, and they also excluded poor-quality OPGs [ 66 ]; however, in our study, we only excluded the metal/motion artefacts that create a challenging image for interpretation, even for radiologists.…”

Section: Discussionmentioning

confidence: 99%

Automatic Segmentation of Teeth, Crown–Bridge Restorations, Dental Implants, Restorative Fillings, Dental Caries, Residual Roots, and Root Canal Fillings on Orthopantomographs: Convenience and Pitfalls

et al. 2023

View full text Add to dashboard Cite

Background: The aim of our study is to provide successful automatic segmentation of various objects on orthopantomographs (OPGs). Methods: 8138 OPGs obtained from the archives of the Department of Dentomaxillofacial Radiology were included. OPGs were converted into PNGs and transferred to the segmentation tool’s database. All teeth, crown–bridge restorations, dental implants, composite–amalgam fillings, dental caries, residual roots, and root canal fillings were manually segmented by two experts with the manual drawing semantic segmentation technique. Results: The intra-class correlation coefficient (ICC) for both inter- and intra-observers for manual segmentation was excellent (ICC > 0.75). The intra-observer ICC was found to be 0.994, while the inter-observer reliability was 0.989. No significant difference was detected amongst observers (p = 0.947). The calculated DSC and accuracy values across all OPGs were 0.85 and 0.95 for the tooth segmentation, 0.88 and 0.99 for dental caries, 0.87 and 0.99 for dental restorations, 0.93 and 0.99 for crown–bridge restorations, 0.94 and 0.99 for dental implants, 0.78 and 0.99 for root canal fillings, and 0.78 and 0.99 for residual roots, respectively. Conclusions: Thanks to faster and automated diagnoses on 2D as well as 3D dental images, dentists will have higher diagnosis rates in a shorter time even without excluding cases.

show abstract

“…Adding the time component into the virtual patient concept by translating findings from research to the clinic setting also presents a future step towards personalised or precision dentistry (Joda et al, 2020;Lahoud et al, 2022;Schwendicke & Krois, 2022). Furthermore, with AI-assisted segmentation of the CBCT images, segmentation became simpler, increasing future possibilities (Fontenele et al, 2022;Shaheen et al, 2021). Superimposition also presents a crucial step in defining the accuracy of the subsequent analysis (Flügge et al, 2017;Kuralt & Fidler, 2021).…”

Section: Discussionmentioning

confidence: 99%

A Novel Digital Method for Evaluation of Mucosa Thickness Changes in Time: Changes Associated With a Removable Partial Denture

Kuralt¹,

Bukleta²,

Kuhar³

et al. 2022

Image Anal Stereol

View full text Add to dashboard Cite

The aim of the study was to propose a method for three-dimensional evaluation and visualisation of mucosa thickness and mucosa thickness changes and to validate it in four different software implementations. Cone-beam computed tomography (CBCT) and digitised mandibular impression, i.e., optical scan, of five patients treated with a mandibular distal extension removable partial denture were acquired at baseline (T0) and 1-year follow-up (T1). CBCT images were automatically segmented and then superimposed with a corresponding optical scan and within a patient. To obtain mucosa thickness changes in the T0-T1 interval, firstly, the distances between T0 and T1 models were computed for mucosa surface, representing surface changes (dSurface) and bone surface, representing bone changes (dBone). The distances were saved as scalar field values in the mesh model. Finally, the mucosa thickness changes (dMucosa) were calculated by subtracting the corresponding dBone from dSurface values. Distance computation algorithms in four different software were tested and compared. No differences were found between all four tested software (p0.001). Mean (and standard deviation) of median dSurface, dBone, and dMucosa of right and left residual ridge (n=10) was -0.47mm (0.43), -0.44mm (0.62), and 0.00mm (0.35), respectively. High local variability of dBone and dMucosa was found on the colour-coded maps. A novel method facilitates precise three-dimensional evaluation and visualisation of mucosa thickness and thickness changes, regardless of the software used.

show abstract

Influence of dental fillings and tooth type on the performance of a novel artificial intelligence-driven tool for automatic tooth segmentation on CBCT images – A validation study

Cited by 24 publications

References 17 publications

Improving performance of deep learning models using 3.5D U-Net via majority voting for tooth segmentation on cone beam computed tomography

Improving performance of deep learning models using 3.5D U-Net via majority voting for tooth segmentation on cone beam computed tomography

Automatic Segmentation of Teeth, Crown–Bridge Restorations, Dental Implants, Restorative Fillings, Dental Caries, Residual Roots, and Root Canal Fillings on Orthopantomographs: Convenience and Pitfalls

A Novel Digital Method for Evaluation of Mucosa Thickness Changes in Time: Changes Associated With a Removable Partial Denture

Contact Info

Product

Resources

About