File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: 78371040-2124 - Artificial Intelligence Driven CBCT Segmentation and 3D Modelling of Anterior Maxillary Wall for Computer-Assisted Surgery: A Comparison of Multiple Algorithms

Title78371040-2124 - Artificial Intelligence Driven CBCT Segmentation and 3D Modelling of Anterior Maxillary Wall for Computer-Assisted Surgery: A Comparison of Multiple Algorithms
Authors
Issue Date1-Jul-2025
Abstract

Artificial Intelligence (AI) has developed rapidly in recent years, and its application in the dental field is expected to expand further. In craniomaxillofacial surgeries, accurate image segmentation is the cornerstone for precise treatment and successful surgical planning. AI algorithms on segmentation have been proposed to improve both the accuracy and efficiency of clinical use. The aim of this study is to compare the accuracy and quality of CBCT segmentation between AI-driven segmentation, 3D modelling, thresholding, and manual segmentation on the anterior maxillary wall.

A total of 20 Cone Beam Computed Tomography (CBCT) scans were selected to perform manual segmentation (reference), thresholding segmentation, 3D modelling, and AI-driven segmentation on the anterior maxillary wall using commercially available software (Mimics, 3-Matics, Blue Sky Plan, and Mimics Viewer). The accuracy of the segmentation was evaluated by (1) DICE Similarity Coefficient (DSC), (2) Hausdorff Distance (HD), and (3) comparing anterior wall thickness of segmentation models with ground truth (manual segmentation).

Additionally, qualitative analysis was carried out with questionnaires completed by experts to assess the suitability for Virtual Surgical Planning (VSP) as well as structural consistency, completeness, accuracy, and smoothness. In the present study, 3D modelling with 3-Matics showed the best performance in both accuracy and quality assessment on the anterior maxillary wall. AI-driven segmentation (Mimics Viewer), designed for craniomaxillofacial surgery, showed satisfactory segmentation accuracy with a DSC of 0.90 ± 0.03. AI segmentation also demonstrated better clinical applicability compared to manual segmentation and 3D modelling, with improved reproducibility, reduced operator variability, and better time efficiency.


Persistent Identifierhttp://hdl.handle.net/10722/358820

 

DC FieldValueLanguage
dc.contributor.authorLo, H-
dc.contributor.authorYang, Weifa-
dc.date.accessioned2025-08-13T07:48:14Z-
dc.date.available2025-08-13T07:48:14Z-
dc.date.issued2025-07-01-
dc.identifier.urihttp://hdl.handle.net/10722/358820-
dc.description.abstract<p>Artificial Intelligence (AI) has developed rapidly in recent years, and its application in the dental field is expected to expand further. In craniomaxillofacial surgeries, accurate image segmentation is the cornerstone for precise treatment and successful surgical planning. AI algorithms on segmentation have been proposed to improve both the accuracy and efficiency of clinical use. The aim of this study is to compare the accuracy and quality of CBCT segmentation between AI-driven segmentation, 3D modelling, thresholding, and manual segmentation on the anterior maxillary wall.</p><p>A total of 20 Cone Beam Computed Tomography (CBCT) scans were selected to perform manual segmentation (reference), thresholding segmentation, 3D modelling, and AI-driven segmentation on the anterior maxillary wall using commercially available software (Mimics, 3-Matics, Blue Sky Plan, and Mimics Viewer). The accuracy of the segmentation was evaluated by (1) DICE Similarity Coefficient (DSC), (2) Hausdorff Distance (HD), and (3) comparing anterior wall thickness of segmentation models with ground truth (manual segmentation).</p><p>Additionally, qualitative analysis was carried out with questionnaires completed by experts to assess the suitability for Virtual Surgical Planning (VSP) as well as structural consistency, completeness, accuracy, and smoothness. In the present study, 3D modelling with 3-Matics showed the best performance in both accuracy and quality assessment on the anterior maxillary wall. AI-driven segmentation (Mimics Viewer), designed for craniomaxillofacial surgery, showed satisfactory segmentation accuracy with a DSC of 0.90 ± 0.03. AI segmentation also demonstrated better clinical applicability compared to manual segmentation and 3D modelling, with improved reproducibility, reduced operator variability, and better time efficiency.</p>-
dc.languageeng-
dc.relation.ispartof26th International Conference on Oral and Maxillofacial Surgery (ICOMS) (22/05/2025-25/05/2025, Singapore)-
dc.title78371040-2124 - Artificial Intelligence Driven CBCT Segmentation and 3D Modelling of Anterior Maxillary Wall for Computer-Assisted Surgery: A Comparison of Multiple Algorithms-
dc.typeConference_Paper-
dc.identifier.doi10.1016/j.ijom.2025.04.1001-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats