Related Experiment Video
Updated: Dec 26, 2025
10:23
Author Spotlight: Three-Dimensional Cephalometric Landmark Annotation Demonstration on Human Cone Beam Computed Tomography Scans
Published on: September 8, 2023
3.5K
Accuracy and reliability of automatic three-dimensional cephalometric landmarking.
G Dot1, F Rafflenbeul2, M Arbotto1
1Institut de Biomecanique Humaine Georges Charpak (IBHGC), Arts et Metiers Institute of Technology, Paris, France.
International Journal of Oral and Maxillofacial Surgery
|March 15, 2020
Summary
Automatic landmarking for 3D craniofacial analysis shows promise, with deep learning algorithms achieving accuracy comparable to manual methods. However, further testing is needed in diverse clinical settings.
Area of Science:
- Radiology
- Medical Imaging
- Computer-Aided Diagnosis
Background:
- Cephalometric analysis is crucial for diagnosing and treating craniofacial abnormalities.
- Manual landmark identification is time-consuming and subject to inter-observer variability.
- Automated methods offer potential for increased efficiency and consistency.
Purpose of the Study:
- To systematically review the accuracy and reliability of automatic landmarking techniques for three-dimensional (3D) craniofacial images.
- To evaluate the performance of various algorithms in cephalometric analysis.
Main Methods:
- Systematic literature search of MEDLINE, Embase, and Web of Science up to March 2019.
- Inclusion of eleven studies reporting on automatic landmarking of computed tomography (CT) or cone beam computed tomography (CBCT) scans.
- Assessment of risk of bias and applicability using the QUADAS-2 tool.
Main Results:
- Eleven studies analyzed, utilizing knowledge-, atlas-, or learning-based algorithms for 2-33 cephalometric landmarks.
- Mean localization errors between manual and automatic landmarks varied from <0.50mm to >5mm.
- Top-performing deep learning algorithms achieved mean errors <2mm, approaching manual operator variability.
Conclusions:
- Deep learning-based automatic landmarking shows potential for accurate and reliable cephalometric analysis.
- Identified risks of bias may lead to overoptimistic performance estimates.
- Robustness requires further validation in challenging clinical environments.
