Evaluation of an artificial intelligence–based algorithm for automated localization of craniofacial landmarks

Abstract

Objectives Due to advancing digitalisation, it is of interest to develop standardised and reproducible fully automated analysis methods of cranial structures in order to reduce the workload in diagnosis and treatment planning and to generate objectifiable data. The aim of this study was to train and evaluate an algorithm based on deep learning methods for fully automated detection of craniofacial landmarks in cone-beam computed tomography (CBCT) in terms of accuracy, speed, and reproducibility. Materials and methods A total of 931 CBCTs were used to train the algorithm. To test the algorithm, 35 landmarks were located manually by three experts and automatically by the algorithm in 114 CBCTs. The time and distance between the measured values and the ground truth previously determined by an orthodontist were analyzed. Intraindividual variations in manual localization of landmarks were determined using 50 CBCTs analyzed twice. Results The results showed no statistically significant difference between the two measurement methods. Overall, with a mean error of 2.73 mm, the AI was 2.12% better and 95% faster than the experts. In the area of bilateral cranial structures, the AI was able to achieve better results than the experts on average. Conclusion The achieved accuracy of automatic landmark detection was in a clinically acceptable range, is comparable in precision to manual landmark determination, and requires less time. Clinical relevance Further enlargement of the database and continued development and optimization of the algorithm may lead to ubiquitous fully automated localization and analysis of CBCT datasets in future routine clinical practice.

Automated localization of mandibular landmarks in the construction of mandibular median sagittal plane

“…3D anatomical landmarks reflect the morphological characteristics of the mandible and are the basis of 3D mandibular analysis. Manual landmark localization relies on doctors' clinical expertise and is tedious [ 24 , 25 ]. The You-Only Look-Once version 3 (YOLOv3) algorithm was applied to 1028 cephalograms to automatically identify 80 cephalometric landmarks with a manual landmark average error of 1.46 ± 2.97 mm [ 25 ].…”

“…In this study, we used the PoseNet algorithm [ 29 ], which has high expansibility and can accurately locate key points in 3D CBCT images without adding additional structure or computations to the model. On average, the error of the 27 landmarks was 1.04 ± 0.28 mm, while the clinical acceptability was 2 mm [ 24 , 25 , 27 ]. The mean error of the central landmarks was 0.63 ± 0.29 mm, which was smaller than that of the lateral landmarks 1.13 ± 0.28 mm, probably because the central landmarks were more accurate and reliable [ 24 ].…”

“…On average, the error of the 27 landmarks was 1.04 ± 0.28 mm, while the clinical acceptability was 2 mm [ 24 , 25 , 27 ]. The mean error of the central landmarks was 0.63 ± 0.29 mm, which was smaller than that of the lateral landmarks 1.13 ± 0.28 mm, probably because the central landmarks were more accurate and reliable [ 24 ]. According to Schlicher et al, landmarks with a distinct anatomical structural contour showed fewer errors than landmarks that were located on curves, which also supported this result [ 28 ].…”

A Comparative Analysis of Artificial Intelligence and Manual Methods for Three-Dimensional Anatomical Landmark Identification in Dentofacial Treatment Planning

“…Another issue in clinical practice involves the amount of supplementary information presented by 3D diagnostics compared to 2D diagnostics, posing significant challenges for clinicians in analysis and treatment planning. Automating the analysis of 3D diagnostics offers a broad spectrum of diagnostic possibilities and enhances accessibility for clinicians, thereby facilitating the transition from 2D to 3D imaging in routine clinical settings [25]. This study aims to validate the efficiency and accuracy of AI by comparing manual and AI tracing using 3D cone beam computed tomography (CBCT).…”

Application of Artificial Intelligence in Orthodontics: Current State and Future Perspectives

“…Thanks to the advancement of computational power, AI has also made progress in three-dimensional (3D) cephalometric landmark detection, and deep learning and CNNs are the most efficient methods [59,60,[71][72][73]. Blum et al utilized a CNN-based model to conduct 3D cephalometric analysis, which yielded a mean error of 2.73 mm and exhibited a 95% reduction in processing time compared with manual annotation [65]. Dot et al proposed a fully CNN, SpatialConfiguration-Net, for the 3D automated detection of 33 landmarks and 15 measurements, achieving superior outcomes.…”