Cephalometric Landmark Localization Model Based on Polarized Self-Attention Mechanism

Shuaichao Feng; Xinpeng Miao; Shukui Ma; Fei Ma; Guangping Zhuo

doi:10.62051/ijcsit.v5n1.12

Authors

Shuaichao Feng
Xinpeng Miao
Shukui Ma
Fei Ma
Guangping Zhuo

DOI:

https://doi.org/10.62051/ijcsit.v5n1.12

Keywords:

Orthodontics, Cephalometric Landmark, DLA-34, Polarized Self-Attention Mechanism

Abstract

Precise localization of cephalometric landmarks is crucial in the fields of orthodontics and craniofacial surgery. Traditional manual cephalometric analysis and computer-aided cephalometric analysis have significant drawbacks, including large errors, low accuracy, and being time-consuming. To achieve efficient and accurate localization of cephalometric landmarks, this study proposes a detection algorithm, CenterNet-PSA, which integrates the Polarized Self-Attention Mechanism. The algorithm first uses a pre-trained DLA-34 as the feature extraction network to extract features, and then incorporates the polarized self-attention mechanism into the DLA-34 feature extraction network to weight the spatial and channel information of the image, thereby improving the accuracy of landmark detection. Finally, the model achieves a mean radial error (MRE) of 1.07mm and a success detection rate (SDR) of 88.14% within a 2mm error range on the ISBI 2015 Grand Challenge cephalometric X-ray test dataset. Compared to other detection methods, CenterNet-PSA can achieve efficient and accurate localization of cephalometric landmarks, meeting the needs of clinical medicine.

Downloads

Download data is not yet available.

References

[1] GRAU V, ALCANIZ M, JUAN M, et al. Automatic localization of cephalometric landmarks [J]. Journal of Biomedical Informatics, 2001, 34(3): 146-56.

[2] KEUSTERMANS J, MOLLEMANS W, VANDERMEULEN D, et al. Automated cephalometric landmark identification using shape and local appearance models; proceedings of the 2010 20th International Conference on Pattern Recognition, F, 2010 [C]. IEEE.

[3] IBRAGIMOV B, LIKAR B, PERNUS F, et al. Computerized cephalometry by game theory with shape-and appearance-based landmark refinement; proceedings of the Proceedings of International Symposium on Biomedical imaging (ISBI), F, 2015 [C].

[4] OKTAY O, BAI W, GUERRERO R, et al. Stratified decision forests for accurate anatomical landmark localization in cardiac images [J]. IEEE transactions on medical imaging, 2016, 36(1): 332-42.

[5] CRIMINISI A, SHOTTON J, BUCCIARELLI S. Decision forests with long-range spatial context for organ localization in CT volumes; proceedings of the Medical Image Computing and Computer-Assisted Intervention (MICCAI), F, 2009 [C]. Citeseer.

[6] LINDNER C, COOTES T F. Fully automatic cephalometric evaluation using random forest regression-voting; proceedings of the IEEE International Symposium on Biomedical Imaging (ISBI) 2015–Grand Challenges in Dental X-ray Image Analysis–Automated Detection and Analysis for Diagnosis in Cephalometric X-ray Image, F, 2015 [C].

[7] LEE H, PARK M, KIM J. Cephalometric landmark detection in dental x-ray images using convolutional neural networks; proceedings of the Medical imaging 2017: Computer-aided diagnosis, F, 2017 [C]. SPIE.

[8] ZHONG Z, LI J, ZHANG Z, et al. An attention-guided deep regression model for landmark detection in cephalograms; proceedings of the Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13–17, 2019, Proceedings, Part VI 22, F, 2019 [C]. Springer.

[9] QIAN J, CHENG M, TAO Y, et al. CephaNet: An improved faster R-CNN for cephalometric landmark detection; proceedings of the 2019 IEEE 16th international symposium on biomedical imaging (ISBI 2019), F, 2019 [C]. IEEE.

[10] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation; proceedings of the Proceedings of the IEEE conference on computer vision and pattern recognition, F, 2014 [C].

[11] DAI X, ZHAO H, LIU T, et al. Locating anatomical landmarks on 2D lateral cephalograms through adversarial encoder-decoder networks [J]. IEEE Access, 2019, 7: 132738-47.

[12] CRESWELL A, WHITE T, DUMOULIN V, et al. Generative adversarial networks: An overview [J]. IEEE signal processing magazine, 2018, 35(1): 53-65.

[13] LIU H, LIU F, FAN X, et al. Polarized self-attention: Towards high-quality pixel-wise regression [J]. arXiv preprint arXiv:210700782, 2021.

[14] ZHANG Q. A novel ResNet101 model based on dense dilated convolution for image classification [J]. SN Applied Sciences, 2022, 4: 1-13.

[15] SUSANTO Y, LIVINGSTONE A G, NG B C, et al. The hourglass model revisited [J]. IEEE Intelligent Systems, 2020, 35(5): 96-102.

[16] YU F, WANG D, SHELHAMER E, et al. Deep layer aggregation; proceedings of the Proceedings of the IEEE conference on computer vision and pattern recognition, F, 2018 [C].