Application research of YOLOv5s_CBCT model in measuring alveolar bone height and width in CBCT images of missing teeth

doi:10.12337/zgkqzzxzz.2025.10.002

Abstract

Abstract: Objective To explore the measurement value of the artificial intelligence model based on YOLOv5s (YOLOv5s_CBCT) for the height and width of alveolar bone at the location of missing teeth, and to provide a precise quantitative tool for preoperative planning of oral implants. Methods Based on the faster implementation of CSP bottleneck with 2 convolutions (C2f), and large selective kernel (LSK) modules, the C2f_CBCT and LSK_CBCT modules were constructed. Using modules and structures such as space-to-depth (SPD), ghost convolution (GhostConv), C3 with ghost convolution (C3Ghost) and bidirectional feature pyramid network (BiFPN), the YOLOv5s_CBCT model was proposed, and the test images were input into the YOLOv5s_CBCT model after training. The predicted average values of alveolar bone height and width for the four implant areas by the model were respectively obtained and error analysis was conducted. Bidirectional analysis of variance was used to explore the significant factors influencing the measurement results, and Bland-Altman consistency analysis and Tukey test were further adopted to strengthen and improve the conclusions of bidirectional analysis of variance. Results (1) Model comparison experiments showed that the proposed model achieved higher detection accuracy with fewer parameters and lower computational complexity. (2) The maximum average error of alveolar bone height detection in the four implant regions was 0.15 mm, and the minimum was 0.03 mm. The maximum average error of width detection was 0.26 mm, and the minimum was 0.10 mm. (3) The results of the two-way ANOVA indicated that the detection results of alveolar bone height and width were significantly affected by the implant region, but not by the measurement method. (4) Bland-Altman consistency analysis further demonstrated good agreement between manual measurements and model prediction. (5) The Tukey test identified the grouping of implant regions that had a significant impact on the measurement results. Conclusion The overall experimental results demonstrated that the measurement errors of the model were all within an acceptable range. The measurement results have high clinical reference value, and the proposed YOLOv5s_CBCT model can be conveniently deployed on medical instruments.

Key words: Dental implants, Alveolar bone, Cone beam computed tomography, Artificial intelligence, YOLOv5s

Liu Wenqi, Wei Wenquan, Mo Hongbing, Shen Xiaorong, Zang Yixin. Application research of YOLOv5s_CBCT model in measuring alveolar bone height and width in CBCT images of missing teeth[J]. Chinese Journal of Oral Implantology, 2025, 30(5): 429-439.

References

[1] Gaviria L, Salcido JP, Guda T, et al.Current trends in dental implants[J]. J Korean Assoc Oral Maxillofac Surg, 2014,40(2):50-60. DOI: 10.5125/jkaoms.2014.40.2.50.
[2] Addy LD.An introduction to dental implants[J]. Br Dent J, 2024,236(10):753-757. DOI: 10.1038/s41415-024-7430-8.
[3] Rushinek H, Cohen A, Casap N, et al.The effect of implant-associated factors on the long-term outcomes of dental implants[J]. Oral Maxillofac Surg Clin North Am, 2025,37(1):51-63. DOI: 10.1016/j.coms.2024.08.009.
[4] Correa LR, Spin-Neto R, Stavropoulos A, et al.Planning of dental implant size with digital panoramic radiographs, CBCT-generated panoramic images, and CBCT cross-sectional images[J]. Clin Oral Implants Res, 2014,25(6):690-695. DOI: 10.1111/clr.12126.
[5] Worthington P, Rubenstein J, Hatcher DC.The role of cone-beam computed tomography in the planning and placement of implants[J]. J Am Dent Assoc, 2010,141(Suppl 3): S19-S24. DOI: 10.14219/jada.archive.2010.0358.
[6] Gaêta-Araujo H, Oliveira-Santos N, Mancini A, et al.Retrospective assessment of dental implant-related perforations of relevant anatomical structures and inadequate spacing between implants/teeth using cone-beam computed tomography[J]. Clin Oral Investig, 2020,24(9):3281-3288. DOI: 10.1007/s00784-020-03205-8.
[7] Siessegger M, Schneider BT, Mischkowski RA, et al.Use of an image-guided navigation system in dental implant surgery in anatomically complex operation sites[J]. J Craniomaxillofac Surg, 2001,29(5):276-281. DOI: 10.1054/jcms.2001.0242.
[8] Bae M, Park JW, Kim N.Fully automated estimation of arch forms in cone-beam CT with cubic b-spline approximation: evaluation of digital dental models with missing teeth[J]. Comput Biol Med, 2021,131:104256. DOI: 10.1016/j.compbiomed.2021.104256.
[9] Kurt Bayrakdar S, Orhan K, Bayrakdar IS, et al.A deep learning approach for dental implant planning in cone-beam computed tomography images[J]. BMC Med Imaging, 2021,21(1):86. DOI: 10.1186/s12880-021-00618-z.
[10] Rezallah N, Luke AM.Evaluating micro-computed tomography in dental implant osseointegration: a systematic review and meta-analysis[J]. Acad Radiol, 2025,32(2):1086-1099. DOI: 10.1016/j.acra.2024.09.011.
[11] Zakirov A, Ezhov M, Gusarev M, et al.Dental pathology detection in 3D cone-beam CT[J]. arXiv, 2018:1810.10309.DOI: 10.48550/arXiv.1810.10309.
[12] Chartrand G, Cheng PM, Vorontsov E, et al.Deep learning: a primer for radiologists[J]. Radiographics, 2017, 37(7): 2113-2131. DOI: 10.1148/rg.2017170077.
[13] Chan HP, Samala RK, Hadjiiski LM, et al.Deep learning in medical image analysis[M]. Switzerland, Cham: Springer International publishing, 2020:3-21. DOI: 10.1016/b978-0-443-15452-2.00014-5.
[14] Sun ML, Liu Y, Liu G, et al.Application of machine learning to stomatology: a comprehensive review[J]. IEEE Access, 2020, 8: 184360-184374. DOI: 10.1109/ACCESS.2020.3028600.
[15] Hu X, Liu Y, Zhao Z, et al.Real-time detection of uneaten feed pellets in underwater images for aquaculture using an improved YOLO-V4 network[J]. Comput Electron Agric, 2021, 185: 106135. DOI: 10.1016/j.compag.2021.106135.
[16] Kwak GH, Kwak EJ, Song JM, et al.Automatic mandibular canal detection using a deep convolutional neural network[J]. Sci Rep, 2020,10(1):5711. DOI: 10.1038/s41598-020-62586-8.
[17] Tuzoff DV, Tuzova LN, Bornstein MM, et al.Tooth detection and numbering in panoramic radiographs using convolutional neural networks[J]. Dentomaxillofac Radiol, 2019,48(4): 20180051. DOI: 10.1259/dmfr.20180051.
[18] Muramatsu C, Kutsuna S, Takahashi R, et al.Tooth numbering in cone-beam CT using a relation network for automatic filing of dentition charts[J]. Med Imaging 2020: Imaging Informatics for Healthcare, Research, and Applications, 2020,11318:162-167. DOI: 10.1117/12.2550933.
[19] Orhan K, Bayrakdar IS, Ezhov M, et al.Evaluation of artificial intelligence for detecting periapical pathosis on cone-beam computed tomography scans[J]. Int Endod J, 2020,53(5):680-689. DOI: 10.1111/iej.13265.
[20] Riberti N, Furlani M, D’Amico E, et al. Deep learning for microstructural characterization of synchrotron radiation-based collagen bundle imaging in peri-implant soft tissues[J]. Appl Sci, 2023, 13(7): 4423. DOI: 10.3390/app13074423.
[21] Elgarba BM, Van Aelst S, Swaity A, et al.Deep learning-based segmentation of dental implants on cone-beam computed tomography images: a validation study[J]. J Dent, 2023,137:104639. DOI: 10.1016/j.jdent.2023.104639.
[22] Lee SJ, Chung D, Asano A, et al.Diagnosis of tooth prognosis using artificial intelligence[J]. Diagnostics (Basel), 2022,12(6):1422. DOI: 10.3390/diagnostics 12061422.
[23] Kong HJ, Yoo JY, Lee JH, et al.Performance evaluation of deep learning models for the classification and identification of dental implants[J]. J Prosthet Dent, 2025,133(6):1521-1527. DOI: 10.1016/j.prosdent.2023.07.009.
[24] Miki Y, Muramatsu C, Hayashi T, et al.Classification of teeth in cone-beam CT using deep convolutional neural network[J]. Comput Biol Med, 2017,80:24-29. DOI: 10.1016/j.compbiomed.2016.11.003.
[25] Jocher G. Ultralytics/YOLOv5 [EB/OL]. (2022-11-22) [2023-06-10]. https://github.com/ultralytics/yolov5.
[26] Sunkara R, Luo T.No more strided convolutions or pooling: a new CNN building block for low-resolution images and small objects[C]//Proc ECML PKDD. Cham: Springer, 2022: 443-459. DOI:10.1007/978-3-031-26409-2_27.
[27] Varghese R, Sambath M.YOLOv8: a novel object detection algorithm with enhanced performance and robustness[C]//2024 IEEE ADICS. Chennai:IEEE. 2024: 1-6. DOI:10.1109/ADICS58448.2024.10533619.
[28] Li Y, Hou Q, Zheng Z, et al.Large selective kernel network for remote sensing object detection[C]//Proc IEEE/CVF ICCV. Paris:IEEE. 2023:16794-16805.
[29] Han K, Wang Y, Tian Q, et al.GhostNet: more features from cheap operations[C]//Proc IEEE/CVF CVPR. Seattle:IEEE. 2020: 1580-1589.
[30] Wang A, Chen H, Lin Z, et al.RepVIT: revisiting mobile CNN from ViT perspective[C]//Proc IEEE/CVF CVPR.Seattle:IEEE. 2024: 15909-15920.
[31] Hu J, Shen L, Sun G.Squeeze-and-excitation networks[C]//Proc IEEE CVPR. Salt Lake City:IEEE. 2018: 7132-7141.
[32] Liu S, Qi L, Qin H, et al.Path aggregation network for instance segmentation[C]//Proc IEEE CVPR. Salt Lake City:IEEE. 2018: 8759-8768.
[33] Tan M, Pang R, Le QV.EfficientDet: scalable and efficient object detection[C]//Proc IEEE/CVF CVPR. Seattle:IEEE. 2020: 10781-10790.
[34] Redmon J, Farhadi A.YOLOv3: an incremental improvement[J]. arXiv, 2018:1804.02767. DOI: 10.48550/arXiv.1804.02767.
[35] Wang CY, Bochkovskiy A, Liao HYM.YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors[C]//Proc IEEE/CVF CVPR. Vancouver:IEEE. 2023: 7464-7475.