Automated Detection of MRONJ Lesions in Panoramic Dental X-rays Using Candidate Region Identification and Semantic Segmentation
Abstract
Medication-related osteonecrosis of the jaw (MRONJ) is a severe adverse effect associated with the administration of bone-modifying agents, such as bisphosphonates (BP) and denosumab (Dmab), and angiogenesis inhibitors. Despite the advancements in therapeutic agents, the incidence of MRONJ has increased, as medication remains a primary risk factor. In most cases, MRONJ is diagnosed at an advanced stage, where portions of the jawbone become exposed in the oral cavity, interfering with both primary disease management and MRONJ treatment. Therefore, early detection and treatment prior to progression are critical for improving patient outcomes and reducing treatment complexity. In Japan, the low penetration of dental CT limits the feasibility of 3D diagnostic imaging in routine practice in dental clinics. Therefore, this study proposes a diagnostic method that relies solely on panoramic X-ray images to automatically predict MRONJ lesions. The proposed method first performs pre-processing to extract the mouth region, and then compares two approaches for MRONJ lesion segmentation. The first approach subdivides the mouth region into patches and utilizes patch-based classification to identify candidate regions before MRNOJ lesion segmentation. The second approach employs the masked vision transformer (Masked-ViT) to estimate the probability of MRONJ lesion presence across the image, and then segmentation is applied to high probability areas. On our panoramic X-ray image dataset consisting of 118 MRONJ patients, the patch-based method achieved a maximum Dice Similarity Coefficient (DSC) of 0.70, outperforming the method using Masked-ViT. Although promising, further enhancements are necessary to meet the requirements for clinical use.
References
O. Nicolatou-Galitis, M. Schiødt, R. A. Mendes, C. Ripamonti, S. Hope, L. Drudge-Coates, D. Niepel and T. Van den Wyngaert, “Medication-related osteonecrosis of the jaw: definition and best practice for prevention, diagnosis, and treatment,” Oral surgery, Oral medicine, Oral pathology and Oral radiology, vol. 127, no. 2, pp. 117–135, 2019.
M. Nashi, H. Kishimoto, M. Kobayashi, A. Tachibana, S. Hashitani, T. Shibatsuji, T. Nishida, K. Fujimura, S. Furudoi, Y. Ishida, S. Ishii, T. Fujita, S. Iwai, T. Shigeta, T. Harada, D. Miyai, D. Takeda, M. Akashi, K. Noguchi and T. Takenobu, “Incidence of antiresorptive agent-related osteonecrosis of
the jaw: A multicenter retrospective epidemiological study in Hyogo Prefecture, Japan,” Journal of Dental Sciences, vol. 18.3, pp. 1156–1163, 2023.
A. Bedogni, S. Blandamura, Z. Lokmic, C. Palumbo, M. Ragazzo, F. Ferrari, A. Tregnaghi, F. Pietrogrande, O. Procopio, G. Saia, M. Ferretti, G. Bedogni, L. Chiarini, G. Ferronato, V. Ninfo, L. L. Russo, L. L. Muizo and P. F. Nocini, “Bisphosphonate-associated jawbone osteonecrosis: a correlation between imaging techniques and histopathology,” Oral surgery, Oral medicine, Oral pathology and Oral radiology, and Endodontology, vol. 105, no. 3, pp. 358–364, 2008.
N. Treister, N. Shehhy, E.H. Bae, B. Friedland, M. Lerman and S. Woo, “Dental panoramic radiographic evaluation in bisphosphonate-associated osteonecrosis of the jaws,” Oral Disease, vol. 15, no. 1, pp. 88–92, 2009.
P. Wongratwanich, K. Shimabukuro, M. Konishi, T. Nagasaki, M. Ohtsuka, Y. Suei, T. Nakamoto, R.G. Verdonschot, T. Kanesaki, P. Sutthiprapaporn and N. Kakimoto, “Do various imaging modalities provide potential early detection and diagnosis of medication-related osteonecrosis of the jaw? A review,” Dentomaxillofacial Radiology, vol. 50, no. 6, 2021.
Treister N. S., Friedland B. and Woo S. B., “Use of cone-beam computerized tomography for evaluation of bisphosphonateassociated osteonecrosis of the jaws,” Oral surgery, Oral medicine, Oral pathology and Oral radiology, vol. 109, no. 5, pp. 753–764, 2010.
J. Redmon, “You Only Look Once: Unified, Real-Time Object Detection,” Proceedings of the IEEE conference on computer vision and pattern recognition, 2016.
K. Simonyan and A. Zisserman, “Very Deep Convolutional Networks for Large-scale Image Recognition,” arXiv preprint arXiv, 1409.1556, 2014.
C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens and Z. Wojna, “Rethinking the Inception Architecture for Computer Vision,” Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 2818–2826, 2016.
M. Tan and Q. Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” International conference on machine learning, pp. 6105–6114, 2019.
S. Mehta and M. Rastegari, “Mobilevit: light-weight, generalpurpose, and mobile-friendly vision transformer,” arXiv preprint arXiv, 2110.02178, 2021.
A. Dosovitskiy, L. Beyer, A. Kolesnilov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit and N. Houlsby, “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint arXiv, 2010.11929, 2020.
K. He, X. Chen, S. Xie, Y. Li, P. Dollar and R. Girshick, “Masked autoencoders are scalable vision learners,” Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp. 16000–16009, 2022.
H. Bao, L. Dong, S. Piao, and F. Wei, “Beit: Bert pre-training of image transformers,” arXiv preprint arXiv, 2106.08254, 2021.
L. Zhou, H. Liu, J. Bae, J. He, D. Samaras and P. Prasanna, “Self pre-training with masked autoencoders for medical image classification and segmentation,” International Symposium on Biomedical Imaging (ISBI).IEEE, pp. 1–6, 2023.
O. Ronneberger, P. Fischer and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” Medical image computing and computer-assisted intervention, pp. 234–241, 2015.
H. Wang, Z. Wang, M. Du, F. Yang, Z. Zhang, S. Ding, P. Mardziel and X. Hu, “Score-CAM: Score-Weighted Visual Explanations for Convolutional Neural Networks,” Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops, pp. 24–25, 2020.
