Computed tomography radiomics-based cross-sectional detection of mandibular osteoradionecrosis in head and neck cancer survivors

Serageldin Kamel; Laia Humbert-Vidan; Zaphanlene Kaffey; Sarah Mirbahaeddin; Abdulrahman Abusaif; David T A Fuentes; Kareem Wahid; Cem Dede; Mohamed A Naser; Renjie He; Ahmed W Moawad; Khaled M Elsayes; Melissa M Chen; Adegbenga O Otun; Jillian Rigert; Mark S Chambers; Andrew Hope; Erin Watson; Kristy K Brock; Katherine Hutcheson; Lisanne van Dijk; Amy C Moreno; Stephen Y Lai; Clifton D Fuller; Abdallah S R Mohamed; MD Anderson Head and Neck Cancer Symptom Working Group

doi:10.1016/j.oraloncology.2025.107337

Computed tomography radiomics-based cross-sectional detection of mandibular osteoradionecrosis in head and neck cancer survivors

Oral Oncol. 2025 Jun 13:167:107337. doi: 10.1016/j.oraloncology.2025.107337. Online ahead of print.

Authors

Serageldin Kamel¹, Laia Humbert-Vidan¹, Zaphanlene Kaffey¹, Sarah Mirbahaeddin¹, Abdulrahman Abusaif¹, David T A Fuentes², Kareem Wahid³, Cem Dede¹, Mohamed A Naser¹, Renjie He¹, Ahmed W Moawad⁴, Khaled M Elsayes⁴, Melissa M Chen⁴, Adegbenga O Otun⁵, Jillian Rigert¹, Mark S Chambers⁵, Andrew Hope⁶, Erin Watson⁷, Kristy K Brock², Katherine Hutcheson⁵, Lisanne van Dijk⁸, Amy C Moreno¹, Stephen Y Lai⁹, Clifton D Fuller¹⁰, Abdallah S R Mohamed¹¹; MD Anderson Head and Neck Cancer Symptom Working Group

Affiliations

¹ The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA.
² The University of Texas MD Anderson Cancer Center, Department of Imaging Physics, Houston, USA.
³ The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA; The University of Texas MD Anderson Cancer Center, Department of Imaging Physics, Houston, USA.
⁴ The University of Texas MD Anderson Cancer Center, Department of Radiology, Houston, USA.
⁵ The University of Texas MD Anderson Cancer Center, Department of Head and Neck Surgery, Houston, USA.
⁶ Department of Radiation Oncology, Princess Margaret Cancer Centre, Toronto, Canada.
⁷ Department of Radiation Oncology, Princess Margaret Cancer Centre, Toronto, Canada; Faculty of Dentistry, University of Toronto, Toronto, Canada.
⁸ Department of Radiation Oncology, University Medical Center Groningen, the Netherlands.
⁹ The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA; The University of Texas MD Anderson Cancer Center, Department of Head and Neck Surgery, Houston, USA. Electronic address: sylai@mdanderson.org.
¹⁰ The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA. Electronic address: cdfuller@mdanderson.org.
¹¹ The University of Texas MD Anderson Cancer Center, Department of Radiation Oncology, Houston, USA; Department of Radiation Oncology, Baylor College of Medicine, Houston, USA. Electronic address: ASMohamed@mdanderson.org.

PMID: 40516152
DOI: 10.1016/j.oraloncology.2025.107337

Abstract

Purpose: This study aims to identify radiomic features from contrast-enhanced CT (CECT) scans that differentiate osteoradionecrosis (ORN) from normal mandibular bone in head and neck cancer (HNC) patients treated with radiotherapy (RT).

Materials and methods: CECT images from 150 patients with confirmed ORN diagnosis (2008-2018) at MD Anderson Cancer Center (MDACC) were analyzed (80 % train, 20 % test). Radiomic features were extracted using PyRadiomics from manually segmented ORN regions and automated contralateral healthy mandible regions. Correlation analysis (r > 0.95) reduced features for model training. A random Forest (RF) classifier with Recursive Feature Elimination identified discriminative features. Explainability was assessed using SHapley Additive exPlanations (SHAP) analysis on the 20 most important features identified by the trained RF classifier.

Results: Of the 1316 radiomic features extracted, 810 features were excluded for high collinearity. From a set of 506 pre-selected radiomic features, 67 were optimal for RF classification, yielding 88% accuracy and a ROC AUC of 0.96. The model well calibrated (Log Loss 0.296, ECE 0.125) and achieved an accuracy of 88% and a ROC AUC of 0.96. The SHAP analysis revealed that higher values of Wavelet-LLH First order Mean and Median were associated with ORN of the jaw (ORNJ). Conversely, higher Exponential GLDM Dependence Entropy and lower Square First-order Kurtosis were more characteristic of normal mandibular tissue.

Conclusion: This study successfully developed a CECT-based radiomics model for differentiating ORNJ from healthy mandibular tissue in HNC patients after RT. Future work will focus on detecting subclinical ORNJ regions to guide earlier interventions.