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

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.

Fig. 1.

Patient exclusion and dataset split workflow.

Fig. 2.

Axial view of a contrast-enhanced CT image with segmented ORN (red) and contralateral mirror-image control normal bone (yellow).

Fig. 3.

Workflow for the radiomic feature extraction and modeling steps. The number of features at each step is denoted by N.

Fig. 4.

A) The cumulative sum of feature importance as determined by the Random Forest model. The horizontal line represents the threshold where cumulative importance reaches 100%, and the vertical line indicates the number of features required to achieve this cumulative importance. B) The top 20 features ranked by their importance in the classification process of the RF model. The length of each bar reflects the relative importance of each feature in the model’s decision-making process, highlighting which features are most influential in differentiating between ORNJ and healthy mandibular tissues.

Fig. 5.

Model performance evaluation. A) Receiver operating characteristic (ROC) curves (mean and per CV fold) of the Random Forest trained on the selected extracted radiomic features (n = 67) from ORN and healthy mandible VOIs on CECT. B) Reliability curve for the Random Forest classifier. Actual outcome probabilities are plotted against predicted probabilities. The thick grey diagonal line represents an ideal calibration, where predicted probabilities align perfectly with the observed outcome frequencies. Deviations from this line indicate overconfidence (points below the diagonal) or underconfidence (points above the diagonal) in the model’s predictions. The tick marks along the x-axis show the distribution of predicted probabilities. A high concentration of ticks in a certain region indicates that many predictions fall within that probability range.

Fig. 6.

SHAP values for the top 20 most influential radiomic features. The y-axis lists each feature, ordered by the average magnitude of their SHAP values. Each dot represents the SHAP value for each individual subject in the dataset illustrating the extent of each feature’s impact on the model’s prediction for differentiating between ORNJ and healthy mandibular tissue. The color gradient, ranging from blue to red, indicates the range of feature values, with red signifying higher values.