Objectives/Goals: Conventional magnetic resonance imaging (MRI) methods are unable to visualize bone comparable to computed tomography (CT). Radiation therapy can result in post-treatment complications, such as osteoradionecrosis (ORN). We aim to integrate black bone (BB) MRI into routine follow-up MRI scans as an alternative to CT to monitor the development of ORN. Methods/Study Population: We assess the CT, T2-weighted (T2W), and BB MRI scans of 21 head and neck cancer patients without active ORN. All images are acquired within 4 weeks of each other to minimize anatomical variability. The mandibles are manually segmented on each image, and the MR images are registered to the CT to map the CT contour onto the MR images for comparison. Image segmentation evaluation metrics, including dice similarity coefficient (DSC), mean distance to agreement (MDA), and Hausdorff distance (HD), assess for spatial differences between each MRI sequence, using CT as reference standard. DSC measures contour volume, while MDA and HD assess the differences in contour boundaries, with HD sensitive to outliers. We employ a Wilcoxon signed-rank test using a significance value of p≤0.05 to assess for statistical significance. Results/Anticipated Results: Using metrics widely accepted in biomedical research, we assess the performance of BB to visualize mandible against its conventional T2W MRI counterpart when compared to CT. Black bone has a significantly different DSC (median=0.878, IQR=0.0397) than T2W (median=0.806, IQR=0.0876; p<0.0001), indicating high volumetric overlap between contours on BB and CT. Black bone has a significantly different MDA (median=0.0731mm, IQR=0.0227mm) than T2W (median=0.119mm, IQR=0.0697; p<0.0001), indicating high agreement between contour boundaries on BB and CT. Black bone has a significantly different HD (median=0.649mm, IQR=0.211mm) than T2W (median=1.80, IQR=0.933; p<0.0001), indicating a submillimeter difference between contour boundaries, including outliers, on BB and CT. Discussion/Significance of Impact: Our results exhibit BB as a promising non-ionizing alternative to CT for bony visualization compared to conventional MRI. BB has potential to detect early anatomical changes in bone, providing a standardized assessment of ORN. This can improve our dose–response understanding of post-treatment toxicities, refining radiation treatment planning.
West et al. (Wed,) studied this question.