Predicting Olfactory Bulb Infiltration and Morphological Changes in Olfactory Neuroblastoma Using MRI

Background and Purpose: Olfactory neuroblastoma (ONB) is a rare malignant tumor originating in the olfactory epithelium. The extent of olfactory bulb infiltration significantly impacts ONB staging, making imaging-based prediction crucial for treatment planning. This study aims to predict olfactory bulb infiltration by evaluating volumetric and morphological changes using magnetic resonance imaging (MRI).

Methods: We collected preoperative MRI scans and postoperative pathology reports from 46 ONB patients. High-resolution coronal T1-weighted and T2-weighted images were analyzed to assess olfactory bulb morphology and volume. Quantitative measurements included olfactory bulb volume and maximum cross-sectional area ([Fig. 1]). Infiltration was classified based on pathological findings. Sensitivity models were developed for volume, maximum cross-sectional area, and tumor/non-tumor side ratios for both volume and area.

Fig. 1 Measurement of olfactory bulb (OB) cross-sectional area and volume in MRI.In MRI, OB cross-sectional areas were measured by manual segmentation of coronal slices passing through the OB. OB volume was calculated by manually outlining the OB on each coronal slice (cross-sectional area in mm²), summing all the outlined surfaces, and multiplying by the slice thickness, resulting in volume in cubic millimeters.

Results: Significant volumetric and cross-sectional area changes were observed in infiltrated olfactory bulbs compared to non-infiltrated ones ([Fig. 2]). The diagnostic models showed strong performance, with area under the ROC curve values ranging from 0.878 to 0.971 ([Fig. 3]). Optimal cut-off values were identified: olfactory bulb volume ≥60 mm³, maximum cross-sectional area ≥9.5 mm², volume ratio (tumor/non-tumor) ≥175%, and cross-sectional area ratio (tumor/non-tumor) ≥175%, all demonstrating high sensitivity and specificity.

Fig. 2 Comparison of infiltrated and noninfiltrated olfactory bulbs based on volume and cross-sectional area using MRI.

Fig. 3 ROC curves for diagnostic models using volume and area metrics to predict olfactory bulb infiltration in olfactory neuroblastoma

Conclusion: MRI effectively assesses olfactory bulb morphological changes and infiltration in ONB, aiding in treatment planning. Further studies with larger cohorts are required to validate these findings and refine MRI-based evaluation criteria.