Objectives: The objective of the study is to assess the correlation between the mandibular canal tracing done on cone beam CT (CBCT) data, with the size, shape, and position of the neurovascular bundle (NB) obtained by magnetic resonance imaging (MRI). Material and methods: Six human cadaver mandibles were scanned with a CBCT Promax® scanner (Planmeca, Helsinki, Finland) and with an Ingenia® 3.0 T MR system (Philips, Amsterdam, The Netherlands). The NB was segmented from the MRI dataset, while the mandibular canal (MC) tracing was done on the CBCT images. Quantitative 3D analysis was made for the full-segmented nerves and for three defined regions of specific clinical interest, namely angle, body, and mental region. Results: From the 3D MRI analysis, the nerve thickness (for the angle, body, and mental region) ranges from 0.8 to 5.2 mm, while the thickness of the mandibular canal tracing is approximately 2.00 mm on both sides as chosen in the tracing software. The mean volume of the NB on the left is 828.49 ± 215.54 mm3 and on the right 792.98 ± 264.57 mm3. For the nerve tracing, the mean value is 351.92 ± 16.42 and 339.69 ± 16.12 mm3 on the left and right sides, respectively. Wilcoxon signed-rank test showed significant differences between NB and MC volume measurements (p = 0.0005). The Bland-Altman plots show an increasing slope for thickness and volume, indicating that the absolute differences between neurovascular bundle, estimated by MRI, and the mandibular canal, drawn on the CBCT images, increase with larger mean values. Conclusions: Surgeons should be aware of the shortcomings of nerve tracing in the different regions of the mandible. Clinical relevance: Tracing of the inferior alveolar nerve (IAN) underestimates shape and volume. Whenever nerve tracing instead of well-recognizable anatomical bony landmarks is used for surgical planning that need precision, a wider safe margin is recommended.
