Related Experiment Video
Updated: Apr 15, 2026
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
Published on: December 15, 2023
Intraparotid Facial Nerve Segmentation: A Reproducibility Study of Manual Segmentation on Neurographic Sequences
Santiago Medrano-Martorell1, Jose C Pariente2, Yensa Rodríguez Álvarez2
1From the Department of Neuroradiology, Hospital Clinic, Barcelona, Spain; Fundació de Recerca Clínic Barcelona-Institut d'Investigacions Biomèdiques August Pi i Sunyer, Barcelona, Spain; Department of Maxillofacial Surgery, Hospital Clinic, Barcelona, Spain; and Department of Otorhinolaryngology, Hospital Clinic, Barcelona, Spain. medrano@clinic.cat.
Manual segmentation of the intraparotid facial nerve is reproducible for the trunk but challenging for peripheral branches. Advanced MR neurography sequences like VFA-TSE may improve delineation for presurgical planning of parotid tumors.
Area of Science:
- Radiology
- Neuroimaging
- Surgical Planning
Background:
- Preoperative facial nerve delineation via MR neurography aids parotid tumor surgery.
- Reliability of manual segmentation for intraparotid facial nerve pathways needs investigation.
Purpose of the Study:
- Assess manual segmentation reproducibility of the intraparotid facial nerve on neurographic MRI.
- Compare segmentation accuracy between DESS and VFA-TSE sequences.
Main Methods:
- Analyzed MR neurography datasets from 7 patients with parotid tumors.
- Two radiologists and two trainees segmented the facial nerve using 3D Slicer.
- Quantified reliability using mean pairwise error and interobserver agreement.
Main Results:
- Segmentation was reproducible for the nerve trunk but challenging for peripheral branches.
- VFA-TSE showed lower errors than DESS for the trunk (1.26 mm vs 1.85 mm).
- Nerve segments near tumors had higher errors and variability; radiologist experience had no significant impact.
Conclusions:
- Manual segmentation is reliable for the facial nerve trunk, less so for branches.
- VFA-TSE may offer advantages for trunk delineation; standardized protocols are beneficial.
- Further research should focus on refining sequences and automated segmentation for improved consistency.
