Virtual Laryngoscopy: Feasibility Studies by CT and MRI

Both computed tomography (CT) and magnetic resonance (MR) imaging protocols and a self-adaptive segmentation algorithm were employed to study the feasibility of virtual laryngoscopy. Several visualization techniques were utilized to achieve a real-time interactive inspection of the larynx. The developed virtual laryngoscopy was tested on 2 volunteers and 2 patients. Spiral CT protocol offers the advantage of high speed (less than 30 seconds of data acquisition) and high resolution (cubic voxel size of 0.3 mm), at the risk of radiation. Fast spoiled-GRASS MR pulse sequence provides a better tissue contrast and a reasonable signal-to-noise ratio, at the cost of lower resolution (of cubic voxel size of 1 mm) and longer acquisition time (close to 7 minutes). The segmentation algorithm performed consistently for the four studies and was not sensitive to motion artifacts in both CT and MR images. The visualization system could interactively navigate inside the larynx lumen with features of inspecting and zooming the geometry, exam the outside of the larynx with associated tissues labeled by different colors, and cut the larynx in any orientation to open the whole volume for a view of the inner surface. The examination of larynx was non-invasive, reproducible, and efficient. It demonstrated the feasibility of virtual laryngoscopy for diagnosis of laryngeal diseases.