Accuracy of upper airway volume measurements using different software products: a comparative analysis

Area of Science:

• Radiology and Imaging Science
• Biomedical Engineering
• Dental and Maxillofacial Imaging

Background:

• Accurate airway volume measurement from cone-beam computed tomography (CBCT) is crucial for clinical diagnosis and treatment planning.
• Various software programs exist for analyzing CBCT data, but their volumetric accuracy can differ significantly.
• NemoStudio is a commonly used software, but its performance in airway volume assessment requires detailed evaluation.

Purpose of the Study:

• To evaluate the accuracy of airway volume measurements derived from CBCT images using different software.
• To compare the performance of NemoStudio against other segmentation tools, including NNT Viewer, ITK-SNAP, and 3D Slicer.
• To establish a benchmark for volumetric accuracy using a water-filled solid model as the gold standard.

Main Methods:

• A 3D-printed airway model was created from CBCT data and scanned 10 times.
• Semi-automatic segmentation was performed using NemoStudio, NNT Viewer, ITK-SNAP, and 3D Slicer.
• Volume measurements were compared to a gold standard (water-filled cavity) using repeated measures ANOVA and Bland-Altman plots.

Main Results:

• NemoStudio exhibited systematic bias and higher variability, leading to lower accuracy compared to other software.
• ITK-SNAP and 3D Slicer demonstrated the highest agreement with the gold standard.
• NNT Viewer showed acceptable performance, while statistical analysis confirmed significant differences in accuracy among the tested software (P < 0.001).

Conclusions:

• Software selection significantly impacts the accuracy of CBCT-based airway volume measurements.
• ITK-SNAP and 3D Slicer are recommended for more reliable volumetric analysis of airways.
• Standardization of methods and refinement of segmentation algorithms are necessary for improved clinical consistency.