Variability of bronchial measurements obtained by sequential CT using two computer-based methods

Area of Science:

• Medical Imaging
• Pulmonary Medicine
• Quantitative CT Analysis

Background:

• Accurate measurement of bronchial lumen (LA) and wall area (WA) is crucial for assessing airway diseases.
• Multidetector CT (MDCT) is commonly used, but measurement variability can impact diagnostic accuracy.
• Comparing different image analysis techniques is essential for reliable longitudinal studies.

Purpose of the Study:

• To evaluate the variability of LA and WA measurements between two successive MDCT acquisitions.
• To compare the performance of energy-driven contour estimation (EDCE) and full width at half maximum (FWHM) approaches.
• To assess the suitability of these methods for studying bronchial remodeling.

Main Methods:

• MDCT data from 42 bronchial segments across 10 slices were analyzed.
• LA and WA were measured using both EDCE and FWHM techniques.
• Variability was assessed using 95% confidence intervals and the coefficient of measurement variation (CV(10)).

Main Results:

• Both EDCE and FWHM showed good reproducibility between repeated MDCT scans (95% CI for LA: -1.59 to 1.5 mm², for WA: -3.31 to 2.96 mm²).
• The coefficient of measurement variation (CV(10)) was strongly correlated between acquisitions (r > 0.72).
• EDCE yielded higher LA values for LA < 15 mm² and lower WA values for WA < 13 mm² compared to FWHM.

Conclusions:

• Measurement variability exists between different CT assessment techniques (EDCE vs. FWHM).
• EDCE and FWHM are not interchangeable for longitudinal studies of bronchial remodeling.
• New parameters like CV(10) may be valuable for studying bronchial remodeling.