Towards local progression estimation of pulmonary emphysema using CT

Area of Science:

• Radiology
• Pulmonary Medicine
• Medical Imaging Analysis

Background:

• Whole lung densitometry using chest CT scans is standard for measuring emphysema in clinical trials.
• Accurate assessment of emphysema progression requires localized density change measurements.
• Understanding regional changes is vital for interpreting drug efficacy and guiding treatment.

Purpose of the Study:

• To develop and validate methods for precise local measurement of lung density changes.
• To model the impact of respiration on lung density measurements.
• To enable accurate tracking of emphysema progression at a localized level.

Main Methods:

• Compared four CT image registration-based methods for analyzing lung density changes.
• Evaluated methods including image subtraction, a dry sponge model with volume correction, and novel adaptations addressing its limitations.
• Validated methods using a lung phantom and CT scans from 21 pulmonary emphysema patients.

Main Results:

• Optimized image registration achieved sub-millimeter accuracy.
• A locally adapted slope model demonstrated the highest accuracy in measuring local progression on phantom data.
• Systematic errors in progression estimation were minimal, even with significant volume differences.
• Patient data confirmed a strong linear relationship between lung volume and density changes.

Conclusions:

• The developed methods offer accurate local registration and effectively remove respiration-induced variability.
• These techniques accurately predict emphysema progression in phantom studies.
• The approach shows consistency with global results in patient data, indicating its potential utility.
• This method is a promising tool for evaluating local emphysema progression in clinical trials and practice.