SU-E-J-120: Characterization of 4D Lung Ventilation Based on 4DCT Lung Density

Area of Science:

• Medical Imaging
• Pulmonary Biomechanics
• Computational Anatomy

Background:

• Accurate characterization of lung ventilation and motion is crucial for developing predictive biomechanical models.
• Four-dimensional computed tomography (4DCT) provides dynamic lung information but is susceptible to motion artifacts.
• Developing robust methods to analyze 4DCT data is essential for improving lung cancer treatment planning.

Purpose of the Study:

• To characterize patient's 4D lung ventilation and motion using 4DCT images.
• To extract pulmonary features for developing a biomechanical model.
• To predict lung tumor motion with enhanced accuracy.

Main Methods:

• Utilized 4DCT images from sixteen patients for ventilation evaluation.
• Developed a program for 4D lung ventilation visualization and quantification.
• Applied free-form deformable image registration (DIR) and developed screening algorithms (Fourier Transformation, Generalized Linear Regression) to assess RPM waveform regularity and minimize motion artifacts.

Main Results:

• Observed significant lung density gradients and estimated gravity pressure effects on pleural pressure.
• Quantified differences in diaphragm motion range between anterior (8±5mm) and posterior (23±12mm) chest walls.
• Demonstrated that audio-coaching improves RPM waveform regularity (0.76 vs. 0.50 Fourier; 0.74 vs. 0.66 Regression), indicating reduced motion artifacts.

Conclusions:

• Characterized fundamental features of lung ventilation and motion from 4DCT data.
• Developed two effective screening methods for selecting high-quality 4DCT datasets.
• These methods facilitate ongoing quantitative studies for improved biomechanical modeling and lung tumor motion prediction.