3DSlicer.CTLungAnalyzer

3DSlicer.CTLungAnalyzer segments and quantifies infiltrated, emphysematous, and collapsed lung regions from chest CT scans to provide spatial visualization and volumetric metrics for assessment of pulmonary involvement such as COVID-19.

Key Features:

• Lung CT Segmenter: Provides a semi-automatic workflow for generating lung masks.
• Hounsfield unit (HU) thresholding: Uses masked thresholds of Hounsfield units (HU) to distinguish non-affected and affected areas, including emphysematous, infiltrated, and collapsed regions.
• Intrapulmonary vessel subtraction: Subtracts intrapulmonary vessels from segmented lung volumes.
• Volume quantification: Quantifies segmented volumes in milliliters (mL).
• 3D reconstruction: Displays segmented volumes in three-dimensional renderings.
• COVID-Q metric: Calculates COVID-Q as the ratio of affected to non-affected lung volume for each lung separately.
• Processing performance: Reported average processing time of approximately eight minutes per patient in a retrospective study.

Scientific Applications:

• Quantification of COVID-19 lung involvement: Provides volumetric metrics and COVID-Q values that correlated with degree of lung involvement in a retrospective study of COVID-19 patients.
• Prognostic assessment: Identifies patients with high lung involvement (reported fatal cases exhibited COVID-Q > 2.0) for potential risk stratification.
• Serial monitoring: Enables serial 3D reconstruction and quantification of infiltrated and collapsed lung areas to monitor disease progression over time.
• Volumetric research and clinical analysis: Supplies per-lung volumetric measurements for research and clinical evaluation of pulmonary pathology.

Methodology:

Implemented as a 3DSlicer extension with two components: a semi-automatic Lung CT Segmenter for lung mask generation, and a Volume Analysis that applies masked Hounsfield unit thresholds to classify emphysematous, infiltrated, and collapsed regions, subtracts intrapulmonary vessels, quantifies volumes in milliliters, displays results in 3D, and computes COVID-Q as the ratio of affected to non-affected lung volume per lung.