Textural analysis and lung function study: Predicting lung fitness for radiotherapy from a CT scan

Area of Science:

• Radiotherapy
• Medical Imaging
• Pulmonary Function

Background:

• Patient fitness assessment is crucial for radical radiotherapy.
• Current methods rely on formal lung function tests like FEV1 and TLCO.
• Advanced image analysis may offer an alternative using standard CT scans.

Purpose of the Study:

• To test if advanced image analysis of CT scans can differentiate radiotherapy-eligible patients.
• To compare image analysis findings with formal lung function tests (FEV1 and TLCO).

Main Methods:

• An apical region of interest (ROI) was extracted from radiotherapy planning CT scans.
• Grey Level Co-occurrence Matrix (GLCM) software calculated entropy scores for each voxel.
• Density and entropy scores were compared between fit and unfit patient groups.

Main Results:

• Significant differences in mean and median density were found between fit and unfit patients (p=0.005, p=0.0008).
• Median entropy scores also significantly differed between the groups (p=0.0418).
• These findings indicate image analysis can distinguish patient fitness.

Conclusions:

• Density and entropy assessments from standard CT imaging can differentiate patient fitness for radical radiotherapy.
• This novel assessment generates additional data from existing CT scans.
• Future work could integrate this into a multiorgan fitness assessment from a single scan.