Texture analysis of technegas lung ventilation images

Area of Science:

• Medical Imaging Analysis
• Nuclear Medicine

Background:

• Technegas lung ventilation images can exhibit 'hot spots', especially in patients with respiratory diseases.
• Quantifying image 'spottiness' is crucial for accurate diagnosis and treatment planning in respiratory conditions.

Purpose of the Study:

• To introduce and evaluate a novel technique for quantifying 'spottiness' in Technegas lung ventilation images.
• To assess the effectiveness of morphological texture analysis in measuring image spottiness compared to expert opinions.

Main Methods:

• Employed morphological texture analysis on 32 Technegas lung ventilation images from patients with respiratory diseases.
• Images were filtered using morphological opening across various scales to derive texture parameters.
• Compared quantitative texture parameters with classifications and rankings from three experienced nuclear medicine physicians.

Main Results:

• A combination of two texture parameters achieved high correlation (rs = 0.71, p < 0.01) with physician rankings for spotty images.
• The technique demonstrated high classification accuracy: 83% for spotty and 90% for non-spotty images.
• The proposed texture analysis method showed superiority over previously published techniques.

Conclusions:

• Morphological texture analysis provides a valuable and objective measure for quantifying image spottiness in Technegas lung ventilation scans.
• This technique offers practical applications and improved diagnostic capabilities in nuclear medicine for respiratory disease assessment.