Comparing Mean-anchored Generalized Linear Binning with Established Methods to Quantify Xenon-129 Ventilation Defect...

Area of Science:

• Medical Imaging
• Pulmonary Medicine
• Quantitative MRI

Background:

• Hyperpolarized xenon-129 (¹²⁹Xe) MRI is a sensitive tool for assessing lung ventilation.
• Standardized quantification of ventilation defects remains a challenge in ¹²⁹Xe MRI.

Purpose of the Study:

• To introduce and evaluate Mean-anchored generalized linear binning (Mean_GLB) for calculating ventilation defect percentage (VDP).
• To benchmark Mean_GLB against reader-segmented VDP and other automated methods.

Main Methods:

• ¹²⁹Xe MRI data from 38 cystic fibrosis patients and 25 healthy controls were analyzed.
• Ventilation defect percentage (VDP) was quantified using Mean_GLB, reader-segmented analysis, and five other automated methods.
• Bias correction was performed using N4 or flip-angle (FA) methods; voxel-wise agreement was assessed using sensitivity, specificity, and Youden's index.

Main Results:

• Mean_GLB demonstrated significant differences in VDP between cystic fibrosis patients and controls, regardless of bias correction method.
• Mean_GLB showed strong correlation with reader-segmented VDP (r=0.97) and achieved the highest Youden's index (∼0.75).
• While Mean_GLB slightly overestimated defects in cystic fibrosis patients, it proved robust across different bias correction techniques.

Conclusions:

• Mean_GLB provides a robust and reliable method for VDP quantification in ¹²⁹Xe ventilation MRI.
• The method is independent of the bias-correction algorithm used.
• Mean_GLB's sensitivity to subtle impairments suggests its utility in pulmonary imaging applications.