Validity of Diaphragm Volume Measurements Using Three-Dimensional Computed Tomography.

OBJECTIVES

The aim of this study was to measure the diaphragm volume using three-dimensional computed tomography (3D-CT) and verify its validity.

DESIGN

This was a retrospective study of existing samples.

METHODS

Participants comprised five male patients, aged 65-70 years, who underwent preoperative chest CT (with a slice thickness of 0.5 mm) before coronary artery bypass surgery. The diaphragm was selectively extracted using a workstation to reconstruct a stereoscopic image, and the total muscle volume was measured. To confirm the accuracy and reproducibility of diaphragm muscle volume measurements on CT, all cases were measured three times by two observers, and intraclass correlation coefficients (ICCs) and interobserver correlations were determined.

RESULTS

Observers #1 and #2 reported an average diaphragm volume of 256.7±33 cm and 259.3±36 cm, respectively. The ICC analyses yielded Cronbach's alphas of 0.992 and 0.981 from both observers, and the interobserver correlation was 0.991. The ICC of a single measurement and the average measurement was 0.984 (95% confidence interval: 0.998-0.884) and 0.992 (95% confidence interval: 0.999-0.939), respectively.

CONCLUSIONS

To our knowledge, this study is the first to standardize the method for measuring the total diaphragm volume and examine the reproducibility and validity of the new method. The diaphragm could be selectively extracted and reconstructed. Measurement of the total diaphragm muscle volume using a workstation to reconstruct a stereoscopic image is feasible and highly reproducible. This technique can be reliably employed to evaluate diaphragm volume, thickness, and morphology.