CT-Based Evaluation of the Shape of the Diaphragm Using 3D Slicer

• Published in: Journal of digital imaging Vol. 37; no. 4; pp. 1980 - 1990
• Main Authors: Taton, Olivier, Van Muylem, Alain, Leduc, Dimitri, Gevenois, Pierre Alain
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.08.2024; Springer Nature B.V
• Subjects: Aged; Chronic obstructive pulmonary disease; Computed tomography; Diaphragm; Diaphragm (anatomy); Diaphragm - anatomy & histology; Diaphragm - diagnostic imaging; Female; Humans; Imaging; Imaging, Three-Dimensional - methods; Male; Medical imaging; Medicine; Medicine & Public Health; Middle Aged; Muscles; Pulmonary Disease, Chronic Obstructive - diagnostic imaging; Radiology; Radius of curvature; Reproducibility; Reproducibility of Results; Software; Statistical analysis; Sternum; Thorax; Tomography, X-Ray Computed - methods; Vertebrae; Computed tomography; Chronic obstructive pulmonary disease; Diaphragm; 3D Slicer
• ISSN: 2948-2933; 0897-1889; 2948-2925; 2948-2933; 1618-727X
• DOI: 10.1007/s10278-024-01069-y

The diaphragm is the main inspiratory muscle and separates the thorax and the abdomen. In COPD, the evaluation of the diaphragm shape is clinically important, especially in the case of hyperinflation. However, delineating the diaphragm remains a challenge as it cannot be seen entirely on CT scans. Therefore, the lungs, ribs, sternum, and lumbar vertebrae are used as surrogate landmarks to delineate the diaphragm. We herein describe a CT-based method for evaluating the shape of the diaphragm using 3D Slicer —a free software that allows delineation of the diaphragm landmarks—in ten COPD patients. Using the segmentation performed with 3D Slicer , the diaphragm shape was reconstructed with open-source Free Pascal Compiler. From this graduated model, the length of the muscle fibers, the radius of curvature, and the area of the diaphragm—the main determinants of its function—can be measured. Inter- and intra-user variabilities were evaluated with Bland and Altman plots and linear mixed models. Except for the coronal length ( p  = 0.049), there were not statistically significant inter- or intra-user differences ( p values ranging from 0.326 to 0.910) suggesting that this method is reproducible and repeatable. In conclusion, 3D Slicer can be applied to CT scans for determining the shape of the diaphragm in COPD patients.