Abstract
INTRODUCTION: CT scan acquisition protocol is pivotal for segmentation covering the utmost periphery. We hypothesize that application of continuous positive airway pressure (CPAP) during CT acquisition could improve segmentation.
OBJECTIVES: Compare quality of segmentations under 4 CT acquisitions: inspiration (INSP), expiration (EXP) and both with CPAP (INS-CPAP and EXP-CPAP).
MATERIALS AND METHODS: 320-detector row CT scans (Aquilion ONE™, Toshiba) were performed in INSP, EXP, INS-CPAP and EXP-CPAP using EzPAP® (Smiths Medical) pressures 6-10 cmH2O for 3' in 4 patients. Segmentations obtained and compared with VBN (LungPoint®, Broncus) and their quality assessed by randomly selecting 2 distal bronchi per lobe and counting bifurcations (BIF) achieved and absolute (DIST) and relative (DIST%) distances from carina to most distal bronchi. DIST% were computed dividing DIST by lung's largest axis.
RESULTS: 2-way ANOVA interaction not significant: p-values and means for column factors. and row
| INSP | INSP-CPAP | EXP | EXP-CPAP | ANOVA (p-value) | |
| BIF (mean) | 6.87 | 6.63 | 4.98 | 5.42 | 4x10e-10 |
| DIST (mean) | 158.30 | 158.33 | 116.88 | 129.97 | 3x10e-12 |
| DIST% | 68.61 | 67.78 | 62.18 | 70.52 | 0.0545 |
Comparison of segmentation quality for CT acquisition
| Right lung | Left lung | ANOVA (p-value) | |
| BIF (mean) | 6.0 | 5.9 | 0.7099 |
| DIST (mean) | 133.93 | 147.81 | 0.0017 |
| DIST% | 65.62 | 68.93 | 0.1384 |
Comparison of segmentation quality for lung
CONCLUSIONS: 1) BIF and DIST might not be accurate enough for comparing segmentations in different phases of the respiratory cycle nor different lungs, due to their varying size. 2) Though not significant, DIST% are larger for EXP-CPAP indicating that use of CPAP might improve segmentation in VBN and encouraging further analysis of CPAP-enhanced CT.
- Copyright ©the authors 2016
