Lung Segmentation¶

To allow for analysis of the lung parenchyma, CIRRUS Lung provides a fully automatic segmentation of the lungs, trachea and main bronchi. To be able to segment the lungs in scans containing a variety of abnormalities, a hybrid lung segmentation approach is applied for each scan. The hybrid lung segmentation approach first segments the lungs using a fast algorithm based on region growing and morphological processing. Next, possible segmentation failures (due to e.g. pathological abnormalities) are automatically detected based on statistical deviations from a range of volume and shape measurements. To scans in which failures are detected, a multi-atlas basd algorithm using non-rigid registration is applied. Details of the algorithm are described in the technical papers by Sluimer et al. (2005) and van Rikxoort et al. (2009) listed below. The automatic segmentation of the lungs is run offline. When the computation is finished, the resulting lung segmentation can be inspected with CIRRUS Lung in an overlay depicting each lung and the trachea and main bronchi. 

Technical publications about lung segmentation algorithms included in CIRRUS Lung¶

E.M. van Rikxoort, B. de Hoop, M.A. Viergever, M. Prokop and B. van Ginneken. "Automatic lung segmentation from thoracic computed tomography scans using a hybrid approach with error detection", *Medical Physics *2009;36:2934-2947. Abstract/PDF DOI PMID

I. Sluimer, M. Prokop and B. van Ginneken. "Towards automated segmentation of the pathological lung in CT",IEEE Transactions on Medical Imaging 2005;24:1025-1038. Abstract/PDF DOI PMID