Automatic Lung Segmentation in CT Images with Accurate Handling of the Hilar Region
- 709 Downloads
A fully automated and three-dimensional (3D) segmentation method for the identification of the pulmonary parenchyma in thorax X-ray computed tomography (CT) datasets is proposed. It is meant to be used as pre-processing step in the computer-assisted detection (CAD) system for malignant lung nodule detection that is being developed by the Medical Applications in a Grid Infrastructure Connection (MAGIC-5) Project. In this new approach the segmentation of the external airways (trachea and bronchi), is obtained by 3D region growing with wavefront simulation and suitable stop conditions, thus allowing an accurate handling of the hilar region, notoriously difficult to be segmented. Particular attention was also devoted to checking and solving the problem of the apparent ‘fusion’ between the lungs, caused by partial-volume effects, while 3D morphology operations ensure the accurate inclusion of all the nodules (internal, pleural, and vascular) in the segmented volume. The new algorithm was initially developed and tested on a dataset of 130 CT scans from the Italung-CT trial, and was then applied to the ANODE09-competition images (55 scans) and to the LIDC database (84 scans), giving very satisfactory results. In particular, the lung contour was adequately located in 96% of the CT scans, with incorrect segmentation of the external airways in the remaining cases. Segmentation metrics were calculated that quantitatively express the consistency between automatic and manual segmentations: the mean overlap degree of the segmentation masks is 0.96 ± 0.02, and the mean and the maximum distance between the mask borders (averaged on the whole dataset) are 0.74 ± 0.05 and 4.5 ± 1.5, respectively, which confirms that the automatic segmentations quite correctly reproduce the borders traced by the radiologist. Moreover, no tissue containing internal and pleural nodules was removed in the segmentation process, so that this method proved to be fit for the use in the framework of a CAD system. Finally, in the comparison with a two-dimensional segmentation procedure, inter-slice smoothness was calculated, showing that the masks created by the 3D algorithm are significantly smoother than those calculated by the 2D-only procedure.
Key wordsCAD image segmentation lung nodules region growing grid 3D imaging biomedical image analysis
This work was supported by the Italian “Istituto Nazionale di Fisica Nucleare” (INFN) and by a grant (“Progetto di Ricerca di Interesse Nazionale”, PRIN, 2005) from the Italian “Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR).
We thank Anna Margiotta, Marinella Foti, Luigi Pignatelli of the “Unità Operativa di Radiologia”, Azienda Ospedaliera Vito Fazzi, Lecce. Thanks to A. De Liperi, Dr. Spinelli of the “Unità Operativa Radiodiagnostica 2”, Azienda Ospedaliera Universitaria Pisana, and M. Barattini, D. Caramella, T. Tarantino of the “Diagnostic and Interventional Radiology” department, University of Pisa. We also wish to thank Claudia Pecoraro for useful discussions.
- 1.The Lancet, vol. 365, May 14, 2005Google Scholar
- 3.Picozzi G, Paci E, Lopez Pegna A, Bartolucci M, Roselli G, De Francisci A, Gabrielli S, Masi A, Villari N, Mascalchi M: Screening of lung cancer with low dose spiral CT: results of a three year pilot study and design of the randomised controlled trial ‘Italung-CT’. Radiol Med (Torino) 109(1–2):17–26, 2005 See also: http://www.cspo.it/ Google Scholar
- 8.Park M, Wilson LS, Jin JS: Automatic extraction of lung boundaries by a knowledge-based method, visualization 2000, Pan-Sydney Workshop on Visual Information Processing, December, 2000. Conferences in Research and Practice in Information Technology, 2. P. Eades and J. Jin, Eds.Google Scholar
- 12.Bellotti R, De Carlo F, Gargano G, Tangaro S, Cascio D, Catanzariti E, Cerello P, Cheran SC, Delogu P, De Mitri I, Fulcheri C, Grosso D, Retico A, Squarcia S, Tommasi E, Golosio B: A CAD system for nodule detection in low-dose lung CTs based on region growing and a new active contour model. Med Phys 34(12):4901–4910, 2007CrossRefPubMedGoogle Scholar
- 15.Insight-users forum, http://www.itk.org/pipermail/insight-users/2003-September/004762.html
- 16.Bellotti R, Bagnasco S, Bottigli U, Castellano M, Cataldo R, Catanzariti E, Cerello P, Cheran SC, De Carlo F, Delogu P, De Mitri I, De Nunzio G, Fantacci ME, Fauci F, Forni G, Gargano G, Golosio B, Indovina PL, Lauria A, Lopez Torres E, Magro R, Martello D, Masala GL, Massafra R, Oliva P, Palmiero R, Preite Martinez A, Prevete R, Quarta M, Ramello L, Raso G, Retico A, Santoro M, Sitta M, Stumbo S, Tangaro S, Tata A, Zanon E: The MAGIC-5 Project: Medical Applications on a Grid Infrastructure Connection. Proc IEEE Nuclear Science Symposium 16–22/10/2004, Rome, ItalyGoogle Scholar
- 17.Bellotti R, Cerello P, Tangaro S, Bevilacqua V, Castellano M, Mastronardi G, De Carlo F, Bagnasco S, Bottigli U, Cataldo R, Catanzariti E, Cheran SC, Delogu P, De Mitri I, De Nunzio G, Fantacci ME, Fauci F, Gargano G, Golosio B, Indovina PL, Lauria A, Lopez Torres E, Magro R, Masala GL, Massafra R, Oliva P, Preite Martinez A, Quarta M, Raso G, Retico A, Sitta M, Stumbo S, Tata A, Squarcia S, Schenone A, Molinari E, Canesi B: Distributed Medical Images Analysis on a GRID Infrastructure. Future Generation Computer System, Special Issue on Life Science Grids for Biomedicine and Bioinformatics, 23, 2006Google Scholar
- 18.Bottigli U, Cerello P, Cheran S, Delogu P, Fantacci ME, Fauci F, Golosio B, Lauria A, Lopez Torres E, Magro R, Masala GL, Oliva P, Palmiero R, Raso G, Retico A, Stumbo S, Tangaro S: GPCALMA: a tool for mammography with a GRID connected distributed database. Proc. of the Seventh Mexican Symp. on Medical Physics, 682/1:67, 2003, also e-preprint physics/0410084 Google Scholar
- 19.Fauci F, Raso G, Magro R, Forni G, Lauria A, Bagnasco S, Cerello P, Cheran SC, Lopez Torres E, Bellotti R, De Carlo F, Gargano G, Tangaro S, De Mitri I, De Nunzio G, Cataldo R: A massive lesion detection algorithm in mammography. Physica Medica XXI(1):21–28, 2005Google Scholar
- 20.Cascio D, Fauci F, Magro R, Raso G, Bellotti R, De Carlo F, Tangaro S, De Nunzio G, Quarta M, Forni G, Lauria A, Fantacci ME, Retico A, Masala GL, Oliva P, Bagnasco S, Cheran SC, Lopez Torres E: Mammogram segmentation by contour searching and mass lesion classification with neural network. IEEE Trans Nucl Sci 53(5):2827–2833, 2006CrossRefGoogle Scholar
- 21.Bellotti R, De Carlo F, Tangaro S, Gargano G, Maggipinto G, Castellano M, Massafra R, Cascio D, Fauci F, Magro R, Raso G, Lauria A, Forni G, Bagnasco S, Cerello P, Zanon E, Cheran SC, Lopez Torres E, Bottigli U, Masala GL, Oliva P, Retico A, Fantacci ME, Cataldo R, De Mitri I, De Nunzio G: A completely automated CAD system for mass detection in a large mammographic database. Med Phys 33(8):3066–3075, 2006CrossRefPubMedGoogle Scholar
- 22.Cheran CS, Cataldo R, Cerello P, De Carlo F, Fauci F, Forni G, Golosio B, Lauria A, Lopez Torres E, De Mitri I, Masala G, Raso G, Retico A, Tata A: Detection and classification of microcalcification clusters in digital mammograms. Proc IEEE Trans Med Imaging Conference, 16–22/10/2004, Rome, ItalyGoogle Scholar
- 23.De Nunzio G, on behalf of MAGIC-5 collaboration, “Lung Parenchima Segmentation in CT Scans as a Preprocessing Step for Automatic Nodule Detection", TCN CAE 2005, International Conference on CAE and Computational Technologies for Industry, Lecce (Italy) October 8, 2005Google Scholar
- 32.Shojaii R, Alirezaie J, Babyn P: Automatic lung segmentation in CT Images using watershed transform. 2005 IEEE International Conference on Image Processing, Genoa (Italy), 2005Google Scholar
- 36.Ko JP, Betke M: Chest CT: automated nodule detection and assessment of change over time—preliminary experience. Radiology 2002:267–273, 2001Google Scholar
- 37.Shen H, Goebel B, Odry B: A new algorithm for local surface smoothing with application to chest wall nodule segmentation in lung CT Data, Medical Imaging 2004: Image Processing, edited by J. Michael Fitzpatrick, Milan Sonka, Proceedings of SPIE 5370 (SPIE, Bellingham, WA, 2004)Google Scholar
- 39.Gargano G, Bellotti R, De Carlo F, Tangaro S, Tommasi E, Castellano M, Cerello P, Cheran SC, Fulcheri C: A novel active contour model algorithm for contour detection in complex objects. Proc. 2007 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, Ostuni (Italy), 2007Google Scholar
- 40.Aktouf Z, Bertrand G, Perroton L: A 3D hole closing algorithm. In “Lectures Notes in Computer Science”, Vol. 1176. Springer, 1996, pp. 36–47Google Scholar