Vessel Attachment Nodule Segmentation Based on Mean-Shift and EM

Abstract:

For solving the segmentation problem of vessel attachment nodule, a new adaptive bandwidth chosen method based on EM is proposed and apply it into Mean-shift algorithm to segment vessel attachment nodule. This method has some advantages such as time low complexity and correct bandwidth when comparing it to the method of bandwidth chosen based on statistical analysis rule or optimized rule, Imposing the vertical orientation vectors of vessel’s gradient submitting to normal distribution and the vertical orientation vectors of nodule’s gradient submitting to uniform distribution, modeling the nodule connected vessel, and estimating model parameter by EM, extract bandwidth parameter in Mean-shift based on the weight of uniform distribution. The proposed method was tested on synthetic data set and the clinical chest CT volumes, and all the results were correct. The results revealed that the proposed method is successful in segmentation lung vessel attachment nodule.

• Periodical: Advanced Materials Research (Volumes 204-210)
• Pages: 589-595
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.204-210.589
• Online since: February 2011
• Authors: Shen Shen Sun
• Keywords: EM, Lung Nodule, Mean-Shift, Vessel Flow Feature

References

[1] Kostis W J, Reeves A P, Yankelewitz D F, et al. Three-dimensioal segmentation and growth-rate estimation of small pulmonary nodules in helical CT images[J]. IEEE Transactions on Medical Imaging 2003, 22(10): 1259-1274.

DOI: 10.1109/tmi.2003.817785

[2] Ning X, Narendra A, Ravi B. Automated lung nodule segmentation using dynamic programming and EM based classification[J]. SPIE Medical Imaging 2002, 4684(5): 666-676.

DOI: 10.1117/12.467209

[3] Okada K, Comaniciu D, Krishnan A. Robust anisotropic Gaussian fitting for volumetric characterization of pulmonary nodules in multislice CT[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence 2003, 25(2)：281-288.

DOI: 10.1109/tmi.2004.843172

[4] Sun Shen-shen, Li Hong, Hou Xin-Ran, Kang Yan, Zhao Hong, Automatic Segmentation of Pulmonary Nodules in CT Images[A], In Proceedings of the IEEE Internatioal Conference on 1st Bioinformatic and Biomedical Engineering[C], Wu Han 2007: 790-793.

DOI: 10.1109/icbbe.2007.206

[5] Gady A, Samuel G, Armato I I I, et al. Vessel tree reconstruction in thoracic CT scans with application to nodule detection[J]. IEEE Transaction on Medical Imaging, 2005, 24(4): 486-499.

DOI: 10.1109/tmi.2005.844167

[6] Jun Z, James W M, Fellow, et al. Maximum-likelihood parameter estimation for unsupervised stochastic model-based Image segmentation [J]. IEEE Transaction on Image Processing, 1944, 3(4): 404-420.

DOI: 10.1109/83.298395