Share Email Print
cover

Proceedings Paper

Automatic segmentation of breast tumor in ultrasound image with simplified PCNN and improved fuzzy mutual information
Author(s): Jun Shi; Zhiheng Xiao; Shichong Zhou
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Image segmentation is very important in the field of image processing. The pulse coupled neural network (PCNN) has been efficiently applied to image processing, especially for image segmentation. In this study, a simplified PCNN (S-PCNN) model is proposed, the fuzzy mutual information (FMI) is improved as optimization criterion for S-PCNN, and then the S-PCNN and improved FMI (IFMI) based segmentation algorithm is proposed and applied for the segmentation of breast tumor in ultrasound image. To validate the proposed algorithm, a comparative experiment is implemented to segment breast images not only by our proposed algorithm, but also by the improved C-V algorithm, the max-entropy-based PCNN algorithm, the MI-based PCNN algorithm, and the IFMI-based PCNN algorithm. The results show that the breast lesions are well segmented by the proposed algorithm without image preprocessing, with the mean Hausdorff of distance of 5.631±0.822, mean average minimum Euclidean distance of 0.554±0.049, mean Tanimoto coefficient of 0.961±0.019, and mean misclassified error of 0.038±0.004. These values of evaluation indices are better than those of other segmentation algorithms. The results indicate that the proposed algorithm has excellent segmentation accuracy and strong robustness against noise, and it has the potential for breast ultrasound computer-aided diagnosis (CAD).

Paper Details

Date Published: 5 August 2010
PDF: 8 pages
Proc. SPIE 7744, Visual Communications and Image Processing 2010, 77441P (5 August 2010); doi: 10.1117/12.863028
Show Author Affiliations
Jun Shi, Shanghai Univ. (China)
Zhiheng Xiao, Shanghai Univ. (China)
Shichong Zhou, Fudan Univ. (China)


Published in SPIE Proceedings Vol. 7744:
Visual Communications and Image Processing 2010
Pascal Frossard; Houqiang Li; Feng Wu; Bernd Girod; Shipeng Li; Guo Wei, Editor(s)

© SPIE. Terms of Use
Back to Top