Local-weighted Citation-kNN algorithm for breast ultrasound image classification

Abstract A new multiple-instance learning (MIL) algorithm which combines local distribution feature of samples with Citation-kNN is proposed. The local distance feature and local sparseness feature are considered. The voters are weighted according to their local distribution. The differently weighted schemes and their combinations are applied to Musk benchmark data set and Breast Ultrasound (BUS) Images. For musk data set, a bag represents a molecule. And instances in a bag represent low-energy conformations of the molecule. For BUS image classification, the image is viewed as a bag and its sub-regions are considered as the instances of the bag. And the image classification problem is converted into a MIL problem. In comparison with Citation-kNN and other methods, the proposed algorithm demonstrates competitive classification accuracy and adaptability.

[1]  Thomas G. Dietterich,et al.  Solving the Multiple Instance Problem with Axis-Parallel Rectangles , 1997, Artif. Intell..

[2]  Catherine Blake,et al.  UCI Repository of machine learning databases , 1998 .

[3]  B. Garra,et al.  Improving the Distinction between Benign and Malignant Breast Lesions: The Value of Sonographic Texture Analysis , 1993 .

[4]  Andrew W. Moore,et al.  Locally Weighted Learning , 1997, Artificial Intelligence Review.

[5]  J. Alison Noble,et al.  Ultrasound image segmentation: a survey , 2006, IEEE Transactions on Medical Imaging.

[6]  Tomás Lozano-Pérez,et al.  A Framework for Multiple-Instance Learning , 1997, NIPS.

[7]  A. Stavros,et al.  Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. , 1995, Radiology.

[8]  Thomas Hofmann,et al.  Support Vector Machines for Multiple-Instance Learning , 2002, NIPS.

[9]  James R. Foulds,et al.  A review of multi-instance learning assumptions , 2010, The Knowledge Engineering Review.

[10]  J W Sayre,et al.  Benign versus malignant solid breast masses: US differentiation. , 1999, Radiology.

[11]  Xianglong Tang,et al.  Fully automatic and segmentation-robust classification of breast tumors based on local texture analysis of ultrasound images , 2010, Pattern Recognit..

[12]  Yixin Chen,et al.  Image Categorization by Learning and Reasoning with Regions , 2004, J. Mach. Learn. Res..

[13]  Oded Maron,et al.  Multiple-Instance Learning for Natural Scene Classification , 1998, ICML.

[14]  Zhi-Hua Zhou,et al.  Solving multi-instance problems with classifier ensemble based on constructive clustering , 2007, Knowledge and Information Systems.

[15]  Zhi-Hua Zhou,et al.  Multi-Instance Learning from Supervised View , 2006, Journal of Computer Science and Technology.

[16]  Jun Wang,et al.  Solving the Multiple-Instance Problem: A Lazy Learning Approach , 2000, ICML.

[17]  Tomás Lozano-Pérez,et al.  Image database retrieval with multiple-instance learning techniques , 2000, Proceedings of 16th International Conference on Data Engineering (Cat. No.00CB37073).