Learning pathological characteristics from user's relevance feedback for content-based mammogram retrieval
Wei, Chia-Hung and Li, Chang-Tsun (2006) Learning pathological characteristics from user's relevance feedback for content-based mammogram retrieval. In: 8th IEEE International Symposium on Multimedia, San Diego, CA, DEC 11-13, 2006. Published in: ISM 2006: EIGHTH IEEE INTERNATIONAL SYMPOSIUM ON MULTIMEDIA, PROCEEDINGS pp. 738-741.Full text not available from this repository.
Content-based image retrieval (CBIR) has been proposed to address the problem of image retrieval from medical image databases. Relevance feedback, explaining the user's query concept, can be used to bridge the semantic gap and improve the performance of CBIR systems. This paper proposes a learning method for relevance feedback, which utilizes probabilistic model to generalize the 2-class problem and provide an estimate of probability of class membership. To build the probabilistic model, support vector machine (SVM is applied to classify the mammograms, and then scale them to the probability of class membership. Experimental results show that the proposed learning method can effectively improve the average precision ratefrom 40% to 62% through five iterations of relevance feedback rounds.
|Item Type:||Conference Item (UNSPECIFIED)|
|Subjects:||Q Science > QA Mathematics > QA76 Electronic computers. Computer science. Computer software
T Technology > TK Electrical engineering. Electronics Nuclear engineering
|Series Name:||IEEE International Symposium on Multimedia-ISM|
|Journal or Publication Title:||ISM 2006: EIGHTH IEEE INTERNATIONAL SYMPOSIUM ON MULTIMEDIA, PROCEEDINGS|
|Publisher:||IEEE COMPUTER SOC|
|Number of Pages:||4|
|Page Range:||pp. 738-741|
|Title of Event:||8th IEEE International Symposium on Multimedia|
|Location of Event:||San Diego, CA|
|Date(s) of Event:||DEC 11-13, 2006|
Actions (login required)