- Title
- Automatic liver segmentation from CT images by combining statistical models with machine learning
- Creator
- Luo, Su-huai; Li, Jia-ming
- Relation
- 2014 International Conference on Computer Science and Artificial Intelligence (ICCSAI 2014). Proceedings of the 2014 International Conference on Computer Science and Artificial Intelligence (Wuhan, China 20-21 December, 2014) p. 190-193
- Publisher
- DEStech Publications
- Resource Type
- conference paper
- Date
- 2015
- Description
- This paper presents a novel automatic liver segmentation algorithm which combines statistical models with machine learning. In the approach, three kinds of statistical models are developed, including statistical pose model (SPM), statistical shape model (SSM), and statistical appearance model (SAM). The algorithm contains three major processes, including prior collecting, statistical models building, and shape detecting. In prior collecting, based on benchmark of liver segmentation, the prior information about the liver is collected, including its position, pose, shape, texture, and the statistical intensity distribution of surrounding area. To fully utilise the prior information for segmentation, the statistical models building will build a support vector machine (SVM) classifier and the three statistical models. The shape detecting process will model the segmentation as a process of model evolution to derive the liver shape. Experiment results of liver segmentation on CT images using the proposed method are presented with performance validation and discussion.
- Subject
- statistical model; support vector machine; liver segmentation
- Identifier
- http://hdl.handle.net/1959.13/1317409
- Identifier
- uon:23412
- Identifier
- ISBN:9781605952147
- Language
- eng
- Reviewed
- Hits: 1033
- Visitors: 1095
- Downloads: 0
Thumbnail | File | Description | Size | Format |
---|