Detection of small bowel tumor based on multi-scale curvelet analysis and fractal technology in capsule endoscopy
详细信息 查看全文
- 作者：Gang Liua ; liugang531557127@163.com" class="auth_mail" title="E-mail the corresponding authorAuthor Vitae ; Guozheng YanaAuthor Vitae ; Shuai KuangaAuthor Vitae ; Yongbing WangbAuthor Vitae
- 关键词：Small bowel tumor ; Multi-scale curvelet ; Fractal technology ; Textural features ; Support vector machine (SVM) ; Genetic algorithm (GA)
- 刊名：Computers in Biology and Medicine
- 出版年：2016
- 出版时间：1 March 2016
- 年：2016
- 卷：70
- 期：Complete
- 页码：131-138
- 全文大小：1210 K
文摘
- •
89 valid capsule videos were collected from 105 clinical trials with new developed WCE system.
- •
A novel automatic computer-aid system is proposed for small bowel tumor detection.
- •
The textural feature set is extracted by using discrete curvelet transform and fractal technology.
- •
The GA–SVM method is proposed for feature selection and WCE image classification.
- •
Extensive comparison experiments demonstrate that the proposed system outperforms recent approaches.