A new motion estimation method for long-term ultrasound free respiration sequences based on multi-scale blobness enhancement filter and level set method

作者：Zhao, Yue; Shen, Yi^*; Li, Jiaxin; Jin, Jing; Wang, Qiucheng

来源：13th IEEE International Conference on Signal Processing, ICSP 2016, 2016-11-06 To 2016-11-10.

DOI：10.1109/ICSP.2016.7877796

• blobness enhancement filter
• Cross correlation methods
• High temporal resolution
• Level Set
• Level set algorithm
• long-term ultrasound sequence
• Medical ultrasound
• multi-scale

Because of its high temporal resolution and safety, medical ultrasound (US) imaging technology is widely used as a guided modality in modern therapies and examinations, especially in human liver treatment. Thus, the precise motion estimation method of the moving target becomes a hot researching domain. In this paper, a new motion estimation method is proposed for liver vessel tracking using long-term ultrasound free respiration sequences. Firstly, a multi-scale blob-like enhancement filter was implemented to extract the coarse contour of the vessel target region. Then, the level set algorithm based on Chan-Vese (CV) model was used to obtain a fine boundary of the tracking target using the coarse one. The center of the tracking target is considered as the geometrical center of the fine boundary. The proposed method is tested using the clinical US sequences and compared with the traditional cross correlation method. The results show that the proposed method is more accurate and stable than the cross correlation method.

• 出版日期2016