Feasibility Study of Signal Similarity Measurements for Improving Morphological Evaluation of Human Brain with Images from Multi-Echo T2-Star Weighted MR Sequences
Signal correlation measurement has been widely used for segmenting specific tissues,localizing abnormal regions and analyzing functional areas in dynamic imaging modalities.In this paper,we discussed the feasibility of similarity mappings derived from six signal coefficient measurements in improving morphological evaluation of human brain.These images are from a digital phantom and four normal volunteers scanned by multi-echo T2-star weighted MR sequences.Simulation studies have shown that similarity mappings from cross-correlation,normalized cross-correlation,mean square error and cubed sum coefficient are not helpful in distinguishing the reference region from its surrounding tissues.Clinical experiments were focused on similarity coefficient mapping (SCM) and improved SCM (iSCM).Final results have demonstrated comparative capacity of SCM and iSCM in improving image quality from quantitative metrics and visual analysis.
Signal correlation measurement similarity mapping morphological evaluation multi-echo T2-star weighted MR sequences
Shaode Yu Xuyin Cheng Yaoqin Xie
Shenzhen Institutes of Advanced Technology, CAS, China. 518055, Shenzhen, China Anqing First Peoples Hospital, China. 246004, Anqing, China
国际会议
The Third International Coference on Health Information Science(HIS2014)2014年第三届健康信息学国际学术会议
深圳
英文
237-247
2014-04-22（万方平台首次上网日期，不代表论文的发表时间）