摘要
目的:研究MRI和PET医学图像融合的方法并且应用在人脑神经解剖中。方法:以Chamfermatching为基本配准方法和以模糊数学为理论基础的融合方法。结果:使得融合后图像有很强的抗配准偏差能力,并且能极大程度地保留原来解剖性信息图像和功能性图像的信息。能根据融合算子的不同组合,得到倾向于不同检查仪器结果的图像。结论:通过本文方法得出的融合后图像,能在人脑解剖结构上显示出该区域的功能状态,对于了解该结构的存在、变异、病变提供了很大的便利。对人脑神经解剖学研究和临床应用都有很大的理论和实践意义。
Objective: To study MRI-PET image fusion procedure and apply it in neuroanatomy of human brain. Methods: Chamfer matching method was used as the basic registration method and fuzzy mathematics was used as the basic rationale to fuse MRI and PET images. Results: The image fusion was capable to balance registration error and could reserve the anatomical and functional information to a great extent. Combination of different operators could obtain various images. Conclusion:The fused images can show functional information of anatomic structure, which makes it easy to find out the existence, variation and pathological changes of some structure. Our findings are of great importance for foundational neuroanatomy research and clinical application.
出处
《解剖学杂志》
CAS
CSCD
北大核心
2005年第2期131-134,共4页
Chinese Journal of Anatomy
基金
国家自然科学基金(30370417)
上海市重大科技项目(03DZ19710)
