基于多模态医学图像处理的多维可视化辅助诊疗关键技术研究

项目名称： 基于多模态医学图像处理的多维可视化辅助诊疗关键技术研究

项目编号： No.U1401252

项目类型： 联合基金项目

立项/批准年度： 2015

项目学科： 管理科学

项目作者： 李伟生

作者单位： 重庆邮电大学

项目金额： 242万元

中文摘要： 基于多模态医学图像处理的多维可视化辅助诊疗技术因为扩展了时空覆盖范围、增强了系统的鲁棒性而成为医学图像领域发展的关键问题之一。本项目主要就多模态医学图像处理面临的器官与组织的准确分割、高精度配准、有效融合，以及传统的可视化大多停留在数据的显示层面，难以精确地掌握器官内管道的分布及病灶信息等问题，从多模态医学图像处理的基本问题和模型着手，利用临床CT、MRI、PET等多模态医学影像建立多维可视化的数字化结构解剖模型，实现对器官及组织的各个参数的精确测量，增强医学影像的可靠性、稳定性及容错能力。主要研究内容包括：基于距离正则项水平集模型的医学图像分割方法、基于感兴趣点的最大互信息多模4D医学图像配准方法、基于多方向多尺度邻域距离分解的多模态医学图像融合方法、多模态医学图像可视化传递函数研究与可视化表达方法等，并将研究成果集成入基于多模态医学图像处理的多维可视化系统进行分析和验证，辅助临床诊疗。

中文关键词： 多模态；医学图像处理；多维可视化；辅助诊疗

英文摘要： Since visualization-assisted diagnosis technology have extended the range of spatio-temporal covering and enhanced the robustness of diagnosis system, the multidimensional visualization-assisted diagnosis and treatment technology based on multimodality medical image processing is becoming one of the key problems in the field of bio-medical images. Traditional visualization methods only focus on how to transfer information from data to view. It is difficult to achieve mastering organ interior vessel distribution and focal information accuracy. In this proposal, we take some fundamental issues and models in multimodality medical image processing as breakthrough point; for instance, organ or tissue images segmentation, registration, fusion etc., using multimodality medical images such as CT, MRI, PET to establish multidimensional visualization based numerical anatomical models. Consequently, each parameter of organ or tissue will be measured accurately. Meanwhile, the reliability, stability and tolerance of visualization-aided diagnosis will be enhanced. The research topics of this proposal mainly include: regularization level set model based multimodality medical image segmentation, multimodality 4D medical image registration based on interesting point detection and maximum mutual information methods, multidirectional and multi-scale neighbor distance decomposition based multimodality medical image fusion, and the transform functions designing as well as visual expression methods in multimodality medical image visualization. The research results of this project will develop a multidimensional visualization system and it will be used to assist clinical medicine diagnosis.

英文关键词： Multimodality;Medical Images Processing;Multidimensional Visualization;Assisted Diagnosis and Treatment