Open source diffusion MRI technology for brain cancer research

 描述(申请人提供)：利用水扩散的测量，dMRI可以对组织的微观结构和细胞取向提供独特的见解。在神经外科脑癌研究中，dMRI是现有的唯一提供脑白质连接(纤维束)轨迹信息的方法。神经外科医生的目标是在手术切除肿瘤时保留关键的纤维束。DMRI还提供了定量测量，有助于确定脑肿瘤的边界，或有助于区分肿瘤的浸润性和水肿性。神经外科领域越来越多的人意识到，扩散模型必须超越目前的扩散张量的临床标准，才能更好地了解纤维束的解剖准确性。但几个信息学挑战阻碍了dMRI的进步，使临床癌症研究人员难以接触到：1)dMRI的进步不受商业临床软件的支持，2)dMRI研究软件不是为临床癌症环境而设计的，以及3)缺乏通用的文件格式标准阻碍了dMRI软件包之间的互操作性。与其他流行的dMRI软件包不同，社区软件包3D Slicer 4.0(www.slicer.org)具有独特的定位，可以支持脑癌的新型临床研究，因为它从一开始就是为特定患者的癌症研究而设计的。3D Slicer软件包是一个基于社区的开源软件平台，2013年全球Slicer总下载量为68629次。虽然Slicer目前的dMRI能力可与商业脑癌神经元导航软件中提供的技术相媲美，但3D Slicer中提供的基本扩散张量模型已不再是最先进的研究成果。 它的缺点包括纤维束的解剖不准确和DTI测量的非特异性。我们建议开发开源软件基础设施和关键的临床相关工作流程，以迈向更先进的dMRI技术，使用3D Slicer进行开源癌症研究。此外，我们建议通过开发一个独立的符合标准的dMRI纤维束成像文件格式的库来提高文件格式的互操作性，该库基于最新提出的用于MR扩散的纤维束成像存储的DICOM补充。我们将与当地和国际神经外科脑癌研究人员以及我们的前列腺癌研究合作者合作，他们都在研究中使用3D Slicer。我们的软件传播将利用基于社区的Slicer软件现有的基础设施。预期的结果是在开源软件3D Slicer中提供一套最先进的dMRI工具和一个符合标准的纤维束照相文件格式库。我们希望这项开源的dMRI技术将使脑癌的新临床研究成为可能。

 DESCRIPTION (provided by applicant): Using measurements of water diffusion, dMRI can give unique insights into the microstructure and cellular orientation of tissues. In neurosurgical brain cancer research, dMRI is the only existing method that provides information about the trajectories of the white matter connections (fiber tracts). Neurosurgeons aim to preserve key fiber tracts when surgically removing tumors. dMRI also provides quantitative measurements that may aid in defining the borders of brain tumors, or in distinguishing tumor infiltration from edema. There is a growing awareness in the neurosurgery community that diffusion models must move beyond the current clinical standard of the diffusion tensor for better anatomical accuracy of fiber tracts. But several informatics challenges prevent advances in dMRI from easily reaching clinical cancer researchers: 1) advances in dMRI are not supported by commercial clinical software, 2) dMRI research software is not designed for clinical cancer settings, and 3) a lack of common file format standards prevents interoperability between dMRI software packages. Unlike other popular dMRI packages, the community software package 3D Slicer 4.0 (www.slicer.org) is uniquely positioned to enable novel clinical research in brain cancer because it was designed from the start for patient-specific cancer research. The 3D Slicer software package is an open-source community-based software platform, with 68629 total Slicer downloads around the world in 2013. While the current dMRI capabilities of Slicer are comparable to the technology available in commercial brain cancer neuron navigation software, the basic diffusion tensor model available in 3D Slicer is no longer state of the art for research. Its drawbacks include anatomical inaccuracies in fiber tracts and non-specificity of DTI-derived measurements. We propose to develop the open-source software infrastructure and key clinically-relevant workflows necessary to move toward more advanced dMRI technologies for open-source cancer research using 3D Slicer. In addition, we propose to improve file format interoperability by developing a standalone standards- compliant library for dMRI tractography file formats, based on the newly proposed DICOM supplement for MR diffusion tractography storage. We will collaborate with local and international neurosurgical brain cancer researchers as well as our prostate cancer research collaborators, all of whom use 3D Slicer in their research. Our software dissemination will leverage the infrastructure in place for the community-based Slicer software. The expected outcome is a state-of-the-art suite of dMRI tools in the open-source software 3D Slicer and a standards-compliant tractography file format library. We expect that this open source dMRI technology will enable novel clinical research in brain cancer.