4D ultrasound image processing for computer aided diagnostic applications

每年有数以千计的加拿大人死于汽车和运动事故造成的看不见的内伤，特别是最近的军事和维和行动。尽可能快速、准确地定位这些内伤的位置和性质，极大地改善了患者的预后。然而，在严峻和可能充满敌意的前线环境中，对医疗和搜救人员来说，快速诊断内伤仍然是一个挑战。此外，便携式和易于使用的4D(即实时3D)非侵入性医学成像系统还没有商业化，主要是因为尚未解决的主要技术和工程挑战。为了应对这些挑战，建议开发一种4D超声计算机辅助诊断成像能力，该能力将具有最低限度的培训要求，从而允许医务人员在急救和创伤诊断操作中自由使用新的4D超声概念。该提案涉及NSERC审查委员会的建议，该建议与203828-2010年的前一项申请有关。NSERC Discovery补助金项下的拟议研究将仅支持研究生参与以下算法的开发：(1)具有体积可视化的自适应3D波束形成算法，这将扩展申请人先前在同一领域的开发；(2)计算机辅助器官分割与决策支持处理相结合，为非侵入性检测内伤提供自动诊断能力并促进图像引导手术；以及(3)在重症监护和急诊室以及对超声成像培训要求最低的偏远地区。这两名研究生(硕士、博士)将接受3D超声成像和先进3D波束形成处理的跨学科融合培训，他们还将从已经到位的价值超过1000万美元的最先进设备中受益。最终，该项目将改善对加拿大人的护理，并将培训新一代研究人员和工程师，他们可以进一步推动这些至关重要的生命技术。

Thousands of Canadians die each year from non-visible internal injuries resulting from automobile & sports accidents, and - especially of late - military and peacekeeping engagement. Pinpointing the location and nature of these internal injuries as rapidly and precisely as possible dramatically improves patient outcomes. However, rapid diagnosis of internal injuries in an austere and possibly hostile front-line environment remains a challenge for medical and search and rescue personnel. Moreover, portable and easy-to-use 4D (i.e. real time 3D) non-invasive medical imaging systems are not yet commercially available, primarily because of unresolved major technological and engineering challenges. To address these challenges, it is proposed to develop a 4D ultrasound computer aided diagnostic imaging capability that will have minimal requirements for training, thus allowing medical practitioners to freely use the new 4D ultrasound concept in Emergency and Trauma diagnostic operations. This proposal addresses the recommendation of the NSERC Review Committee - 1510, relevant with the previous application, 203828-2010. The proposed investigations under this NSERC Discovery Grant will support only Graduate Students to be involved in algorithmic development of (1) adaptive 3D beaforming algorithms with volumetric visualization, which will extend the applicant's previous developments in the same area, (2) computer aided organ segmentation integrated with decision support processing for providing automated diagnostic capabilities for non-invasively detecting internal injuries and facilitating image guided surgery, in intensive care & emergency units and in remote areas with minimum requirements for training on ultrasound imaging. The two graduate (MASc, PhD) students that will be trained in the interdisciplinary fusion of 3D ultrasound imaging and advanced 3D beamforming processing will benefit also from over $10 million dollars in state-of-the-art equipment already in place. Ultimately, the project will lead to improvement of care for Canadians, and will also train a new generation of researchers and engineers who can further advance such life-critical technologies.