Approach-specific, multi-GPU, multi-tool, high-realism neurosurgery simulation

描述（申请人提供）：特定入路，多gpu，多工具，高真实感神经外科模拟神经系统疾病的外科治疗是复杂的，需要高级培训。然而，医院紧凑的实习日程给年轻外科医生掌握如此复杂手术所需的技术带来了挑战。手术模拟器是为复杂手术提供高级培训的强大工具；在不降低发病率和死亡率的情况下加快对住院医生的培训；提高病人的治疗效果与手术经验明显相关的技能。手术模拟训练已经提出了近十年，计算技术和算法发展的进步使其成为一种更现实的训练选择。双盲验证研究表明，接受过模拟器训练的住院医生在执行复杂手术（如腹腔镜胆囊切除术）方面，系统地优于没有接受过模拟器训练的住院医生。然而，尽管将模拟与传统框架结合使用显示出优势，但模拟在外科教育中仅取得了有限的进展，特别是在神经外科领域。本提案的总体目标是开发和评估一种交互式神经外科模拟器，该模拟器在临床上是现实的，并且得到了很好的验证。在第一阶段，我们实现了一个交互式神经外科模拟器的原型，用于训练医学生切除脑干深部肿瘤。该仿真器由多分辨率、区域可分网格划分算法、高效且物理逼真的非线性模型有限元公式和GPU加速组成。我们的临床合作者认为我们的1期模拟器在肿瘤切除手术的主要组成部分是成功的。然而，他们也指出，考虑到血管系统将会受益，因为血管系统的操作对肿瘤切除和许多其他神经外科手术至关重要。在这个第二阶段的项目中，我们建议用模拟血管系统所需的额外组件来扩展和完善我们的模拟器，并构建一个用于动静脉畸形（AVMs）手术治疗的模拟器，动静脉畸形是涉及脑血管系统的最复杂的手术之一。我们将与北卡罗来纳大学神经外科学系的神经外科医生合作，对神经外科模拟器进行有效性研究。将开发的交互式模拟器将促进复杂且需要高级培训的神经外科手术的培训。这将转化为更少的手术室错误，降低患者发病率和改善患者预后。

DESCRIPTION (provided by applicant): Approach-specific, Multi-GPU, Multi-tool, High-realism Neurosurgery Simulation Surgical treatment of neurological diseases is complicated and requires advanced training. However, compressed internship schedules in hospitals creates a challenge to young surgeons to master techniques needed for such complex procedures. Surgical simulators are powerful tools to provide advanced training for complex procedures; accelerate the training of residents without a penalty in morbidity and mortality; and improve skills where patient outcome clearly correlates with surgical experience. Surgical simulation training has been proposed for nearly a decade, and advancement in computing technology and algorithm development has now made it a more realistic training option. Double-blind validation studies have shown that surgical residents trained with a simulator systematically outperformed residents without such training in performing complex procedures such as laparoscopic cholecystectomies. However, despite demonstrated advantages of using simulation in conjunction with traditional frameworks, simulation has only made limited inroads in surgical education, particularly in the field of neurosurgery. The overall goal of this proposal is to develop and evaluate an interactive neurosurgery simulator that is clinically realistic and well-validated. During Phase I, we implemented a prototype interactive neurosurgery simulator for training medical students on the resection of deep seated tumors in the brain stem. The simulator consists of a multi-resolution, sulcal- separable meshing algorithm for pterional approach, an efficient and physically realistic non- linear model FEM formulation, and GPU acceleration. Our clinical collaborators deemed our Phase 1 simulator a success for major components of tumor resection procedures. However, they also noted that it would benefit from consideration of vasculature, as manipulation of vasculature is critical to tumor resection and many other neurosurgical procedures. In this Phase 2 project, we propose to extend and refine our simulator with additional components required to simulate vasculature and build a simulator application for surgical treatment of arteriovenous malformations (AVMs), one of the most complex surgeries involving brain vasculature. We will conduct a validity study of the neurosurgery simulator in collaboration with neurosurgeons in the Department of Neurosurgery at the University of North Carolina. The interactive simulator that will be developed will facilitae training in neurosurgical procedures that are complicated and require advanced training. This will translate to fewer operating room errors, reduced patient morbidity and improved patient outcomes.