VR-planSys: towards a viable 3D surgical planning system for treating lung cancer

手术切除肺癌是治疗的首选。通过阅读计算机断层扫描（CT）图像，通过手术计划获得详细的解剖信息，对于精确确定癌症病变是否可切除以及手术如何最大限度地保留肺功能至关重要。现代多层CT扫描仪已经严重影响了外科医生处理所有信息的能力——想象一下，阅读大约800张肺腔的CT图像并将其解剖结构以三维方式映射到实际的肺腔上的工作量！目前的临床做法是将多张高分辨率CT图像（厚度一般为0.6 mm）压缩成一张厚度为2.5 - 5.0 mm的标准CT图像。这种压缩导致详细和有价值的信息丢失，原则上，应该用于更好的患者护理。基于标准CT图像的手术计划也导致大约10%的患者在接受侵入性手术后才发现他们的肺癌无法切除。基于先进的虚拟现实（VR）技术，我们的长期目标是创建VR- plansys——世界上第一个用于治疗肺癌的3D VR手术计划系统。为了适应办公室，VR- plansys将是一个独立的桌面虚拟现实系统，可以从高分辨率CT图像中生成肺腔的3D立体显示。VR-planSys的基础将是先进的触觉技术（模拟触觉），以增强与3D显示器的直观交互。我们的短期目标是开发：(a)可行的分割算法，从高分辨率肺腔CT图像中自动提取所有解剖结构；(b)直观的人机界面，正确整合视觉和触觉信息，没有感知错觉。我们的可行性研究证实，利用现有的技术和知识，VR-planSys是可以实现的。初步结果表明，在手术计划中，观看肺腔的3D显示比阅读标准CT图像具有显着优势。VR-planSys将改善患者护理；促进发展新的外科手术；搭建基于vr的手术计划和模拟的桥梁，用于治疗其他疾病。

Surgical removal of lung cancer is the treatment of choice. Detailed anatomic information, acquired via surgical planning by reading computed tomographic (CT) images, is crucial to precisely determine whether a cancer lesion is removable and how a surgical procedure preserves maximal lung function. Modern multi-slice CT scanners have severely taxed the surgeon's ability to handle all information - imagine the workload of reading about 800 CT images of a lung cavity and mapping anatomic structures from them onto the actual lung cavity in three dimensions! The current clinical practice is to compress multiple high-resolution CT images (typically 0.6 mm thickness) into one standard CT image of 2.5 - 5.0 mm thickness. This compression results in the loss of detailed and valuable information and, in principle, should be used for better patient care. Surgical planning based on standard CT images also causes about 10% of patients undergoing invasive procedures only to discover that their lung cancer is not removable. Based on advanced virtual reality (VR) technologies, our long-term objective is to create VR-planSys - the world's first 3D VR surgical planning system for treating lung cancer. To fit into an office, VR-planSys will be a self-contained desktop VR system and generate 3D stereoscopic displays of lung cavities from high-resolution CT images. Underpinning VR-planSys will be advanced haptics (simulating the sense of touch) to enhance intuitive interaction with the 3D displays. Our short-term objectives are to develop: (a) viable segmentation algorithms to automatically extract all anatomic structures from high-resolution CT images of lung cavities; and (b) intuitive human-computer interfaces to correctly integrate visual and haptic information without perceptual illusions. Our feasibility work confirms that VR-planSys is achievable using current technologies and knowledge. The initial results demonstrate that viewing 3D displays of lung cavities has significant advantages over reading standard CT images for surgical planning. VR-planSys will improve patient care; facilitate the development of new surgical procedures; and bridge VR-based surgical planning and simulation for treating other diseases.