EasyVis: Flexible, immersive three-dimensional laparoscopic surgical visualization through multi-camera arrays

点击翻译按钮获取中文摘要

Abstract Laparoscopic or “minimally invasive” procedures are frequently performed in surgery today. It provides numerous clinical benefits to patients and has become the gold standard surgical procedure for many intraabdominal and thoracic procedures. Existing laparoscopes provide visualization of the operative field through a video camera inserted into the patient’s abdomen. It occupies a dedicated laparoscopic port and requires a camera operator to maneuver and navigate the camera for visualization. The field of view is narrow and the depth of the scene is difficult to infer. To facilitate surgery, the camera needs to be in proximity to the operative field, making it prone to smudging, fogging, and splatter that may obscure visualization and require cleaning. The position of the camera may also interferes with the operating surgical instruments. To address these deficiencies associated with current laparoscopic visualization, in this project, we will develop a paradigm-shift, integrated, panoramic, flexible, immersive 3D laparoscopic visualization system called EasyVis that could significantly improve the efficiency of laparoscopic surgery. EasyVis directly integrates multiple microcameras and their peripherals including light sources and miniaturized projectors with the surgical ports. It provides an uninterrupted, intra-abdominal, flexible, and immersive 3D view from arbitrary virtual viewpoints and viewing angles, and close-up 3D visualization of any specific area, all under direct, hands-free, easy and full control of the operating surgeon through voice commands, eliminating the need of camera navigation. EasyVis does not occupy any extra surgical port and solves the problems associated with smudging, fogging and splatter, and interference with and occlusion of instruments. With EasyVis, the surgeon is provided with enhanced visualization (better quality, easy control) and less interruption due to laparoscope manipulation during the operation. Three specific aims will be pursued. First, we will develop an imaging system including heterogeneous microcamera arrays, light sources, and miniature projectors. This imaging system provides flexibly selectable viewpoints and viewing angles with auto-focus and zooming capabilities and real-time scene stabilization. A system that can deploy the imaging system through a surgical port will also be developed. Second, we will develop algorithms to achieve flexible global and local views, and immersive, any-view 3D visualization, all through a human-computer interface worn by the operating surgeon. Third, we will validate our laparoscopic visualization system with inanimate and animal models.