Sci-Thur PM: YIS - 02: Intraoperative Guidance for Minimally Invasive Abdominal Surgery Using Fused Video and Ultrasound Images: A Phantom Study

Many abdominal surgery procedures are now performed minimally invasively. We consider tumour resection, where surgeons use a laparoscopic camera to view the organ surface and a laparoscopic ultrasound(US) probe to visualize the tumour to plan and perform the excision. Conventionally, images are displayed separately and are typically presented in 2D. Therefore, the surgeon has to look back and forth between the images and mentally map the US onto the video to determine the tumour location relative to the surface. Furthermore, the 2D nature of the images decreases depth perception. To address these limitations, we developed an augmented reality visualization that fusesimages in a common 3D environment. Instruments were tracked using sensors spatially identified with a magnetic field generator. Through calibration, their image locations were determined in real time. The accuracy of the camera and UScalibrations was determined both relative to the tracking system and to each other using target localization. We evaluated the efficacy of the fusion with a phantom experiment. A surgeon performed tumour resections on polyvinyl alcohol‐cryogel phantoms under the guidance of the conventional visualization and the fusion system presented in 2D and in 3D. The target localization error was 1.20±0.08mm for the camera, 1.85±0.14mm for the US, and 2.38±0.11mm between the camera and the US. Early results demonstrate a faster resection planning time using fusion compared to the conventional setup while maintaining similar margin accuracy. This study supports the implementation of fusion for guidance of time‐sensitive resection tasks performed under conditions of warm ischemia.