Current treatment for colorectal cancer with hepatic involvement, as well as for primary hepatocellular carcinoma, is hepatic resection. New surgical resection schemes have significantly improved the cure rates attained by hepatic resection, yet, the procedure continues to be limited, in part, by the ability of the surgeon to comprehend, in 3D, the location and extent of tumor with respect to liver anatomy. Detection of lesions is by radiological imaging techniques, with computed tomography (CT) with intra-arterial injection the current gold standard. Conventional display of transaxial CT data doesn't fully address the surgeon's needs. The ultimate objective of an imaging system for surgical planning is realistic simulation of the surgical procedure. One can imagine, some decades from now, a system in which a holographic 3D image of the patient floats in front of the surgeon who, with electronic analogs of surgical instruments, can perform the entire procedure in advance of any actual surgery. Developments in computer science and image processing lead towards this ultimate objective. Manipulation of image data into a 3D model is the impetus for development of techniques that can help accurately determine operative resectability and provide a road map for that surgery. 3D imaging has been applied to orthopedic applications, but has not been routinely used to image the liver. Although surgical techniques in orthopedics differ from those used for the liver, the overall visualization of the clinical situation in 3D is shared in both applications, and motivates our use of 3D imaging for the liver. The final goal is to develop methods for imaging the liver that will aid the surgeon in determining resectability and planning an operative approach to resection which both increases the accuracy of selection of patients who are resectable and improves the outcome in those who undergo surgical resection. In order to achieve this goal, 4 specific aims have been identified: (1) To refine acquisition techniques which optimize detection and anatomical definition of the normal liver, tumor, and hepatic vascularity for use in the construction of 3D images, (2) to develop algorithms for segmentation of liver anatomy and pathology from CT data, (3) to create an imaging system using 2D and 3D images for surgical planning for the surgeon, oncologist and radiologist, (4) to evaluate the effectiveness of the imaging system by a pilot study of the techniques in potential candidates for surgical liver resection. Achieving these aims will combine helical CT scanning, probabilistic segmentation, the volumetric rendering technique, and advanced real time imaging. Approximately 20 patients with primary hepatocellular carcinoma or metastases scheduled for liver resection will be used to evaluate and refine the techniques. It is the goal of this project to provide optimal care for these patients by combining state of the art radiologic imaging and surgical technique with advanced computer imaging.
