Sensing and Manipulation Problems in Endoscopic Surgery: Experiment, Analysis, and Observation

Abstract

The minimally invasive nature of endoscopic surgery allows operations to be performed through small incisions, producing significantly less damage to good tissue than in conventional open surgery. Patient recovery times are thus greatly reduced. This major advantage has driven a dramatic growth of endoscopic techniques in a wide range of surgical applications. Unfortunately, the surgeon's abilities are severely hampered by the limitations of current endoscopic technology. In many ways endoscopic surgery is similar to teleoperation of a remote manipulator. Although the surgeon is physically close to the patient, the surgical environment is effectively “remote,” with sensing and manipulation transmitted through the endoscope and long instruments. Existing solutions from teleoperation applications could likely be applied to endoscopic surgery. This paper attempts to identify the major problems of current endoscopic technology, particularly in vision and manipulation. Vision issues are discussed in the first half, motivated by an experiment comparing surgeons' performance under different visual conditions. Surgeon subjects perform a three-dimensional positioning task with binocular direct vision, monocular direct vision, and a video endoscope. The lack of a stereoscopic view through the endoscope significantly slows performance of the tasks, but there are additional factors that make endoscopic viewing worse than monocular direct viewing. Relevant previous results which demonstrate the significance of interocular spacing in stereo viewing and the effect of display-control axes misalignment are also discussed. Issues of endoscopic manipulation are discussed in the second half of this paper, motivated by a suturing task comparing surgeons' performance with hand and endoscopic instruments. The constraint of endoscopic instruments' passing through the skin reduces their usable degrees of freedom and significantly increases the time to tie a suture knot. Kinematics of endoscopic instruments are compared to the human hand and arm. A model of the surgeon's precision grasp demonstrates the role of stiffness in fine motion control as an example of an advantage of the hand which is lost in the use of endoscopic instruments. The work described in this paper represents an initial effort in identifying problems and evaluating solutions in endoscopic surgery based on objective measurement of performance.