Answers are grounded in this paper and cite their sources.
In robot minimally invasive surgery, soft tissue keeps moving. It moves because of breathing, heartbeat, and when tools pull on it. If the control system is just designed to correct only one target position it cannot track the tissue for an extended period of time. This paper looks at three kinds of studies. The first is visual localization without markers, the second is tactile sensing at the end of the instrument, and the third is how vision and touch work together. The paper also compares three types of control methods, which are vision-led, touch-led, and fusion at information level or control level. The reviewed studies indicate that vision is good at telling where the target is and where the instrument is. Touch is good at telling the force after the instrument touches tissue, and when contact happens. The two are not the same thing, and they cannot just replace each other. When they work together, people still need to think about time delay, who is in control, and what to do when a sensor fails. Right now, not many studies test these things in a real clinical environment, or test what happens when one sensor breaks. Later, people should use tissue phantoms, animal tests, and real patients to check how big the position error is, how big the contact force is, and what the system does when something goes wrong.
No takes yet. Share an insight, caveat, or question.
Jingzhan Hu (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: