After the breakthrough of modern minimal-invasive keyhole surgery in the mid-1970s, the laparoscopic technique has been the gold standard for the majority of procedures in abdominal and gynecological surgery. The method requires only small incisions and utilizes a set of tubes to serve as tunnels for a variety of surgical instruments which minimize trauma, infection risk, healing time and overall surgical effort. Typically, two to four of these trocars are used to insert a camera or other instrumentation such as clamps, electrosurgical Bovie scalpels, forceps, or tissue removal bags into the inflated abdomen. The operating surgeon then looks at a screen that displays a live broadcast from within the patient’s body while he manually applies the instruments. However, because the surgeon performs the procedure without direct contact and using instruments snaked through small trocars, the procedure can be cumbersome and difficult to perform without a high degree of training.

Layers of computation between physician and patient calls into question our typical notion of authority and liability.

In 2000, the Food and Drug Administration approved a novel device for medical practice that quickly replaced many forms of conventional laparoscopic surgery: the American corporation Intuitive Surgical’s surgery robot “da Vinci,” aptly named after the artist whose work included the invention of the automaton. The device aimed to overcome some of the difficulties common to conventional laparoscopic surgery¹ by providing an easier, more immediate approach to the patient’s body. Though several shortcomings still limit its adoption in all cases, machines like “da Vinci,” enable surgeons to operate at a distance, minimizing the risk of infection and improving patient outcomes. By peering into a console that enables tele-vision and tele-steering, the robot can manipulate instruments under the operator’s guidance with greater consistency and precision.

Fourteen years after its commercial launch the $2 million device has been approved for urological, gynaecological, non-cardiovascular thoracoscopic, thoracoscopically assisted cardiotomy and general surgical procedures. As of 2008, the western hemisphere saw robotic surgery become a standard therapy for prostate cancer and other abdominal cancers with 85 percent of American patients who require radical removal of the prostate and uterus receiving robotically assisted treatment.²

Complications

Naturally as devices like “da Vinci” become more common and increasingly advanced, they begin to raise a variety of medical and ethical questions, they also, however, become the domain of artistic inquiry. Now that these elaborate machines are used to perform surgical procedures more precise and sparing,^3,4 than previously deemed achievable by the human hand it becomes difficult to attribute surgical actions wholly to the physician. Moreover, when these machines are mediated through real-time imagery and haptic input that puts layers of computation between physician and patient, our typical notion of authority and liability is called into question. Who or what is operating on the patient and thus responsible for the success or failure of the medical procedure? In the event of a complication, who is legally actionable? The surgeon, the robot, the manufacturer? Shockingly, these questions came years after the first introduction of medical robots and largely as the result of a small contingent of artists rather than physicians. Critical analysis of assistive robots still remains largely absent inside the medical community.

Several of the advancements Intuitive Surgical made with “da Vinci” bare heavily on these issues and serve to illustrate where our current sensibilities about assistive technologies break down. Unlike previous devices, “da Vinci” is split into two parts: an “input” and an “effector” effectively decoupling surgical input to enable precision work on a smaller scale. “da Vinvi’s” console for instance uses two cameras and two displays to create a fused 3D image that replicates the space within the body which the surgeon’s use to navigate.

Rather than the rigid instruments used in laparoscopy, the robot’s arms have joints that allow for wrist-like movements as registered by elaborate, glove-like controllers. These controllers mimic the ergonomics of conventional surgical instruments but can do so with dampened movement to reduce the jitter of the hand and scaled motion appropriate for “da Vinci’s” miniaturized instrumentation. The robot can thus follow the surgeon’s lead, carrying out his movements and actions and enabling him to reach behind objects, knot faster and cut along surfaces not directly facing the camera system.

Furthermore, the connection between console and effector robot is established through a high speed data cable enabling the surgeon to work in another room, another hospital, or even another country. He is usually situated in the same operating theatre, though early feasibility trials have already shown that trans-Atlantic surgery⁵ and teleoperation with the help of airborne radio link via unmanned aerial vehicles⁶ are possible.

It’s conceivable in the near future international teams of experts might perform complicated operations together remotely. A patient may only get to know the surgeons who do the basis work at his local hospital in Germany, while a specialist from Taiwan would be on hand to take over at critical moments. Legislation prohibits such international surgeries in many countries today, but once studies sufficiently demonstrate their benefit, they might become legal, perhaps even common procedures.

Imagery & Authority

While many urologists and surgeons appreciate these innovations and argue that they alleviate former ergonomic obstacles (though the efficiency is much-debated), concerns about the role of the image are seldom expressed. A surgeon peering at a screen while a robot reaches over the patient’s exposed body presents an entirely new relationship to medical imagery, far from the conventional usage of x-ray or MRI. When images are used as physician-patient interfaces requiring vision and action simultaneously, they cease to be merely diagnostic tools but also therapeutic recipients, manipulable visual objects that transfer pictorial changes into reality.⁷ In this regard, they gain the agency to instruct the surgeon – a notion that is reflected by the medical community’s umbrella term for several such technologies: “image guidance.”

Image guided techniques thus sit in a grey area between bodily performance of medical procedures and a surgeon simply giving instruction for machines to execute under its own volition. We cannot not leap to the conclusion that these procedures are merely the manipulation of a visual interface interpreted by the medical device, nor can we say with confidence that they are performed entirely under the surgeon’s control.

The majority of the physicians who work with surgery robots refute the later notion, emphasizing such technological novelties simply provide windows into the body for performance of minimally-invasive procedures, but frequently neglect to acknowledge these images as efficacious agents that present a highly artificial, constructed reality. Moreover, they rarely grant any authority or liability to the machine, regarding it as merely an extension of their own hands.

In this way, the simple conditions for laparoscopy and its relatively simplex relations between surgeon, image, tool and patient are often applied to the more complex case of robotic surgery. But while the manual craft of conventional laparoscopic keyhole surgery requires no computer aid, robotic surgeries employ a sophisticated computational apparatus to mimic 3-dimensional vision as well as interpret the surgeon’s hand movements. As a result, both the procedure’s use of augmented vision and its surgical dexterity can be attributed in-part to the algorithms’ own agency rather that of the operator.

Intrusion

Any surgery – as detached as it might seem on the screen – constitutes an act of penetration. While surgeons fully agree on the fundamental responsibility and caution involved with such a procedure, the patient’s impression of a robotic treatment may differ. In Liz Magic Laser’s video Mine (2009) a staged performance of the “da Vinci” robot ripping apart her purse, spilling her cosmetics, and browsing through her credit cards provides an impressive depiction of the intrusive, forceful, and violative nature of surgery, especially if performed by a non-human being. Curiously, the “Mine” presents the internals of the body as foreign material when replaced with the familiar context of a woman’s handbag which results in an unexpectedly visceral reaction to what should be a less-invasive act.

The notion that it’s the machine rather than the surgeon performing the surgery occasionally lingers with patients who have had robotic operations. And to stay this fear, surgical robots should remain reliable, predictable and comprehensible enough to be regarded as tools rather than obscure machines. Due to competitive conditions within the privatized medical industry however, the inner workings of these machines typically remains opaque, inexplicable, even suspicious. In fact, the market leader for surgical robots, Intuitive Surgical, incurred a number of lawsuits due to technical errors which has already created a cottage industry of lawyers in the US who specialize in claims against “da Vinci.”

Conflicting Opinions

This Scepticism will likely continue to grow as future generations begin to incorporate more advanced operator feedback. Already in development are robots with pressure sensors that gather information about tissue stiffness by palpating arteries or tumours at which point that information is processed and fed to a visual display and fused to create a diagnostic map of the operational area, effectively rendering a hybrid view of the appearance and touch of a tissue. Paired with the ability to give haptic feedback through the controllers, an intelligent robot could tell surgeons where crucial structures are threatened and even hinder his movements.

If we imagine a complicated, high-risk surgery (e.g. the removal of a highly perfused organ such as the liver due to an abscess) a carefully thought out strategy to prevent blood loss and future underperfusion would be essential. Haptic feedback could prove helpful in avoiding vital blood vessels. However, it is uncertain how reliable the algorithms that determine the controller’s resistance might be: While today, surgeons rely on pre-operatively obtained images to plan ahead, and will continue to use them once robotic liver surgery is possible, they will likely come find themselves in a situation when the robot identifies crucial structures where the MRI does not.

We can anticipate authoritative conflicts of this nature but will have to find reasonable and empirically proven ways to resolve them. While the development of such features is still in the early stages, the National Science Foundations has been making a concerted effort to push this research forward. Most notably their project entitled “Complementary Situational Awareness of Human-Robot Partnerships” which began last October.^{8 9} With substantial funds and motivation to push assistive medical devices forward, it’s only a matter of time before further innovation raises serious questions about the credibility and liability of medical decisions.

Intimacy & Agency

Yuri Ancarani’s short film da Vinci” (2012) seeks to explore this delicate relationship between surgeon and robot. The 25-minute video follows the work routine of Dres Franca Melfi and Olivia Fanucchi performing robotic lung surgery at the university hospital Pisa. Ancarani arranges blue-tinted laparoscopic images, scenes from the operation theatre and the robot, adding a low-pitched ambient soundscape of faint heartbeats. Shown at the Arsenale at Venice Biennale 2013, the cool-coloured video piece highlights the sterile, clinical setting of what appears more like a quiet laboratory than an operating room. Surgeons and nurses remain still and calm as the whirring robot arms breeze about the patient’s body. Meanwhile, the surgery staff fold their hands, look at the screens, and await further instructions. Contrasted by scenes of the robot’s arms, they appear to stand impassively outside of the intimate relationship of surgeon and robot.

Vincent Tiley’s video piece Theotokos (2010) instead focuses on aspects of the surgery robots “artificial intelligence.” The 14-minute video provides no indication of the robot’s controller, but rather creates the impression that the machine operates its instruments autonomously. Tiley follows the robot’s movements in close-up shots set in a dark studio, where “da Vinci” is depicted as a spider spinning its web, carefully examining the artist’s body, and operating on the transparent skin of an icosahedron. As the robot gently strokes Tiley’s hair, pinches his cheeks and carefully lifts his lips to reveal the teeth, it appears as an autonomous and seemingly intelligent machine. Tiley keeps still as the tweezers cautiously stroke his eyelashes and along his brow but the robot’s uncanny anonymity resembles that of an alien probing a human being and instills a lingering sense of helplessness and terror.

While physicians over the years have been mostly concerned with the outcome and efficiency of their procedures, we are now entering an era of novel challenges posed by technology in which ethical and philosophical concerns may not necessarily coincide with clinical efficacy. As medicine continues to advance, it will become increasingly important to understand how new procedures are situated in relation to scientific, ethical, medical, and legal issues that arise – and to do so while ensuring these practices coincide with pragmatic medical perspectives.

Advances in medical interfaces for instance, require we gain knowledge about how they are involved in the immediate diagnostic and therapeutic decision-making processes, what their limitations are, and how much authority we are willing to concede. In order to use new technologies responsibly, our understanding of their operation and how they are perceived must be manifold. Artistic materials may provide one means to gain this kind of insight, be it conveying an outsider’s deeply emotional reaction to a new technology, or revealing nuances of a physician-patient relationship we once believed to be straightforward. As the complexity of our medical systems grows however, it will demand a corresponding increase in our effort to understand them.

---