| Article Access Statistics|
| Viewed||347 |
| Printed||14 |
| Emailed||0 |
| PDF Downloaded||13 |
| Comments ||[Add] |
Click on image for details.
|LETTER TO EDITOR
|Year : 2019 | Volume
| Issue : 6 | Page : 1549-1550
Risk Homeostasis and Automation Bias in Neurosurgery
George C Vilanilam, Gopikrishnan Rajasekar
Department of Neurosurgery, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivandrum, Kerala, India
|Date of Web Publication||20-Dec-2019|
Dr. George C Vilanilam
Department of Neurosurgery, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Trivandrum - 695 011, Kerala
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Vilanilam GC, Rajasekar G. Risk Homeostasis and Automation Bias in Neurosurgery. Neurol India 2019;67:1549-50
”There will be fewer and fewer jobs that a robot cannot do better.”
A sense of awe with a tinge of worry was instilled in our minds by Ganpathy et al.'s review article  on the future impact of artificial intelligence (AI) in neurosciences. The fear of losing our jobs to artificially intelligent robots made us lose some sleep till we gathered our senses and decided to examine this prospective threat more objectively.
| » Futuristics|| |
Future science predictions are often based on current scientific trend analysis, pragmatic possibilities and intelligent guesswork. Ganapathy et al.'s futuristic thinking  states that humankind is in a transition phase with regard to the impact of AI in the neurosciences. However, it is still too early to predict if AI would have a positive impact on patient outcomes. Yet, the risk of the robot ousting the surgeon at his skilled occupation is a risk too significant to ignore. Stereotactic depth electrode implantation, deep brain stimulation and biopsies have already seen the advent of the neurosurgical robot. Narrow spaces and critical structures make neurosurgical robotic access more challenging than other surgical specialities.
| » Risk Homeostasis|| |
The Risk Homeostasis Theory (RHT) proposed by Gerald Wilde has an important role in risk reduction especially in complex multifactorial operations like neurosurgery. The concept was first proposed with reference to road traffic accidents, but it can be extrapolated to surgical adverse events as well. It states that the greater the level of perceived risk, road accidents decrease correspondingly. Thus, when road safety engineering measures are introduced, they may not translate into a lower accident rate, as drivers now tend to take greater risks.
Similarly, automated neurosurgery led by AI may also create a culture of throwing caution to the winds and a depreciation of the risk factors. An increased adverse event rate could be a natural fallout of the same. The recent case of a robotic cardiac surgical procedure gone awry is a clear case of misjudgment of the level of perceived risks and overreliance on automation.
| » Automation Bias|| |
Neurosurgical operations present a complex heterogeneity that calls for innovative and creative solutions. Over automation can lead to boredom, lack of innovation, complacency and above all more accidents. When there is too much trust placed on the infallibility of AI either in diagnosis or planning of a therapeutic intervention, there is always a risk of healthcare providers lowering their guard in every step of the process due to a decreased perceived risk. Neurosurgical tasks call for eternal vigilance, caution and calculated risks. The inclusion of AI and machine learning at each step should be deeply thought about before implementation as the standard of care. Caution and care most diligently exercised while dealing with surgical operations of the delicate brain should not lapse at all when surgical automation makes matters easier.
| » Uniquely Human|| |
Uniquely human traits have their strengths and weaknesses. Understanding language, nonverbal cues, complex image recognition, or using our body in highly flexible and adaptive ways are considered uniquely human. The physician's healing touch and empathetic words are still quite irreplaceable by any machine., Uniqely human limitations such as physiological tremors limiting surgical dexterity can easily be overcome by AI-guided robots. A rapid merging of man and machine via new types of interfaces such as augmented reality, virtual reality, brain–computer interfaces, artificial organs, and “smart” implants engineered with nanotechnology and synthetic biology may pave the way for less uniquely human traits in patient care. Nanobots in our bloodstream or communications implants in our brains have thinned the fine line between the human and the machine [Table 1].
| » Cobotics|| |
A complete elimination of the neurosurgeon -patient interaction by AI- seems improbable at present. However, a cobotic future (collaborative human–robot surgical effort) is quite exciting. Cobotics is a neologism formed by “collaboration” and “robotics” (Peshkin and Colgate, 1999). Robots chip in where human frailties exist. In surgical access precision planning and fine, repetitives technical tasks, human–robotic collaboration may be the way forward to overcome human limitations.
Thus, AI neurosurgical initiatives are an attempt to miniaturize access while enhancing speed, accuracy and control. A balance should be maintained while embracing automation and algorithms to surgical sciences, for optimum safety and effective outcomes. Neurosurgery in the current era may not appear 100% automatable but a prospective roboticized future with neurosurgeons on a permanent vacation may not be too distant. Nevertheless, such an automated future may not be as attractive in practice, as it sounds.
”Technology has no ethics, but humanity depends on them”
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| » References|| |
Ganapathy K, Abdul SS, Nursetyo AA. Artificial intelligence in neurosciences: A clinician's perspective. Neurol India 2018;66:934.
] [Full text]
Wilde JGS. Risk homeostasis theory and traffic accidents: Propositions, deductions and discussion of dissension in recent reactions. Ergonomics 1988;31:441-68.
Dyer C. Robot assisted surgery is blamed for heart patient's death. BMJ2018;363:k4791.
Vilanilam GC, Kumar KK, Aggrawal V, Sudhir BJ, Nair P, Easwer HV, et al
. Simulated neurosurgical reality: Could it fall short of the real thing? Neurol India 2016;64:1108.
] [Full text]
Iype B, Vilanilam GC. Patellar jerks in the 3-Tesla era: No knee-jerk excitement anymore!. Neurol India 2017;65:1445-7.
] [Full text]
Wang MY, Goto T, Tessitore E, Veeravagu A. Introduction. Robotics in neurosurgery. Neurosurg Focus 2017;42:E1.
Kochan A. A cobotic solution for surgical applications. Industrial Robot: An Int J 2004;31:478-80.