The advent of artificial intelligence in medicine is pushing boundaries, with a recent study showing how AI techniques, similar to those used for ChatGPT, are being employed to train surgical robots to operate independently. Researchers at Johns Hopkins University and Stanford University have created a model where robots learn from videos of human-controlled robotic arms executing surgical tasks, potentially simplifying the programming required for each surgical maneuver.
This evolution from assistive to potentially autonomous robotics in surgery builds on years of technology where robots have been used to enhance precision in surgical procedures, famously demonstrated by the “surgery on a grape” meme. By 2020, robot-assisted surgeries were already common, with over 876,000 procedures recorded. Robots offer benefits like reaching inaccessible areas and providing tremor-free precision, but traditionally, they’ve been under the direct control of a surgeon.
However, the shift towards more autonomous surgical robots brings significant concerns. Critics argue that AI, like that in ChatGPT, merely replicates observed data without a deep understanding of underlying medical concepts. This raises the alarm: what happens when an AI encounters a scenario it has never been trained for? Surgery often involves split-second decisions where mistakes can be catastrophic. Furthermore, for autonomous robots to be legally used in surgeries, they would need to pass stringent FDA regulations, unlike AI tools for summarizing patient visits, which do not require such oversight because a doctor’s review is technically mandatory.
The potential for error in AI recommendations or autonomous actions is not trivial. There’s evidence that AI can “hallucinate” or provide incorrect data, which might not be thoroughly reviewed by overburdened doctors. This scenario eerily mirrors military applications where AI’s predictions were taken at face value, leading to tragic outcomes due to human complacency.
In healthcare, where mistakes can mean life or death, the stakes are incredibly high. Dr. Dipen J. Parekh from the University of Miami highlights the complexity, noting that each patient’s anatomy and disease presentation is unique. An autonomous robot would need to interpret complex medical imaging like CT scans or MRIs, a task currently managed by highly trained professionals.
The notion of infallible AI in surgery seems far-fetched given technology’s inherent limitations. The aftermath of a surgical error by an AI could be monumental, raising questions about liability and accountability. Who would be responsible if an autonomous robot errs? Doctors have the training and accountability that AI lacks, alongside a moral compass that machines do not possess. There’s also the risk that reliance on robots might lead to a decline in human surgical skills, akin to how dating apps might diminish social interaction skills.
The argument for autonomous surgery often cites the need to alleviate overworked doctors, but this might be addressing symptoms rather than the root causes of healthcare system issues. The looming doctor shortage, expected to reach up to 20,000 surgeons by 2036 in the U.S., underscores the need to tackle systemic problems like the inaccessibility of medical education and practice rather than just substituting with technology.
As we venture into this new era of medical robotics, it’s crucial to balance technological innovation with ethical considerations, ensuring that human oversight, expertise, and accountability remain central to the practice of medicine.