5 days ago
The scalpel's new partner: When AI surgeons step into the operating room
Recently, at a laboratory at Johns Hopkins University, a machine rewrote medical history. A pair of robotic arms, guided not by a surgeon's steady hand but its own artificial intelligence, removed a gallbladder entirely on its own. The robot – Hierarchical Surgical Robot Transformer (SRT-H), identified structures, applied clips, made precise incisions, and sutured the wound with precision. Though performed on a pig cadaver, this was hailed as "the first realistic surgery by a machine with almost no human intervention," it shattered assumptions about what machines could understand in the chaotic, fluid world of biological bodies.
This breakthrough is the next step in the 'evolution' of the remote-controlled robots like the da Vinci system, where surgeons operate instruments from a console. The SRT-H, though, is a landmark simply because here it is a machine that interprets visual data, decides actions, and self-corrects errors in real-time. Its intelligence comes from a large language model like ChatGPT, its expertise is learned from watching and absorbing 17 hours of surgical footage where human experts performed the same gallbladder removals. It internalised 16,000 individual motions, learning the dance of dissection, clipping, and extraction. When tested, it succeeded flawlessly eight times, adapting to obscured views, synthetic blood, and shifted starting positions. Most remarkably, it detected and fixed its own mistakes, like a gripper slipping off an artery, without human prompting.
I scouted around for the best person to talk about this, and all fingers pointed to Dr. Rajiv Santosham, a pioneering minimally invasive and robotic thoracic surgeon at Apollo Hospitals, Chennai. So, I talked to him about the same. "We never imagined we could fly. So robots performing fully independent surgery feels unimaginable right now. But what they can do is already transforming how we operate. Imagine navigating near a critical vessel. I, as a surgeon, operate blind to what lies immediately behind it. But an AI, fed the patient's pre-op CT scan, can visualise that hidden anatomy in real-time. It could warn me, guide me, prevent a tear. That's not replacement; that's revolutionary assistance."
The Surgeon's Perspective – Pragmatism Meets Potential: Dr. Santosham's voice carries the weight of experience as a pioneer of uniportal VATS (video-assisted thoracic surgery) in India – a technique requiring a single 4-centimetre incision. Yet, his excitement about autonomous AI is measured, grounded in daily realities, and acknowledges AI's current limitations. "To test its medical judgment," he shares, "I once fed ChatGPT an ECG image. The diagnosis it gave was completely wrong. I asked it to re-check, double-check, triple-check. It still made a mess of it. So yes, there's a vast chasm between pattern recognition and true clinical understanding. It's a tool, a powerful one, but still evolving."
His speciality, thoracic surgery, also tempers his enthusiasm for near-term autonomy. "I do robotic surgery. But frankly, for many lung procedures, they can be a bit… fancy. My uniportal technique often requires smaller access points than robotic arms. Why make three or four larger holes when one small one suffices? For straightforward cases, the robot might actually add complexity, not value."
However, he lights up when discussing the specific applications of the tech. "There are procedures where robotic precision combined with AI's spatial awareness could be transformative. Like in urology, gynaecological oncology, and deep pelvic cancers. Places human hands and eyes struggle to reach and see clearly." Crucially, he sees robotics as a democratizing force. "The beauty isn't just precision; it's consistency. Outcomes become less dependent on the individual surgeon's experience or fatigue level. A well-trained machine also reduces the learning curve for a surgeon. Suddenly, complex procedures become safer and more accessible to a wider pool of surgeons, especially in settings with limited specialist access."
Despite this, his conclusion on the human role is definitive: "Surgeons won't lose their jobs to machines anytime soon. But surgeons who actively learn to harness AI? They will become the leaders, the innovators. They'll have an undeniable edge over those clinging purely to conventional methods. Hence, adaptation isn't optional; it's the future."
Beyond the Lab – Cost, Blame, and the Indian Context: The promise of autonomous surgery collides with practical hurdles: affordability, accountability, and adoption. Dr. Santosham, deeply familiar with India's healthcare landscape, offers a unique perspective on the scepticism about such expensive medical tech reaching the masses.
"Remember laparoscopy?" Dr. Santosham asks. "They said it would never reach tier 2 and tier 3 cities in India. It did. Then they said robotic surgery was too expensive, destined only for elite metros. It's now percolating across the country. The Da Vinci system is brilliant, but it's not the only player in the market. China is producing high-quality robotic systems at phenomenally lower costs. India absolutely has the capability to innovate and manufacture affordable versions too. So, bridging this divide will happen. It's a matter of when, not if." He envisions a future where indigenous robotics makes precision surgery accessible far beyond the Metros.
The Accountability Question: We all know AI is prone to errors, and worse – hallucinations. So, when an autonomous robot errs, whose responsibility would that be? Dr. Santosham addresses this head-on, drawing a sharp distinction from the rapid, crisis-driven deployment of technologies like COVID-19 vaccines. "Medical technology, especially something as critical as autonomous surgery, undergoes incredibly rigorous testing before approval," he states. "Think of the scrutiny applied in the US by the FDA. By the time a system is cleared for human use, the chances of catastrophic error are minimised. And let's be clear: human surgeons make mistakes too. Perfection is a myth, whether flesh or silicon. The key is robust failsafes." He draws parallels to existing safety features in current robotic systems: "If I glance away during a robotic suture, the system detects my diverted attention and freezes all instrument movement instantly. They filter out hand tremors. These are designed to prevent human error. Autonomous systems will build layers upon layers of such safeguards."
Regulation and Indian Adoption: While looking towards US and EU regulatory benchmarks, Dr. Santosham is bullish on India's embrace of the technology. "India has a remarkable capacity for technological leapfrogging," he asserts. "We often focus on the challenges of poverty, but underestimate the sheer scale of wealth and technological ambition in our major cities. Chennai, Hyderabad, Bangalore, Mumbai, Delhi—these are hubs with world-class hospitals and patients demanding the latest innovations. They will invest in, adopt, and eventually even build and export advanced autonomous surgical systems. Affordability, driven by local innovation and scale, will follow."
The Road Ahead – Collaboration, Not Conquest: The vision emerging from labs like Johns Hopkins and the insights of surgeons like Dr. Santosham point not to a dystopian replacement of humans, but to a powerful, evolving partnership.
The SRT-H robot wasn't designed for isolation. During its successful trials, human surgeons remained present, offering verbal guidance: "Move the left arm slightly," or "Switch to the curved scissors." The AI understood and complied. This interaction is the blueprint – autonomy that enhances human oversight, not eliminate it. The immediate future hence lies in collaborative autonomy, where AI handles predictable, precision-critical tasks under a surgeon's supervisory command, freeing the human expert to manage the overall strategy, complex decision-making, and unexpected complications.
The statistics supporting this hybrid approach are compelling. Meta-analyses of existing robot-assisted surgery (still human-controlled) already show tangible benefits: operations completed 25% faster, a 30% reduction in complications during surgery, and patients recovering 15% quicker. Autonomous systems, once matured, promise to amplify these gains while tackling the persistent shortage of highly skilled surgeons, particularly in specialised fields and underserved regions.
The final goal transcends mere technical achievements. It's about democratisation of medical tech. As Dr. Santosham implies, it's about ensuring that a child in a remote village doesn't face a life-threatening condition simply because an experienced surgeon isn't at hand. It's about making the collective genius of global surgical expertise accessible through intelligent machines, guided by local medical professionals. The autonomous incision at Johns Hopkins wasn't just into tissue; it was the first cut into a future where the best possible surgery isn't a privilege of geography or wealth, but distributed equally for all.
The journey will demand rigorous validation, ethical frameworks, cultural acceptance, and continued human ingenuity. But as Dr. Santosham concludes with characteristic pragmatism and foresight: "India embraced robots; it'll embrace autonomy too. We'll afford it, build it, master it. The surgeon's role will evolve, but the need for human judgment, compassion, and responsibility? That remains eternal."
The scalpel has a new partner, and together, they are rewriting the rules of healing. One successfully operated body at a time.
