Author Name : Aradhya Pundir
While the concept of minimally invasive and robot-assisted surgery may sound intriguing, it is currently out of reach of the common man. India currently has over 180 to 200 da Vinci surgical robotic systems. The da Vinci surgical system equipment itself costs hospitals between ₹14 crore and ₹20 crore. It is evident from these numbers that robotic systems in the healthcare sector have not yet been able to bridge this socio-economic gap. RAS systems are largely dependent upon demographic and socioeconomic status. One way to bridge these disparities is by integrating AI systems with robotic systems. As of today, the physical robotic machine inside the operating theatre is a purely mechanical system with zero autonomous decision-making capabilities. Dr. Ajit Pai, a highly prominent Indian Senior Consultant Surgical Oncologist and Robotic Surgeon, refers to these systems as “slave robots” as they have no independent capacity for thought.
Integrating AI with these robots would mean incorporating technologies like Machine Learning, Computer Vision, Digital Twin simulation, and the usage of annotated surgical datasets for machine learning purposes. This would mean fusing kinematic data along with image data. Fusing kinematic data with image data in robotics combines internal joint and pose metrics with external visual feedback using tools like Extended Kalman Filters and visual servoing frameworks to improve state estimation, tracking precision, and spatial control during surgical procedures. This integration enables robotic systems to interpret both movement and visual information simultaneously, thereby making surgical assistance more accurate and reliable.
High-quality studies, including systematic reviews and clinical trials, consistently demonstrate that AI integration enhances surgical precision and patient outcomes. Dr. Ajit Pai supports this statement by taking the example of one of his patients. The patient was a 40-year-old man from Bangladesh who had come with right colon cancer. He had been offered open surgery locally. However, he was in great distress as the open surgery would have taken months of recovery, and he had to get back to work at the earliest as his family’s livelihood depended upon him. Dr. Ajit’s team walked the patient through the benefits of robotic surgery, which included smaller incisions, less blood loss, faster recovery, and an earlier return to normal function. Fortunately, the patient was discharged within 48 hours of surgery and was able to travel back comfortably. His reports revealed that the cancer was completely removed, and he did not require any further treatments like chemotherapy or radiation, which turned out to be a blessing for the patient as he was able to get back to work much sooner than originally assumed.
AI-enhanced robotic surgery not only offers the potential to democratize surgical care but also brings superior precision and efficiency if ethical, technical, and economic hurdles are addressed. On the technical front, integrating AI platforms with the existing robotic systems also raises concerns over interoperability, software compatibility, and cybersecurity. For instance, the massive cost and labour involved in expert medical labelling, along with corporate or institutional proprietary ownership, make surgical annotated datasets difficult to produce and access. Operation theatres are highly unpredictable environments, as bleeding, smoke, and shifting organs make standard computer vision tracking extremely difficult in real time. Every patient’s internal anatomy varies widely, making it hard for machine learning models to apply generalized training data safely and consistently. Additionally, purely mechanical and computerized scaling provides immediate, fail-safe feedback loops that complex AI processing can sometimes delay, posing another important technical challenge.
In conclusion, robotic-assisted surgeries cannot be democratized unless they move from purely mechanical and computerized operation to AI-assisted robotic surgeries, provided the technical hurdles of AI-assisted RAS are successfully overcome.
