Surgeons traditionally rely on approximate estimations when determining distances and anatomical structures during operations, which can lead to complications. In 2023, Dr. Vinayak Rengan, a practicing pediatric and neonatal surgeon, joined forces with Balachandran Sitharaman to solve this problem. They founded the startup anatomy, based in Bengaluru, which develops a surgical intelligence platform to help surgeons estimate distances.
Dr. Rengan noted that the problem is relevant because approximately 300 million operations are performed annually, 50 million of which encounter complications. The company's goal is to prevent the recurrence of these complications. Curium's first product, named SurgiMeasure, uses deep learning and computer vision to provide real-time anatomical measurements during laparoscopic and robotic procedures. The nine-member startup aims to create a fundamental model for surgical applications rather than focusing on a single-task AI tool.
According to Dr. Rengan, this issue is particularly significant for minimally invasive surgery. Laparoscopic cameras generate two-dimensional images, making it difficult for surgeons to determine depth and distance without physical reference. He reported that the average measurement error among surgeons using rulers, threads, needles, and calipers is currently about 30%.
He compares the Curium platform to a navigation system for the operating room, stating that it is like 'Google Maps or a parking assistant for the abdomen.' The system guides surgeons during laparoscopic and robotic surgery by indicating where to cut, what to avoid, where certain structures might move, and by how much.
SurgiMeasure integrates deep learning, computer vision, anatomical segmentation, and depth estimation to help surgeons understand spatial relationships during interventions. According to Dr. Rengan, the system has been implemented in real operating rooms and demonstrated 98% accuracy in tool segmentation, real-time output at 30 frames per second, and an end-to-end latency of less than 300 milliseconds. Testing was conducted in eight live surgeries at Dr. Mehta's hospitals in Chennai.
Curium is also working on technologies for applications beyond simple measurement. The current roadmap includes eight types of operations in hepatobiliary, gynecological, urological, thoracic, abdominal, colorectal, bariatric, and hernia repair fields. The company's long-term plan is to develop an AI system capable of assisting surgeons in various procedures, not being limited to one function.
The B2B startup plans to offer its software bundled with laparoscopic and robotic platforms to hospitals and surgical centers via subscription. Curium has also won the Johnson & Johnson Polyphonic Surgical AI Challenge 2026 and is building a network of surgeons and clinical partners. The startup is entering a relatively new market: the surgical robotics market in India was estimated at around $851 million in 2023, and Dr. Rengan notes that only a few companies globally focus on real-time surgical AI.