How the Polyphonic™ AI Fund for Surgery QuickFire Challenge awardees are aiming to empower surgeons

For years, much of what was learned during surgery stayed within the walls of a single operating room (OR). That means valuable information—from surgical video to device data and clinical insights—hasn’t always been easy to share or use to improve patient care.

But what if these separate pieces could be connected, creating a collective surgical intelligence that can inform standards and future procedures, as well as decrease variability?

That’s the vision behind Johnson & Johnson’s Polyphonic™ digital ecosystem for surgery and the company’s targeted investments in artificial intelligence (AI), says Raymond Fryrear II, M.D., Vice President and Chief Scientific Officer of Polyphonic at Johnson & Johnson.

Polyphonic is an open digital ecosystem that connects the OR like never before—bringing together surgical data, video, devices and AI insights into a single surgical intelligence layer so that healthcare teams can collaborate more effectively. Its open approach allows a broad range of partners, applications and innovations to work together, helping break down longstanding silos and enabling insights to flow across the surgical journey. By connecting data, people and tools, it aims to help transform how surgery is organized, performed, taught and continuously improved.

“When you’re a surgeon, it can feel like you are trying to conduct an orchestra while the sheet music keeps changing and the instruments are all tuned differently,” says Dr. Fryrear. “What we’re trying to do with Polyphonic is bring the orchestra into harmony—not replacing surgeon judgment but augmenting it, turning every case into shared wisdom in a standardized manner, thereby reducing cognitive load, decreasing variability and improving patient outcomes at scale.”

That is why the Polyphonic AI Fund for Surgery QuickFire Challenge (QFC) exists.

The QFC is a global initiative that invites researchers, startups, developers, healthcare institutions and established companies to submit AI-powered proposals that address real clinical problems and real-world scale.

Selected by a panel of experts, awardees receive funding, mentorship and technical support to help accelerate development. While innovations aren’t required to be built on Polyphonic, the QFC is intended to foster solutions that can benefit from—and potentially integrate with—the Polyphonic open digital ecosystem, helping bring AI innovations into ORs around the world.

“The QFC matters because many great ideas in the world languish in the gap between cool demos and those that actually change practice in the OR,” says Dr. Fryrear. “Polyphonic is closing that gap, accelerating the ideas that have the potential to make a real impact and connecting the parts of surgery to elevate the entire field at scale.”

Read on to meet this year’s QFC awardees, explore the clinical problems they’re tackling and learn how they’re harnessing the power of AI to help shape the future of surgery.

Curium Life’s SurgiMeasure

“Like a parking assistant for surgeons, the tool emits a warning sign when a surgeon is too close to another part of the anatomy.”

After 10 years working as a pediatric and neonatal surgeon, Vinayak S. Rengan, M.D., found himself frustrated by how much guesswork was still involved in procedures. Recently, he was performing surgery on a baby with an undescended testicle and needed to measure the distance from the testicle to the opening in the lower abdominal wall through which it normally descends before birth.

“Because I’ve done a couple hundred of these procedures, I can reasonably guesstimate this measurement,” says Dr. Rengan. “If a surgeon is less experienced, however, they may not be able to do that guesstimation with as much accuracy.”

That’s why Dr. Rengan built a real-time anatomical measurement tool called SurgiMeasure. It uses AI—specifically, a combination of computer vision, deep learning and depth algorithms—to convert surgical video into structured geometric data, measuring anatomical distances during laparoscopic and robotic surgeries. This is crucial, as anatomy differs slightly from person to person, and the anatomical environment itself can be difficult to discern.

Consider a common abdominal surgery such as hysterectomy (the removal of the uterus). The ureter is a thin tube—about the thickness of a phone charger cable—that connects the kidney to the bladder and sits just behind the uterus. During the surgery, a surgeon’s instruments can get very close to the ureter, and even the most highly skilled surgeons are inherently at risk of nicking it and causing ureteral injury, says Dr. Rengan.

“What we are building is a foundational model that’s able to identify anatomy at a granular level,” he says. “The vision is that when the surgeon is about to get too close to the ureter, a warning signal occurs—sort of like a glorified parking assist, where it starts beeping when you get too close to another car.”

The idea for SurgiMeasure came to Dr. Rengan in 2016, when he was a surgical resident and had the task of deciding which surgeon would handle each operation.

“I quickly realized that some surgeons are much better than others—irrespective of how long they’ve trained—at identifying complications before they occur,” he says. “I thought, ‘What if we were able to bring in that kind of experience for every procedure in every operating room?’”

This is what pushed Dr. Rengan and a small team of fellow surgeons to launch Curium Life, a deep-tech surgical AI startup that is developing multiple products, in addition to SurgiMeasure, to build applications that transform surgery as we know it today to improve outcomes for both surgeons and patients alike.

“Fifteen years ago, we didn’t have parking assist in cars and now most of us can’t live without it,” Dr. Rengan says. “That’s the direction we’re building toward.”

C1DO1 experiential learning platform

“An AI-powered research assistant that watches surgical procedures and gives feedback in both real-time and post-surgery.”

As a professor of surgery and the head of surgical simulations at Catholic University of Chile, Julián Varas, M.D., began studying how specialized training could accelerate surgical skill. His research showed that after just 16 training sessions packed with detailed feedback and targeted mentoring, new surgeons performed as well as traditionally trained surgeons with five years’ experience. These impressive results spread via word of mouth to surgeons around the world, who started traveling to Dr. Varas’ training facility in Chile.

This inspired Dr. Varas to consider how he might offer these specialized training sessions remotely—a tall task given the kind of hands-on practice and feedback that’s historically been needed to optimize surgical training.

First, he created a digital interface between teachers and trainees that was focused on improving and enhancing surgical feedback. When Large Language Models (LLMs) came out, Dr. Varas fed LLMs all the info he’d gathered about what constitutes good feedback for surgeons. He also leveraged computer vision—a branch of AI that enables computers to interpret and analyze visual information from images and videos, among other inputs—to assess surgical training videos.

In 2019, C1DO1 (pronounced “see one, do one”) was launched. The name was inspired by the surgeon William Stewart Halsted, who is widely credited with creating the first formal surgical residency program in the U.S. in the late 1800s.

“Halsted used to say, ‘You have to see me, then I will supervise you, then you will perform the surgery on your own and finally you will return home and teach others,’” says Dr. Varas.

Here’s how the C1DO1 platform works: Surgeons from any specialty submit videos of themselves performing real or simulated procedures, which are then reviewed independently by human experts as well as AI. They then receive materials such as video tutorials, procedure checklists and personalized feedback on areas of improvement.

C1DO1 also accumulates the information clinicians upload, so the technology can continually assess a surgeon’s skills and gauge their improvement. This significantly reduces a surgeon’s learning curve, says Dr. Varas, who explains that while it can typically take hundreds of cases to master a procedure, C1DO1 is helping surgeons reach a similar level of mastery after 10 to 20 cases.

Dr. Varas is excited to see how tapping into Johnson & Johnson’s Polyphonic ecosystem will lead to more and even better information to help the C1DO1 system provide more personalized and detailed assessments and feedback over time—not only for surgeons, but for all healthcare professionals.

“Right now, we’re helping surgeons decrease the number of cases they need to do to become experts,” says Dr. Varas. “If we can do this with surgeons, why not with nurses, physical therapists and any healthcare profession?”

iQ Surgical’s AI copilot for the operating room

“Like a wayfinder for surgeons, helping augment their understanding, progress and execution during surgery.”

When Jay Park, M.D., was in his early surgical training, he was like any other resident: eager to see as many cases as possible, assist with as many surgeries as he could and work with the best mentors. This wasn’t just because Dr. Park has always been an enthusiastic learner.

“It’s also because every day of medical school and beyond, you’re building a database in your head to help you judge what is right or wrong when you’re operating,” he says.

This kind of judgment is crucial when operating on the brain—arguably the body’s most complex organ—where anatomy can shift significantly during surgery, a phenomenon neurosurgeons call brain shift. “The brain has a lot of soft tissue, which can be impacted by things like gravity, fluid changes and brain swelling,” explains Dr. Park.

Today, neurosurgeons utilize a technology called neuronavigation, which uses imaging taken before a procedure to give the neurosurgeon a map of the patient’s brain. Neuronavigation is incredibly helpful, improving a surgeon’s precision when removing tumors, for example, and protecting healthy brain tissue.

“Yet because of brain shift, that image taken before the procedure and what the surgeon is actually looking at throughout the procedure are often two very different things,” says Dr. Park. “It’s kind of like using Google Street View instead of a real-time GPS.”

The other tool neurosurgeons rely on during procedures is intraoperative MRI. The name is a bit of a misnomer, however, as the MRI isn’t actually in the OR—it’s usually in a nearby room—and it can take an hour or two to get the images. Considering some brain surgeries require two or even three images, that means a lot of time is spent trying to capture what’s happening in a patient’s brain, explains Dr. Park.

Enter iQ Surgical’s copilot—continuously updated to maintain a more current understanding.

The technology is being developed with the aim of providing surgeons a set of detailed data—a more objective and consistent layer of information that could complement what an individual surgeon can accumulate over time through intuition, experience and pattern recognition—to potentially support decision-making during surgery.

“As surgeons, we don’t currently have the technological equivalents to a real-time GPS, a dashboard, lane assists or even a speedometer,” says Dr. Park. “We have to develop an intrinsic ability to know where we are and how fast we’re going.”

With iQ Surgical’s copilot, Dr. Park aims to give every surgeon “a real-time layer of surgical intelligence during complex procedures,” he says. The goal is not to replace surgical judgment, but to augment it by helping surgeons make more precise, consistent and data-informed decisions in the OR.

Polyphonic’s mission to bring information together can enable the company to make the next generation of AI-powered products for surgeons a reality, says Dr. Park.

What iQ Surgical is building is about giving surgeons better intelligence at the moment they need it most, he adds.