Dartmouth Startup Wins Phase II SBIR Grant to Develop Next-Generation Surgical Navigation Systems

IST is developing navigation systems for surgeons with a combination of real-time image updating and quantitative fluorescence imaging to enable faster, less invasive, and more accurate surgical treatment of tumors.

"We're starting with the brain, but it can apply to more than that," says Dartmouth's MacLean Professor of Engineering Keith Paulsen, one of IST's founders. "The overall goal is to guide surgeons to do a better job. A lot of times that involves oncology, and trying to get all the cancer out that you can."

Currently, IST is based in Lebanon, NH at the Dartmouth Regional Technology Center near Dartmouth Hitchcock Medical Center. The founders, including Professor of Surgery and Neurology David Roberts, are part of a long-standing NIH-funded collaboration between researchers at Dartmouth, as well as Brian Wilson, a fluorescence imaging expert from the University Health Network of Toronto.

"There's pre-operative image guidance, and then there's fluorescence," explains Paulsen. "You want those two forms of surgical navigation to work together as an integrated system, and nobody's doing that."

For image guidance, preoperative patient scans are taken and oriented with the operating room. A GPS-like system then tracks the tools and equipment and provides the surgical team with a map, but during surgery the patient's tissues often shift, and that's been an ongoing problem since the advent of surgical image guidance.

"It's like you're driving around and all of a sudden a detour pops up. We've all been there, you're looking at the map and you're looking at the detour, trying to navigate between the two. We want surgeons to get an updated map while they work so they don't have to guess where to go," Paulsen says.

The Dartmouth team has been working on that navigation problem for many years, and they had developed a solution, but the computing power needed to execute it fast enough for the surgeon to use in real-time has only more recently become available. "Right now, our system can update the map in just a few minutes, with the help of an operator," says Paulsen.

The next step is to make it completely autonomous and ultimately continuous. IST studied how researchers processed intraoperative images and worked to translate those decisions into algorithms. The result is an autonomous, on-demand workflow in which the surgeon requests an update, the system collects the required data, and produces updated guidance.

Fluorescence imaging, on the other hand, is about localization at the microscopic level to identify tumor infiltration. "The image guidance gets you to the right spot, but once you're there, a tumor often doesn't look any different than normal brain. Especially at the edges, it's really hard to tell it apart," Paulsen says.

Fluorescence imaging works because tumors are metabolically more active than normal tissue, so they'll take up or produce more of certain agents or molecules, in this case a fluorophore which is a fluorescent chemical compound that emits light upon excitation.

"When you get down to really small amounts of the fluorophore, you can't see it, but you can detect it," he says. "So we have a point probe that takes that measurement and can detect the smallest amount of chemical agent. And then we have what we call a wide-field imaging system, which shows the whole surgical field with an overlaid map of the fluorescence."

Paulsen and his team expect the technology will make surgery more complete and more accurate for more cases. And it could make high-quality surgery less dependent on the skill of the surgeon. "We can make better surgeons out of more surgeons. It can be a democratizer of surgeon skill in that sense," says Paulsen.

Ultimately, the goal is to extend both life and quality of life for patients. And even if a tumor isn't perfectly removed, often just getting a higher percentage of the cancer cells out can give follow-up treatments, such as immunotherapies, a much better chance of working.

"We've been doing this work for a while, and had developed a lot of technology," says Paulsen. "We finally realized, if we want to see the work have any impact, we were going to have to take on the commercialization challenges. So that's our plan."