Hypervision Surgical, UK, has entered into a strategic development agreement with the Belgian digital technologies R&D hub imec to deliver scalable intraoperative imaging technologies. According to Hypervision, the partnership will combine “imec’s cutting-edge on-chip spectral sensing technology with Hypervision’s expertise in real-time, surgery-certified AI analytics.”

A vision for better intraoperative imaging

Hypervision was spun out from King’s College London, UK, in 2021 by a team of medical imaging and AI experts and clinicians with the goal of improving surgical decision-making with a new generation of intraoperative imaging technology. The team aimed to bring hyperspectral imaging—which can provide information on factors like tissue oxygenation, perfusion and differentiation—into the operating room by tackling challenges like hardware complexity, slow processing speeds and lack of compatibility with surgical workflows by developing an on-chip device that can integrate into existing surgical vision platforms.

In 2022, Hypervision’s technology was successfully used as part of an in vivo surgery study to guide a “keyhole” surgical procedure. Earlier this year, the Hypervision team demonstrated its technology for Her Royal Highness the Princess Royal during her official visit to King’s College, an opportunity the company says highlights, “the vital role of next-generation medical technologies in advancing surgical care and improving patient outcomes.”

The collaboration announcement comes on the heels of Hypervision’s obtaining US Food and Drug Administration (FDA) clearance and United Kingdom Conformity Assessed (UKCA) certification for its first hyperspectral vision platform, HYPERSNAP, which combines on-chip hyperspectral imaging with real-time AI analytics.

The technology and scalability 

HYPERSNAP is built on NVIDIA IGX, an industrial-grade edge AI platform. It works by processing a live data stream of the tissues being operated on, looking at tens of spectral bands across the visible spectrum. A challenge of using technology that requires data processing during surgery is there can’t be any lag time, or the surgeon’s precision and decision-making will be negatively affected. Hypervision says it achieved low-latency data processing (60 frames per second) by leveraging GPU parallelism, which can analyze multiple streams of data at the same time.

The device shows surgeons side-by-side videos of the tissue being imaged and a “heatmap” representing quantitative tissue perfusion from hyperspectral information of relative tissue oxygenation. The addition of the real-time mapping of tissue oxygenation provides surgeons with information that can benefit their decision-making while operating, which the company says improves patient outcomes.

By partnering with imec, Hypervision plans to combine its product with imec’s on-chip spectral sensing technology, with the end goal of scalability and the production of high-volume, clinically integrated products. “We are uniting deeptech nanotechnology with advanced AI-driven surgical intelligence... enabling a new era of precision-guided surgery and improved patient care,” says Olivier Rousseaux, director of venture development at imec.

As part of the collaboration, imec is leveraging its expertise in semiconductor fabrication, equipment and process technology to develop on-chip spectral imaging and to design and manufacture wafer-level interference-based optical filters. According to the company, imec’s “unique CMOS-based infrastructure provides very compact, clean and high-yield optical filter integration with scalability to high-volume production at low cost.”

“At Hypervision, we are driven by a vision to redefine what’s possible in the operating room,” said Michael Ebner, CEO and cofounder of Hypervision. “This strategic partnership with imec allows us to integrate groundbreaking chip-level spectral technology into scalable surgical solutions—accelerating our mission to bring advanced surgical intelligence to every procedure.”