Theator raises $15.5 million to analyze surgical footage with computer vision

Surgical platform developer Theator today closed a $15.5 million series A round led by Insight Partners, bringing the company’s total raised to date to over $18 million. Theator says the funds will be used to scale its commercial operations and partnerships with U.S. providers as it looks to grow its R&D team.

Some experts assert that surgeons’ efforts to overcome inequalities are stunted by an apprenticeship model that overemphasizes scope of practice, limiting surgeons’ knowledge. In the U.S., Black children are three times as likely as white children to die within a month of surgery, according to a study published in the journal Pediatrics. An estimated 5 billion people lack access to safe surgical care worldwide. And of the at least 7 million people who suffer complications following surgery each year, including more than 1 million who die, at least 50% of these deaths and complications are preventable.

Theator, a two-year-old company based in San Mateo, California, claims its product addresses these challenges by enabling surgeons to view video recordings of their procedures as lists (e.g., “Decompression, Cystic duct leak, Critical view of safety”) and jump to steps from which they wish to learn. Users get annotated summaries of their surgical performance as well as KPIs, analytics, and rankings to identify where skills training might be needed.

Theator was founded in 2018, but the company’s backstory dates back to 2015, when CEO and founder Tamir Wolf’s wife and then-boss both suffered from appendicitis a few months apart. As a physician, Wolf was able to diagnose their condition and took them to the emergency room for treatment. “They were treated at two different hospitals, and I was struck by the wildly different levels of care they experienced at each one, even though they were a mere 7 miles apart,” he told VentureBeat via email. “As I tried to understand why the approach to a seemingly straightforward illness was so different, I began to unearth the widespread disparity and variability that is plaguing the medical field, and surgery specifically.”

The use of AI and machine learning in the operating theater is on the rise. In 2019, McGill University developed an algorithm that was able to tell when trainee surgeons were ready to perform a real operation. And last April, researchers from Stanford created a model to help protect health care professionals from COVID-19 in the operating room.

As for Theator, it leverages computer vision to scan video footage of real-world procedures and identify key moments in order to annotate them with metatags. Through desktop and mobile apps, surgeons gain access to an indexed library of over 400,000 minutes of surgical video encompassing over 80,000 “intraoperative” moments.

“[We use] AI to complete high-granularity indexing of large scale surgical datasets, which enables big data analysis of surgical patterns for the first time, introducing a scientific data-driven understanding of best practices,” Wolf explained. “At Theator, we’re using computer vision, on edge, to detect any frames that might have identifying features (everything that is extra-cavitary), removing these before uploading to our cloud-based system … Theator’s research team [also] recently developed VTN (Video Transfer Network), a first of its kind Transformer-based framework for video action recognition that yields a significantly reduced training runtime (x16.1) and an accelerated inference process (x5.1) while maintaining state-of-the-art performance. VTN has the potential to significantly accelerate the video input and review component of Theator’s platform, in turn improving our ability to roll it out at scale in hospitals worldwide.”

Other investors in Theator’s latest round of fundraising include 23andMe cofounder and CEO Anne Wojcicki, former Netflix chief product officer Neil Hunt, and Zebra Medical Vision’s cofounder Eyal Gura.

“The lack of in-person contact with experienced surgeons has accelerated the need for Theator’s solution, which broadens surgeon expertise, even remotely or asynchronously. The pandemic has highlighted a reality wherein surgeons barely have time for in-person interaction, allowing them a greater degree of freedom to continuously improve performance,” Wolf said. “Surgeons also used to  learn new techniques and approaches at conferences, but these have all but been removed from the equation. Theator’s comprehensive surgical video dataset, the ability to upload your own procedures, view others, and receive and provide feedback has made it an increasingly vital tool when more and more continuous development is being carried out remotely.”

By VentureBeat Source Link