Artificial intelligence technology offers a huge opportunity for diagnostics in medicine: with the right training, AI systems can quickly process large numbers of scans and images and identify problems with remarkable accuracy. But there is a problem: training the AI is time consuming and labor intensive. Take RedBrick AI, a US start-up, today announcing a $4.6 million funding round to accelerate its scale-up; its tools and technologies can make a huge difference, it believes.
“AI is remarkably effective in making diagnoses; for example, with AI you can automate 40% of breast cancer diagnoses,” explains RedBrick AI CEO and co-founder Shivam Sharma. “However, there is a real challenge: these systems are not easy to build, and healthcare in particular poses unique challenges.”
Simply put, to train an AI system, researchers need to show it as much data as possible – images and scans if your goal is to train it to read it. Each scan needs to be annotated to tell the system what it represents — an image of a cancer-free patient, perhaps, or an image with a potential problem area to investigate — so the AI can learn what it’s looking for.
The problem here, says Sharma, is that no one has developed tools to help clinicians quickly and easily annotate images so that large amounts of data can be quickly fed into the AI system. “Due to the complexity, size and unique nature of medical images, clinicians must resort to traditional and difficult-to-use clinical tools to perform annotations,” he explains.
In that regard, Redbrick AI’s unique selling point is that it has developed a suite of specialized annotation tools designed specifically for the healthcare profession. It believes that using its tools, clinicians and programmers can reduce the time it takes to train an AI system by as much as 60%.
This is an important breakthrough that opens up the possibility of accelerating the application of AI in healthcare. The medical profession is very open to such applications. In 2021 alone, the US Food and Drug Administration approved 115 AI algorithms for use in medical environments, an 83% increase from 2018, but there is room to go much further and faster.
Redbrick AI believes it improves on existing technology in a number of key areas. First, the tools are custom designed for the medical field, rather than relying on more generic techniques that don’t always reflect the nuances and specialties of the healthcare industry. In addition, the tools are quickly accessible through the platform and can be used without any prior training. The platform also includes a number of automation facilities, which can manage and speed up workflows.
It’s a value proposition that’s quickly gaining popularity in the healthcare industry, with customers from the US, Europe and Asia signing up during the company’s first year of trading. Redbrick AI offers its tools through a software-as-a-service model, where customers pay monthly subscriptions based on their user numbers to access the platform.
“With the rapid growth of AI in clinical settings, researchers need excellent tools to build high-quality datasets and models at scale,” Sharma added. “Our customers are at the forefront of this growth, pioneering everything from surgical robots to the automated detection of cancer.”
Today’s fundraising should help Redbrick AI reach even more of these customers over the next 12 months. Sharma expects to use some of the money raised to further develop the company’s tools. It has also freed up money for its go-to-market strategy, with Sharma seeing opportunities to work with larger numbers of corporate clients – the major medical research and technology companies – as well as smaller teams of healthcare specialists.
The $4.6 million seed round is led by Surge, Sequoia Capital India’s scale-up program, with participation from Y Combinator and a number of business angels.
Sharma and his co-founder Derek Lukacs are excited about the opportunity to scale the company faster. “In this room, everything begins and ends with the hospital,” says Sharma. “It’s the source of the raw data, but it’s also where our technology will ultimately have the most impact — for better patient outcomes.”