Nvidia backs Flywheel to accelerate medical AI development

Hours after teaming up with data cloud provider Snowflake, Nvidia has backed a vendor working to overcome data hurdles associated with medical AI development. NVentures, the venture capital arm of the computing giant, has co-led Minneapolis-based Flywheel’s $54 million series D round with Novalis LifeSciences.

Flywheel said it plans to use the capital to further grow its footprint and help healthcare and medical research organizations focus more on their AI projects, instead of preparing the data for them. Microsoft also participated in the round, along with insiders such as Invenshure, 8VC, Beringea and Hewlett Packard Enterprise.

The investment comes at a time when enterprises across all sectors are exploring the potential of AI to drive efficiencies and save time. In healthcare, this technology can even help with improving patient care and defining appropriate treatment interventions.

So, how exactly does Flywheel help?

Organizations have to put their data in order when they want to build and train an AI model. This means paying close attention to the data sources and aggressively organizing, curating and cleaning the information to make it training-ready. The task sounds easy but can easily take up to 80% of data scientists’ time, leaving just 20% for the main job of analysis and creating insights.

Founded in 2012, Flywheel addresses this gap with a special focus on healthcare and medical research. The company offers an end-to-end platform that automates ingestion, curation, management and computation of medical imaging data, allowing healthcare organizations to efficiently scale AI development and other research discoveries.

“Flywheel allows providers to extract and mobilize valuable insights from data by making it more structured, accessible and conducive to integrated applications, including AI,” Jim Olson, the CEO of the company, told VentureBeat.

Three key technologies enable this system: Flywheel Enterprise to manage imaging data, Flywheel Discovery to explore and analyze this data and Flywheel Exchange to browse and use pre-curated public datasets.

While Flywheel has not shared how many customers it has, its website does mention some important success stories, including those from Roche, Genentech, the University of California, San Francisco, CMU-Pitt Bridge Center and Octave Bioscience. Among use cases, the company is also building a federated data platform for the development of an unbiased breast imaging AI. 

“We don’t have a lot of direct competitors that are able to offer the complete end-to-end solution focused on medical imaging like we are. Oftentimes, our biggest competitor is a homegrown solution that has been built internally but reaches a point where it can no longer scale or integrate with other applications like ours can,” Olson added.

Plan ahead

With this round of funding, which takes Flywheel’s total capital raised to over $150 million, the company plans to expand beyond the public sector and pharmaceutical companies to cover more healthcare segments seeking to harness the value of their data for AI development. These include healthcare providers, payers, and software companies.

Flywheel also said it will focus on growing its presence in key geographies, particularly across Europe. 

“The application of AI has led to the discovery of new drugs, identifying patterns in disease, and
improvements in patient care. Flywheel uses AI to unlock the value within medical imaging data, signaling the continued benefit of applying AI across the healthcare industry,” Mohamed “Sid” Siddeek, head of NVentures, said.

In the near future, tools like Flywheel will become critical to keeping data affairs in order and compliant when developing AI for healthcare. According to a recent survey from GE HealthCare, 61% of clinicians believe the technology can help with decision-making, while nearly 55% said it can enable faster health interventions and improve operational efficiencies.

Originally appeared on: TheSpuzz