Flywheel has been helping 10 of the top 20 biopharma companies accelerate their imaging data workflows and add value to their data. But what does that mean in practice? The answer is that it can really vary depending on each organization’s needs and priorities. Below, we’ll detail what some of our biopharma customers are doing with Flywheel to maximize their imaging data.
Saving On Development Costs and Increasing Flexibility
One top 10 pharmaceutical company needed a way to interrogate imaging data earlier in the clinical trial process so they could make better go/no-go decisions and streamline their oncology research. They had been working for three years to develop an in-house imaging data platform to get around clinical trial data bottlenecks. But development costs soon spiraled out of control, and any extra customization plus maintenance would have pushed these costs up further.
The company decided to work with Flywheel, and we were able to drive those development costs down by 33%. We also provide easier customization via software development kits (SDKs), which are included with the platform. Plus, our team of software engineers is available to discuss more complex customization, meaning the company doesn’t have to do it all in-house or find another vendor to do any development work for them.
Our imaging data platform has helped the company’s efforts on several fronts:
- More informed decision-making around trials
- Exploration of secondary endpoints
- Establishing baseline assessments for AI and radiomic analyses
Learn more about our work with this client in our case study.
Powering AI Development Efforts
Another biopharma company’s AI development efforts were being significantly slowed down by data delays. Their team was shipping raw imaging data on physical drives to reading centers, and those readers would then perform annotations on their own platforms and have to send data back months later. When medical AI is being developed and implemented at a breakneck pace, waiting around for months wouldn’t cut it.
Working with Flywheel helped them save an average of three months per annotation campaign by allowing reading centers to work on the same platform as the company, eliminating the need for data shipping while maintaining data privacy and blinding. They’re able to use this advantage to create high-quality ground truth annotations for Optical Coherence Tomography (OCT) images that can train and test machine-learning algorithms to advance their ophthalmology R&D. With research from seven clinical trials ingested at the outset, they’re just getting started.
Read more about this company’s AI development efforts with Flywheel in our case study.
Creating a FAIR Data Repository
Clinical research yields scores of data, but too much data can present a problem. A leading biopharma company had images from years of trials spanning oncology, ophthalmology and neurology. This data was scattered across disconnected systems with different naming conventions and formats, so their teams struggled to find the data they already had and use it efficiently.
Implementing Flywheel helped the company make their data FAIR (findable, accessible, interoperable and reusable) by sorting and processing all of their data in one place. Part of this process involves using Flywheel Gears, containerized algorithms that automate tasks like file conversion, QA and QC. By running Gears, they have saved countless hours standardizing datasets including MRI, CT, PET, OCT and Fundus imaging.
The company has now processed 200 million images and counting. And it has successfully developed a new algorithm that was cleared by the FDA. Learn more about their story in our case study.
How Other Biopharma Companies are Leveraging Flywheel
Other biopharma companies are using Flywheel at various stages of the imaging journey, from preclinical through FDA submittal.
Data Orchestration
A top 20 biopharma company uses Flywheel as the orchestration layer for all preclinical imaging from ingest through radiomics and AI inference.
AI-Driven Analysis
A top 20 biopharma company imports third party algorithms to measure airway and parenchymal metrics from non-contrast CTs. By bringing the models to their curated data, the company eliminates the need to export and send the data out to a third party, saving time and maintaining data security and provenance.
Automated Curation
A top five biopharma company automates data processing at ingest for multiple neurology modalities, including de-identification. Then, individual processing pipelines are applied, triggering human-in-the-loop review for curated, analysis-ready datasets.
How Can My Biopharma Company Optimize Efficiency and Add Value to Imaging Data?
These are just examples of what some of our customers have been able to accomplish thus far with Flywheel. But we’re always ready to discuss individual needs and determine a solution that will help you achieve your goals around clinical research and AI development. Get in touch with one of our experts to start the conversation.
