Insights on Research Data Management

Medical devices help clinicians and patients by enhancing diagnostics, treatment, monitoring and overall delivery of quality healthcare. But manufacturers face numerous challenges when developing new devices, in particular when it comes to best utilizing medical imaging and video data. Below, we’ve rounded up some of the top challenges medical device manufacturers face with regard to medical imaging. 

Slow R&D and product development 

Medical device research and testing can take years—by some estimates, three to seven years. New devices need to be rigorously tested to ensure quality control before going to market. Depending on the class of device, this can also involve clinical trials involving thousands of human participants—say, if it’s a device that will be implanted in the human body. Given this process, it can be not only time-consuming but expensive to access the diverse, real-world imaging data needed to adequately test and perfect novel devices.

A centralized imaging data management platform can speed this process along by accelerating data retrieval, validation and analysis. This lets you aggregate, view and analyze loads of imaging data in one location, on one screen, helping you move from idea to prototype to submission faster by making each step of the process more efficient.  

Data silos and interoperability  

Collecting all this data often involves multiple departments, institutions and locations in order to get the robust datasets needed for FDA approval. But medical imaging data is often scattered across different systems, including PACS, EMRs and cloud storage, making integration and accessibility difficult. Some of this data may even live on laptops and be transferred via flash drives, which isn’t ideal for either access or data privacy. 

That’s why it’s crucial to work with a platform that can be reached by various parties (with access controls, of course, to adhere to data privacy standards). This platform should also let you harmonize data, including CT, PET and MRI, cleanly converting it into the DICOM standard and back again, so you can compare apples to apples. 

Regulatory compliance and data security 

Strict regulations, such as HIPAA, GDPR, 21 CFR Part 11, and SOC2 and SOC3 compliance, require manufacturers to ensure data security and auditability. Imaging data needs to be de-identified, access to PHI needs to be strictly controlled, and changes to data need to be logged to adhere to these regulations, among other measures. 

The medical platform your research and development team uses needs to be built with data privacy in mind, allowing for easy de-identification and creation of audit trails. Doing this helps manufacturers ensure they won’t hit snags in the submission process when they come up against costly violations or breaches that can end up sinking promising developments. 

Scalability and storage costs 

As devices move through the phases of clinical trials, more data needs to be collected for submission—meaning greater storage costs. The platform you use to store and access data collected during clinical trials should allow for cloud-based storage so it can scale with your needs.  

Cloud storage is generally cheaper than on-premise by about 30-40%. And with added scale, you can keep all that valuable data collected during clinical trials for downstream uses, such as marketing, research and development of new devices, and optimization of current devices on the market. 

AI and machine learning integration 

It isn’t just data capture and management that takes up valuable time, but data processing as well. It’s important to leverage AI algorithms to cut the time it takes to process and analyze imaging data, for faster results. But putting these algorithms in place is another story. 

A platform like Flywheel allows researchers to deploy containerized algorithms as ready-made Gears to automate processing pipelines. This helps them annotate, de-ID and extract metadata from patient data faster, among other uses. Flywheel even lets medical device developers create their own Gears, with help from our Professional Services team. And the data collected in Flywheel can easily be exported as cohorts for further annotation, training and validation of AI models—which can then be fed back into medical devices for greater efficiency and differentiation in the market.

Why Flywheel? 

Flywheel is the only platform that addresses all of these challenges facing medical device manufacturers. With our secure, web-based UI, CVAT integration, automation and cloud storage options, we empower research teams to ingest, access, aggregate and analyze the copious amount of imaging and video data needed to develop new medical devices, all while maintaining compliance with protocols and regulations such as HIPAA, SOC 2 and SOC 3, GDPR and 21 CFR Part 11. 

We’ve worked with top manufacturers such as Siemens Healthineers to find the exact imaging datasets it needs for analysis, allowing for superior upload, access and project management capabilities. In one instance, Siemens was able to identify 200 relevant cases of liver lesions among all its imaging data and tailor software to run tests on it. 

What could you do with the ability to quickly ingest, retrieve and analyze the data you need, when you need it?  

Flywheel is here to help you develop the next breakthrough in medical device technology. Reach out to our team for a discovery call to find out if we can help you accelerate your device development workflows for faster time to market.