Suzhou International Expo Centre

Medtec Innovation Suzhou

2024.12.23-24 | Hall A Suzhou International Expo Centre

EN | 中文
   

3 Ways Privacy-Enhancing Technology Will Improve Medical Products Innovation

While a massive amount of healthcare data is produced every day, comprising nearly 30 percent of all data being created around the world, differing privacy regulations and forms of storage of health data have slowed innovation.

Privacy-enhancing technologies (PET) mathematically enforce privacy for data and algorithms, allowing companies to collaborate around sensitive information without the privacy and legal risks.

During collaboration using PET, data and algorithms are one-way encrypted with no possibility of decryption, thus the information is de-identified, eliminating the risk of patient data ever being re-identified. More effective PET technologies allow for the use of the data and execution of the algorithms in that one-way encrypted state.

Here are three ways that wider adoption of PET stands to improve collaboration in healthcare and medical product manufacturing, and increase innovation:

1. Creating More Accessible Products and Equipment

Quality and design risk often ranks as one of the top challenges for medical device makers.

Will the device fit the average person, and how can designers best determine what is “average?” It’s impractical, nearly impossible for most companies, to test every design on thousands of actual people, but it’s not impossible to design better products and improve the quality of medical devices by using real patient data including imaging data, health records, and other information.

Historically, the problem is gathering and using that protected data from healthcare partners and others in the broader healthcare industry. At times, the data comes from across borders, requiring compliance with multiple data privacy laws based on industry, location and other features such as content. Healthcare enterprises have just begun to scratch the surface of internationally collaborating using patient data.

2. Ability to Safeguard PHI, PII and Other Protected Data

HIPAA in the U.S. and GDPR in the EU set the requirements for using and safeguarding PHI and PII. When it comes to stewarding patient data, many smaller medical companies have difficulty ensuring compliance when working with partners, sub-contractors and other interested parties. BAAs and expensive third-party risk management programs will never absolve you of penalties, bad press and other repercussions if one of your partners loses, exposes or otherwise mismanages protected data. Even sharing your own protected data internally presents similar challenges.

There is no doubt the effort to protect data while still ensuring the quality and validity of the data is a challenge when it comes to using PHI/PII for making better and higher quality products. For example, the data needed to build a better heart-related, implanted medical device, such as a pacemaker or a defibrillator, might require the hospital’s records of select patients, radiology’s dicom images, pharmacy’s prescription records and other data. This might also include protected but unregulated data such as the intellectual property of the battery subcontractor.

3. Faster and Cheaper Product Creation and Ongoing Validation

The ability to collaborate quickly and safely with any kind and type of data is key to building better products faster and cheaper. Privacy enhancing technologies help speed the data collaboration processes both technically and with respect to your partner’s data governance and information security requirements. If partners can manage the security and privacy of their own data and guarantee that the users of the data will never get access to PHI/PII or IP, the cost and complexity of data sharing drop while the accuracy and speed of data increases.

PET builds on well understood mathematical principles, such as federated learning and multi-party computation, and compares favorably with other privacy preserving technologies, such as homomorphic encryption, synthetic data and tokenization. These mathematical approaches help guarantee privacy.

Through PET, de-identified source files, like X-Ray image data, become more easily accessible without having to navigate tricky Business Associate Agreements and privacy regulations. More quality data results in more accurate AI models for training medical products, making training quicker and cheaper.

PET also allows for multi-modal collaboration using any type of data, including image, text, voice, video and more, allowing source data from multiple different medical products to be used to better create and train predictive and generalizable AI models. And with less time needed for research and development overall, PET enables manufacturers to create better products and get them into the market faster. Once available to the market, PET allows for seamless, ongoing reporting and analysis for product validation.

Conclusion

For medical product manufacturers and their information-supply chain partners, PET unlocks innovation, speed and faster paths to revenue by allowing analysis and use of real patient data from any partner as well as guarded industry secrets – all while letting everyone keep their data private. Article source: MPO

X